diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,140988 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 10.0,
+  "eval_steps": 500,
+  "global_step": 201340,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 4.966722956193504e-05,
+      "grad_norm": 3.921875,
+      "learning_rate": 0.0008,
+      "loss": 3.0588,
+      "step": 1
+    },
+    {
+      "epoch": 0.0004966722956193504,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0007999642395947154,
+      "loss": 2.3153,
+      "step": 10
+    },
+    {
+      "epoch": 0.0009933445912387007,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0007999245058110659,
+      "loss": 1.8684,
+      "step": 20
+    },
+    {
+      "epoch": 0.001490016886858051,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0007998847720274163,
+      "loss": 1.8397,
+      "step": 30
+    },
+    {
+      "epoch": 0.0019866891824774015,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0007998450382437668,
+      "loss": 1.7934,
+      "step": 40
+    },
+    {
+      "epoch": 0.0024833614780967518,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0007998053044601173,
+      "loss": 1.7761,
+      "step": 50
+    },
+    {
+      "epoch": 0.002980033773716102,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0007997655706764677,
+      "loss": 1.7596,
+      "step": 60
+    },
+    {
+      "epoch": 0.0034767060693354523,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.0007997258368928181,
+      "loss": 1.7166,
+      "step": 70
+    },
+    {
+      "epoch": 0.003973378364954803,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0007996861031091686,
+      "loss": 1.6598,
+      "step": 80
+    },
+    {
+      "epoch": 0.004470050660574153,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0007996463693255191,
+      "loss": 1.6145,
+      "step": 90
+    },
+    {
+      "epoch": 0.0049667229561935035,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0007996066355418695,
+      "loss": 1.5857,
+      "step": 100
+    },
+    {
+      "epoch": 0.005463395251812854,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00079956690175822,
+      "loss": 1.5669,
+      "step": 110
+    },
+    {
+      "epoch": 0.005960067547432204,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0007995271679745704,
+      "loss": 1.5253,
+      "step": 120
+    },
+    {
+      "epoch": 0.006456739843051554,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0007994874341909208,
+      "loss": 1.5324,
+      "step": 130
+    },
+    {
+      "epoch": 0.006953412138670905,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0007994477004072714,
+      "loss": 1.4808,
+      "step": 140
+    },
+    {
+      "epoch": 0.007450084434290256,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0007994079666236218,
+      "loss": 1.4696,
+      "step": 150
+    },
+    {
+      "epoch": 0.007946756729909606,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0007993682328399723,
+      "loss": 1.45,
+      "step": 160
+    },
+    {
+      "epoch": 0.008443429025528956,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0007993284990563226,
+      "loss": 1.4663,
+      "step": 170
+    },
+    {
+      "epoch": 0.008940101321148307,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0007992887652726731,
+      "loss": 1.4412,
+      "step": 180
+    },
+    {
+      "epoch": 0.009436773616767657,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0007992490314890237,
+      "loss": 1.3908,
+      "step": 190
+    },
+    {
+      "epoch": 0.009933445912387007,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.000799209297705374,
+      "loss": 1.3938,
+      "step": 200
+    },
+    {
+      "epoch": 0.010430118208006357,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0007991695639217245,
+      "loss": 1.3474,
+      "step": 210
+    },
+    {
+      "epoch": 0.010926790503625708,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0007991298301380749,
+      "loss": 1.3319,
+      "step": 220
+    },
+    {
+      "epoch": 0.011423462799245058,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0007990900963544253,
+      "loss": 1.3428,
+      "step": 230
+    },
+    {
+      "epoch": 0.011920135094864408,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0007990503625707759,
+      "loss": 1.3505,
+      "step": 240
+    },
+    {
+      "epoch": 0.012416807390483758,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0007990106287871263,
+      "loss": 1.3445,
+      "step": 250
+    },
+    {
+      "epoch": 0.012913479686103109,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0007989708950034767,
+      "loss": 1.3059,
+      "step": 260
+    },
+    {
+      "epoch": 0.013410151981722459,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0007989311612198272,
+      "loss": 1.3184,
+      "step": 270
+    },
+    {
+      "epoch": 0.01390682427734181,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0007988914274361776,
+      "loss": 1.2753,
+      "step": 280
+    },
+    {
+      "epoch": 0.01440349657296116,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0007988516936525282,
+      "loss": 1.2823,
+      "step": 290
+    },
+    {
+      "epoch": 0.014900168868580511,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0007988119598688786,
+      "loss": 1.3007,
+      "step": 300
+    },
+    {
+      "epoch": 0.015396841164199862,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000798772226085229,
+      "loss": 1.2959,
+      "step": 310
+    },
+    {
+      "epoch": 0.015893513459819212,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0007987324923015795,
+      "loss": 1.2481,
+      "step": 320
+    },
+    {
+      "epoch": 0.01639018575543856,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0007986927585179299,
+      "loss": 1.2824,
+      "step": 330
+    },
+    {
+      "epoch": 0.016886858051057912,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0007986530247342804,
+      "loss": 1.242,
+      "step": 340
+    },
+    {
+      "epoch": 0.01738353034667726,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0007986132909506309,
+      "loss": 1.2588,
+      "step": 350
+    },
+    {
+      "epoch": 0.017880202642296613,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0007985735571669812,
+      "loss": 1.2736,
+      "step": 360
+    },
+    {
+      "epoch": 0.01837687493791596,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0007985338233833317,
+      "loss": 1.2348,
+      "step": 370
+    },
+    {
+      "epoch": 0.018873547233535314,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0007984940895996822,
+      "loss": 1.2293,
+      "step": 380
+    },
+    {
+      "epoch": 0.019370219529154662,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0007984543558160325,
+      "loss": 1.1832,
+      "step": 390
+    },
+    {
+      "epoch": 0.019866891824774014,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0007984146220323831,
+      "loss": 1.2007,
+      "step": 400
+    },
+    {
+      "epoch": 0.020363564120393366,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0007983748882487336,
+      "loss": 1.2085,
+      "step": 410
+    },
+    {
+      "epoch": 0.020860236416012715,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0007983351544650839,
+      "loss": 1.2127,
+      "step": 420
+    },
+    {
+      "epoch": 0.021356908711632067,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0007982954206814344,
+      "loss": 1.1983,
+      "step": 430
+    },
+    {
+      "epoch": 0.021853581007251415,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0007982556868977848,
+      "loss": 1.2332,
+      "step": 440
+    },
+    {
+      "epoch": 0.022350253302870767,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0007982159531141354,
+      "loss": 1.1835,
+      "step": 450
+    },
+    {
+      "epoch": 0.022846925598490116,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0007981762193304858,
+      "loss": 1.21,
+      "step": 460
+    },
+    {
+      "epoch": 0.023343597894109468,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0007981364855468362,
+      "loss": 1.1797,
+      "step": 470
+    },
+    {
+      "epoch": 0.023840270189728816,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0007980967517631867,
+      "loss": 1.1599,
+      "step": 480
+    },
+    {
+      "epoch": 0.024336942485348168,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0007980570179795371,
+      "loss": 1.1898,
+      "step": 490
+    },
+    {
+      "epoch": 0.024833614780967517,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0007980172841958876,
+      "loss": 1.1622,
+      "step": 500
+    },
+    {
+      "epoch": 0.02533028707658687,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0007979775504122381,
+      "loss": 1.182,
+      "step": 510
+    },
+    {
+      "epoch": 0.025826959372206217,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0007979378166285885,
+      "loss": 1.1502,
+      "step": 520
+    },
+    {
+      "epoch": 0.02632363166782557,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0007978980828449389,
+      "loss": 1.1264,
+      "step": 530
+    },
+    {
+      "epoch": 0.026820303963444918,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0007978583490612895,
+      "loss": 1.1711,
+      "step": 540
+    },
+    {
+      "epoch": 0.02731697625906427,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0007978186152776398,
+      "loss": 1.1233,
+      "step": 550
+    },
+    {
+      "epoch": 0.02781364855468362,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0007977788814939903,
+      "loss": 1.1435,
+      "step": 560
+    },
+    {
+      "epoch": 0.02831032085030297,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0007977391477103408,
+      "loss": 1.1424,
+      "step": 570
+    },
+    {
+      "epoch": 0.02880699314592232,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0007976994139266911,
+      "loss": 1.1642,
+      "step": 580
+    },
+    {
+      "epoch": 0.02930366544154167,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0007976596801430416,
+      "loss": 1.11,
+      "step": 590
+    },
+    {
+      "epoch": 0.029800337737161023,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0007976199463593922,
+      "loss": 1.152,
+      "step": 600
+    },
+    {
+      "epoch": 0.03029701003278037,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0007975802125757426,
+      "loss": 1.1054,
+      "step": 610
+    },
+    {
+      "epoch": 0.030793682328399723,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.000797540478792093,
+      "loss": 1.1078,
+      "step": 620
+    },
+    {
+      "epoch": 0.03129035462401907,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0007975007450084434,
+      "loss": 1.1475,
+      "step": 630
+    },
+    {
+      "epoch": 0.031787026919638424,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.000797461011224794,
+      "loss": 1.1248,
+      "step": 640
+    },
+    {
+      "epoch": 0.032283699215257776,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0007974212774411444,
+      "loss": 1.1052,
+      "step": 650
+    },
+    {
+      "epoch": 0.03278037151087712,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0007973815436574948,
+      "loss": 1.1349,
+      "step": 660
+    },
+    {
+      "epoch": 0.03327704380649647,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0007973418098738453,
+      "loss": 1.1056,
+      "step": 670
+    },
+    {
+      "epoch": 0.033773716102115825,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0007973020760901957,
+      "loss": 1.1263,
+      "step": 680
+    },
+    {
+      "epoch": 0.03427038839773518,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0007972623423065461,
+      "loss": 1.1334,
+      "step": 690
+    },
+    {
+      "epoch": 0.03476706069335452,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0007972226085228967,
+      "loss": 1.1171,
+      "step": 700
+    },
+    {
+      "epoch": 0.035263732988973874,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0007971828747392471,
+      "loss": 1.1007,
+      "step": 710
+    },
+    {
+      "epoch": 0.035760405284593226,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0007971431409555975,
+      "loss": 1.1092,
+      "step": 720
+    },
+    {
+      "epoch": 0.03625707758021258,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.000797103407171948,
+      "loss": 1.1299,
+      "step": 730
+    },
+    {
+      "epoch": 0.03675374987583192,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0007970636733882984,
+      "loss": 1.1215,
+      "step": 740
+    },
+    {
+      "epoch": 0.037250422171451275,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0007970239396046489,
+      "loss": 1.0959,
+      "step": 750
+    },
+    {
+      "epoch": 0.03774709446707063,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0007969842058209994,
+      "loss": 1.0947,
+      "step": 760
+    },
+    {
+      "epoch": 0.03824376676268998,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0007969444720373498,
+      "loss": 1.0743,
+      "step": 770
+    },
+    {
+      "epoch": 0.038740439058309324,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0007969047382537002,
+      "loss": 1.0936,
+      "step": 780
+    },
+    {
+      "epoch": 0.039237111353928676,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0007968650044700507,
+      "loss": 1.0967,
+      "step": 790
+    },
+    {
+      "epoch": 0.03973378364954803,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0007968252706864012,
+      "loss": 1.0531,
+      "step": 800
+    },
+    {
+      "epoch": 0.04023045594516738,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0007967855369027516,
+      "loss": 1.0668,
+      "step": 810
+    },
+    {
+      "epoch": 0.04072712824078673,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.000796745803119102,
+      "loss": 1.0908,
+      "step": 820
+    },
+    {
+      "epoch": 0.04122380053640608,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0007967060693354525,
+      "loss": 1.0599,
+      "step": 830
+    },
+    {
+      "epoch": 0.04172047283202543,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0007966663355518029,
+      "loss": 1.0601,
+      "step": 840
+    },
+    {
+      "epoch": 0.04221714512764478,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0007966266017681534,
+      "loss": 1.0303,
+      "step": 850
+    },
+    {
+      "epoch": 0.04271381742326413,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0007965868679845039,
+      "loss": 1.0721,
+      "step": 860
+    },
+    {
+      "epoch": 0.04321048971888348,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0007965471342008543,
+      "loss": 1.0544,
+      "step": 870
+    },
+    {
+      "epoch": 0.04370716201450283,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0007965074004172047,
+      "loss": 1.0499,
+      "step": 880
+    },
+    {
+      "epoch": 0.04420383431012218,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0007964676666335552,
+      "loss": 1.0578,
+      "step": 890
+    },
+    {
+      "epoch": 0.044700506605741534,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0007964279328499057,
+      "loss": 1.103,
+      "step": 900
+    },
+    {
+      "epoch": 0.04519717890136088,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0007963881990662561,
+      "loss": 1.0236,
+      "step": 910
+    },
+    {
+      "epoch": 0.04569385119698023,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0007963484652826066,
+      "loss": 1.0282,
+      "step": 920
+    },
+    {
+      "epoch": 0.04619052349259958,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.000796308731498957,
+      "loss": 1.0686,
+      "step": 930
+    },
+    {
+      "epoch": 0.046687195788218935,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0007962689977153074,
+      "loss": 1.0468,
+      "step": 940
+    },
+    {
+      "epoch": 0.04718386808383828,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.000796229263931658,
+      "loss": 1.0379,
+      "step": 950
+    },
+    {
+      "epoch": 0.04768054037945763,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0007961895301480084,
+      "loss": 1.0553,
+      "step": 960
+    },
+    {
+      "epoch": 0.048177212675076984,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0007961497963643588,
+      "loss": 1.0768,
+      "step": 970
+    },
+    {
+      "epoch": 0.048673884970696336,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0007961100625807093,
+      "loss": 1.0261,
+      "step": 980
+    },
+    {
+      "epoch": 0.04917055726631568,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0007960703287970597,
+      "loss": 1.0311,
+      "step": 990
+    },
+    {
+      "epoch": 0.04966722956193503,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0007960305950134102,
+      "loss": 1.0556,
+      "step": 1000
+    },
+    {
+      "epoch": 0.050163901857554385,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0007959908612297607,
+      "loss": 1.034,
+      "step": 1010
+    },
+    {
+      "epoch": 0.05066057415317374,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0007959511274461111,
+      "loss": 1.0009,
+      "step": 1020
+    },
+    {
+      "epoch": 0.05115724644879309,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0007959113936624615,
+      "loss": 1.0109,
+      "step": 1030
+    },
+    {
+      "epoch": 0.051653918744412435,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0007958716598788119,
+      "loss": 1.0085,
+      "step": 1040
+    },
+    {
+      "epoch": 0.052150591040031787,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0007958319260951625,
+      "loss": 1.0898,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05264726333565114,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.000795792192311513,
+      "loss": 1.0379,
+      "step": 1060
+    },
+    {
+      "epoch": 0.05314393563127049,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0007957524585278633,
+      "loss": 0.9888,
+      "step": 1070
+    },
+    {
+      "epoch": 0.053640607926889836,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0007957127247442138,
+      "loss": 0.9989,
+      "step": 1080
+    },
+    {
+      "epoch": 0.05413728022250919,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0007956729909605642,
+      "loss": 1.046,
+      "step": 1090
+    },
+    {
+      "epoch": 0.05463395251812854,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0007956332571769147,
+      "loss": 0.9776,
+      "step": 1100
+    },
+    {
+      "epoch": 0.05513062481374789,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0007955935233932652,
+      "loss": 1.0245,
+      "step": 1110
+    },
+    {
+      "epoch": 0.05562729710936724,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0007955537896096156,
+      "loss": 1.0557,
+      "step": 1120
+    },
+    {
+      "epoch": 0.05612396940498659,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.000795514055825966,
+      "loss": 1.0043,
+      "step": 1130
+    },
+    {
+      "epoch": 0.05662064170060594,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0007954743220423165,
+      "loss": 0.9955,
+      "step": 1140
+    },
+    {
+      "epoch": 0.05711731399622529,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.000795434588258667,
+      "loss": 1.0221,
+      "step": 1150
+    },
+    {
+      "epoch": 0.05761398629184464,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0007953948544750174,
+      "loss": 0.9994,
+      "step": 1160
+    },
+    {
+      "epoch": 0.05811065858746399,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0007953551206913679,
+      "loss": 1.031,
+      "step": 1170
+    },
+    {
+      "epoch": 0.05860733088308334,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0007953153869077183,
+      "loss": 0.9968,
+      "step": 1180
+    },
+    {
+      "epoch": 0.059104003178702694,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0007952756531240688,
+      "loss": 1.0061,
+      "step": 1190
+    },
+    {
+      "epoch": 0.059600675474322046,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0007952359193404193,
+      "loss": 1.0162,
+      "step": 1200
+    },
+    {
+      "epoch": 0.06009734776994139,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0007951961855567697,
+      "loss": 0.9925,
+      "step": 1210
+    },
+    {
+      "epoch": 0.06059402006556074,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0007951564517731202,
+      "loss": 1.0197,
+      "step": 1220
+    },
+    {
+      "epoch": 0.061090692361180095,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0007951167179894705,
+      "loss": 1.0029,
+      "step": 1230
+    },
+    {
+      "epoch": 0.06158736465679945,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.000795076984205821,
+      "loss": 0.9652,
+      "step": 1240
+    },
+    {
+      "epoch": 0.06208403695241879,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0007950372504221716,
+      "loss": 1.0034,
+      "step": 1250
+    },
+    {
+      "epoch": 0.06258070924803814,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0007949975166385219,
+      "loss": 1.0125,
+      "step": 1260
+    },
+    {
+      "epoch": 0.0630773815436575,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0007949577828548724,
+      "loss": 0.999,
+      "step": 1270
+    },
+    {
+      "epoch": 0.06357405383927685,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0007949180490712229,
+      "loss": 0.9831,
+      "step": 1280
+    },
+    {
+      "epoch": 0.0640707261348962,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0007948783152875732,
+      "loss": 0.9889,
+      "step": 1290
+    },
+    {
+      "epoch": 0.06456739843051555,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0007948385815039238,
+      "loss": 0.9954,
+      "step": 1300
+    },
+    {
+      "epoch": 0.06506407072613489,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0007947988477202742,
+      "loss": 1.0029,
+      "step": 1310
+    },
+    {
+      "epoch": 0.06556074302175424,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0007947591139366246,
+      "loss": 0.9666,
+      "step": 1320
+    },
+    {
+      "epoch": 0.0660574153173736,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0007947193801529751,
+      "loss": 0.9921,
+      "step": 1330
+    },
+    {
+      "epoch": 0.06655408761299295,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0007946796463693255,
+      "loss": 0.9799,
+      "step": 1340
+    },
+    {
+      "epoch": 0.0670507599086123,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0007946399125856761,
+      "loss": 1.0026,
+      "step": 1350
+    },
+    {
+      "epoch": 0.06754743220423165,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0007946001788020265,
+      "loss": 0.9782,
+      "step": 1360
+    },
+    {
+      "epoch": 0.068044104499851,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0007945604450183769,
+      "loss": 1.0086,
+      "step": 1370
+    },
+    {
+      "epoch": 0.06854077679547035,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0007945207112347274,
+      "loss": 0.9763,
+      "step": 1380
+    },
+    {
+      "epoch": 0.0690374490910897,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0007944809774510778,
+      "loss": 1.0011,
+      "step": 1390
+    },
+    {
+      "epoch": 0.06953412138670904,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0007944412436674283,
+      "loss": 0.9751,
+      "step": 1400
+    },
+    {
+      "epoch": 0.0700307936823284,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0007944015098837788,
+      "loss": 1.0073,
+      "step": 1410
+    },
+    {
+      "epoch": 0.07052746597794775,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0007943617761001291,
+      "loss": 0.9551,
+      "step": 1420
+    },
+    {
+      "epoch": 0.0710241382735671,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0007943220423164796,
+      "loss": 0.9414,
+      "step": 1430
+    },
+    {
+      "epoch": 0.07152081056918645,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0007942823085328301,
+      "loss": 0.9908,
+      "step": 1440
+    },
+    {
+      "epoch": 0.0720174828648058,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0007942425747491804,
+      "loss": 0.9451,
+      "step": 1450
+    },
+    {
+      "epoch": 0.07251415516042516,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.000794202840965531,
+      "loss": 0.9761,
+      "step": 1460
+    },
+    {
+      "epoch": 0.07301082745604451,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0007941631071818815,
+      "loss": 0.9389,
+      "step": 1470
+    },
+    {
+      "epoch": 0.07350749975166385,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0007941233733982319,
+      "loss": 0.9704,
+      "step": 1480
+    },
+    {
+      "epoch": 0.0740041720472832,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0007940836396145823,
+      "loss": 0.9748,
+      "step": 1490
+    },
+    {
+      "epoch": 0.07450084434290255,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0007940439058309328,
+      "loss": 0.982,
+      "step": 1500
+    },
+    {
+      "epoch": 0.0749975166385219,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0007940041720472833,
+      "loss": 0.9661,
+      "step": 1510
+    },
+    {
+      "epoch": 0.07549418893414125,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0007939644382636337,
+      "loss": 0.9424,
+      "step": 1520
+    },
+    {
+      "epoch": 0.0759908612297606,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0007939247044799841,
+      "loss": 0.9927,
+      "step": 1530
+    },
+    {
+      "epoch": 0.07648753352537996,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0007938849706963346,
+      "loss": 0.9794,
+      "step": 1540
+    },
+    {
+      "epoch": 0.07698420582099931,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.000793845236912685,
+      "loss": 0.9505,
+      "step": 1550
+    },
+    {
+      "epoch": 0.07748087811661865,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0007938055031290355,
+      "loss": 0.9565,
+      "step": 1560
+    },
+    {
+      "epoch": 0.077977550412238,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.000793765769345386,
+      "loss": 0.9998,
+      "step": 1570
+    },
+    {
+      "epoch": 0.07847422270785735,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0007937260355617364,
+      "loss": 0.9816,
+      "step": 1580
+    },
+    {
+      "epoch": 0.0789708950034767,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0007936863017780868,
+      "loss": 0.9523,
+      "step": 1590
+    },
+    {
+      "epoch": 0.07946756729909606,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0007936465679944374,
+      "loss": 0.9512,
+      "step": 1600
+    },
+    {
+      "epoch": 0.07996423959471541,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0007936068342107878,
+      "loss": 0.9186,
+      "step": 1610
+    },
+    {
+      "epoch": 0.08046091189033476,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0007935671004271382,
+      "loss": 0.9518,
+      "step": 1620
+    },
+    {
+      "epoch": 0.08095758418595411,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0007935273666434887,
+      "loss": 0.9269,
+      "step": 1630
+    },
+    {
+      "epoch": 0.08145425648157346,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0007934876328598391,
+      "loss": 0.9667,
+      "step": 1640
+    },
+    {
+      "epoch": 0.0819509287771928,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0007934478990761896,
+      "loss": 0.947,
+      "step": 1650
+    },
+    {
+      "epoch": 0.08244760107281215,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0007934081652925401,
+      "loss": 0.9563,
+      "step": 1660
+    },
+    {
+      "epoch": 0.0829442733684315,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0007933684315088905,
+      "loss": 0.9397,
+      "step": 1670
+    },
+    {
+      "epoch": 0.08344094566405086,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0007933286977252409,
+      "loss": 0.965,
+      "step": 1680
+    },
+    {
+      "epoch": 0.08393761795967021,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0007932889639415913,
+      "loss": 0.9461,
+      "step": 1690
+    },
+    {
+      "epoch": 0.08443429025528956,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0007932492301579419,
+      "loss": 0.9537,
+      "step": 1700
+    },
+    {
+      "epoch": 0.08493096255090891,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0007932094963742923,
+      "loss": 0.9146,
+      "step": 1710
+    },
+    {
+      "epoch": 0.08542763484652827,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0007931697625906427,
+      "loss": 0.9558,
+      "step": 1720
+    },
+    {
+      "epoch": 0.0859243071421476,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0007931300288069932,
+      "loss": 0.9614,
+      "step": 1730
+    },
+    {
+      "epoch": 0.08642097943776696,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0007930902950233436,
+      "loss": 0.9227,
+      "step": 1740
+    },
+    {
+      "epoch": 0.08691765173338631,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.000793050561239694,
+      "loss": 0.9533,
+      "step": 1750
+    },
+    {
+      "epoch": 0.08741432402900566,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0007930108274560446,
+      "loss": 0.9488,
+      "step": 1760
+    },
+    {
+      "epoch": 0.08791099632462501,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.000792971093672395,
+      "loss": 0.9396,
+      "step": 1770
+    },
+    {
+      "epoch": 0.08840766862024436,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0007929313598887454,
+      "loss": 0.9199,
+      "step": 1780
+    },
+    {
+      "epoch": 0.08890434091586372,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0007928916261050959,
+      "loss": 0.9593,
+      "step": 1790
+    },
+    {
+      "epoch": 0.08940101321148307,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0007928518923214463,
+      "loss": 0.9639,
+      "step": 1800
+    },
+    {
+      "epoch": 0.08989768550710242,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0007928121585377968,
+      "loss": 0.9267,
+      "step": 1810
+    },
+    {
+      "epoch": 0.09039435780272176,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0007927724247541473,
+      "loss": 0.9629,
+      "step": 1820
+    },
+    {
+      "epoch": 0.09089103009834111,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0007927326909704977,
+      "loss": 0.9051,
+      "step": 1830
+    },
+    {
+      "epoch": 0.09138770239396046,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0007926929571868481,
+      "loss": 0.9445,
+      "step": 1840
+    },
+    {
+      "epoch": 0.09188437468957981,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0007926532234031987,
+      "loss": 0.9441,
+      "step": 1850
+    },
+    {
+      "epoch": 0.09238104698519917,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0007926134896195491,
+      "loss": 0.9459,
+      "step": 1860
+    },
+    {
+      "epoch": 0.09287771928081852,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0007925737558358995,
+      "loss": 0.944,
+      "step": 1870
+    },
+    {
+      "epoch": 0.09337439157643787,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00079253402205225,
+      "loss": 0.9664,
+      "step": 1880
+    },
+    {
+      "epoch": 0.09387106387205722,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0007924942882686004,
+      "loss": 0.9384,
+      "step": 1890
+    },
+    {
+      "epoch": 0.09436773616767656,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0007924545544849508,
+      "loss": 0.936,
+      "step": 1900
+    },
+    {
+      "epoch": 0.09486440846329591,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0007924148207013013,
+      "loss": 0.9484,
+      "step": 1910
+    },
+    {
+      "epoch": 0.09536108075891526,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0007923750869176518,
+      "loss": 0.8862,
+      "step": 1920
+    },
+    {
+      "epoch": 0.09585775305453462,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0007923353531340023,
+      "loss": 0.933,
+      "step": 1930
+    },
+    {
+      "epoch": 0.09635442535015397,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0007922956193503526,
+      "loss": 0.9179,
+      "step": 1940
+    },
+    {
+      "epoch": 0.09685109764577332,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0007922558855667031,
+      "loss": 0.9129,
+      "step": 1950
+    },
+    {
+      "epoch": 0.09734776994139267,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0007922161517830536,
+      "loss": 0.9578,
+      "step": 1960
+    },
+    {
+      "epoch": 0.09784444223701202,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.000792176417999404,
+      "loss": 0.8939,
+      "step": 1970
+    },
+    {
+      "epoch": 0.09834111453263136,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0007921366842157545,
+      "loss": 0.9061,
+      "step": 1980
+    },
+    {
+      "epoch": 0.09883778682825071,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0007920969504321049,
+      "loss": 0.9174,
+      "step": 1990
+    },
+    {
+      "epoch": 0.09933445912387007,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0007920572166484553,
+      "loss": 0.9491,
+      "step": 2000
+    },
+    {
+      "epoch": 0.09983113141948942,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0007920174828648059,
+      "loss": 0.9257,
+      "step": 2010
+    },
+    {
+      "epoch": 0.10032780371510877,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0007919777490811563,
+      "loss": 0.8943,
+      "step": 2020
+    },
+    {
+      "epoch": 0.10082447601072812,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0007919380152975067,
+      "loss": 0.9324,
+      "step": 2030
+    },
+    {
+      "epoch": 0.10132114830634747,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0007918982815138572,
+      "loss": 0.9209,
+      "step": 2040
+    },
+    {
+      "epoch": 0.10181782060196683,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0007918585477302076,
+      "loss": 0.9143,
+      "step": 2050
+    },
+    {
+      "epoch": 0.10231449289758618,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0007918188139465581,
+      "loss": 0.8881,
+      "step": 2060
+    },
+    {
+      "epoch": 0.10281116519320552,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0007917790801629086,
+      "loss": 0.9361,
+      "step": 2070
+    },
+    {
+      "epoch": 0.10330783748882487,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.000791739346379259,
+      "loss": 0.8759,
+      "step": 2080
+    },
+    {
+      "epoch": 0.10380450978444422,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0007916996125956095,
+      "loss": 0.9115,
+      "step": 2090
+    },
+    {
+      "epoch": 0.10430118208006357,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0007916598788119598,
+      "loss": 0.9272,
+      "step": 2100
+    },
+    {
+      "epoch": 0.10479785437568293,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0007916201450283104,
+      "loss": 0.9501,
+      "step": 2110
+    },
+    {
+      "epoch": 0.10529452667130228,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0007915804112446609,
+      "loss": 0.8966,
+      "step": 2120
+    },
+    {
+      "epoch": 0.10579119896692163,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0007915406774610112,
+      "loss": 0.9069,
+      "step": 2130
+    },
+    {
+      "epoch": 0.10628787126254098,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0007915009436773617,
+      "loss": 0.9063,
+      "step": 2140
+    },
+    {
+      "epoch": 0.10678454355816032,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0007914612098937123,
+      "loss": 0.9479,
+      "step": 2150
+    },
+    {
+      "epoch": 0.10728121585377967,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0007914214761100626,
+      "loss": 0.8811,
+      "step": 2160
+    },
+    {
+      "epoch": 0.10777788814939902,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0007913817423264131,
+      "loss": 0.9181,
+      "step": 2170
+    },
+    {
+      "epoch": 0.10827456044501838,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0007913420085427635,
+      "loss": 0.8811,
+      "step": 2180
+    },
+    {
+      "epoch": 0.10877123274063773,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0007913022747591139,
+      "loss": 0.8769,
+      "step": 2190
+    },
+    {
+      "epoch": 0.10926790503625708,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0007912625409754644,
+      "loss": 0.9298,
+      "step": 2200
+    },
+    {
+      "epoch": 0.10976457733187643,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0007912228071918149,
+      "loss": 0.8929,
+      "step": 2210
+    },
+    {
+      "epoch": 0.11026124962749578,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0007911830734081653,
+      "loss": 0.9101,
+      "step": 2220
+    },
+    {
+      "epoch": 0.11075792192311514,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0007911433396245158,
+      "loss": 0.9398,
+      "step": 2230
+    },
+    {
+      "epoch": 0.11125459421873447,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0007911036058408662,
+      "loss": 0.9077,
+      "step": 2240
+    },
+    {
+      "epoch": 0.11175126651435383,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0007910638720572167,
+      "loss": 0.9285,
+      "step": 2250
+    },
+    {
+      "epoch": 0.11224793880997318,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0007910241382735672,
+      "loss": 0.8824,
+      "step": 2260
+    },
+    {
+      "epoch": 0.11274461110559253,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0007909844044899176,
+      "loss": 0.8869,
+      "step": 2270
+    },
+    {
+      "epoch": 0.11324128340121188,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0007909446707062681,
+      "loss": 0.9341,
+      "step": 2280
+    },
+    {
+      "epoch": 0.11373795569683123,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0007909049369226184,
+      "loss": 0.9054,
+      "step": 2290
+    },
+    {
+      "epoch": 0.11423462799245059,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0007908652031389689,
+      "loss": 0.91,
+      "step": 2300
+    },
+    {
+      "epoch": 0.11473130028806994,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0007908254693553195,
+      "loss": 0.9317,
+      "step": 2310
+    },
+    {
+      "epoch": 0.11522797258368928,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0007907857355716698,
+      "loss": 0.9031,
+      "step": 2320
+    },
+    {
+      "epoch": 0.11572464487930863,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0007907460017880203,
+      "loss": 0.9089,
+      "step": 2330
+    },
+    {
+      "epoch": 0.11622131717492798,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0007907062680043708,
+      "loss": 0.9035,
+      "step": 2340
+    },
+    {
+      "epoch": 0.11671798947054733,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0007906665342207211,
+      "loss": 0.8807,
+      "step": 2350
+    },
+    {
+      "epoch": 0.11721466176616668,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0007906268004370717,
+      "loss": 0.8563,
+      "step": 2360
+    },
+    {
+      "epoch": 0.11771133406178604,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0007905870666534221,
+      "loss": 0.8702,
+      "step": 2370
+    },
+    {
+      "epoch": 0.11820800635740539,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0007905473328697726,
+      "loss": 0.9268,
+      "step": 2380
+    },
+    {
+      "epoch": 0.11870467865302474,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.000790507599086123,
+      "loss": 0.902,
+      "step": 2390
+    },
+    {
+      "epoch": 0.11920135094864409,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0007904678653024734,
+      "loss": 0.9036,
+      "step": 2400
+    },
+    {
+      "epoch": 0.11969802324426343,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.000790428131518824,
+      "loss": 0.9041,
+      "step": 2410
+    },
+    {
+      "epoch": 0.12019469553988278,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0007903883977351744,
+      "loss": 0.8601,
+      "step": 2420
+    },
+    {
+      "epoch": 0.12069136783550213,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0007903486639515248,
+      "loss": 0.8791,
+      "step": 2430
+    },
+    {
+      "epoch": 0.12118804013112149,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0007903089301678753,
+      "loss": 0.9051,
+      "step": 2440
+    },
+    {
+      "epoch": 0.12168471242674084,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0007902691963842257,
+      "loss": 0.8775,
+      "step": 2450
+    },
+    {
+      "epoch": 0.12218138472236019,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0007902294626005762,
+      "loss": 0.8495,
+      "step": 2460
+    },
+    {
+      "epoch": 0.12267805701797954,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0007901897288169267,
+      "loss": 0.9162,
+      "step": 2470
+    },
+    {
+      "epoch": 0.1231747293135989,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.000790149995033277,
+      "loss": 0.8772,
+      "step": 2480
+    },
+    {
+      "epoch": 0.12367140160921823,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0007901102612496275,
+      "loss": 0.8825,
+      "step": 2490
+    },
+    {
+      "epoch": 0.12416807390483758,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.000790070527465978,
+      "loss": 0.9092,
+      "step": 2500
+    },
+    {
+      "epoch": 0.12466474620045694,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0007900307936823284,
+      "loss": 0.8917,
+      "step": 2510
+    },
+    {
+      "epoch": 0.1251614184960763,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0007899910598986789,
+      "loss": 0.8767,
+      "step": 2520
+    },
+    {
+      "epoch": 0.12565809079169563,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0007899513261150294,
+      "loss": 0.8803,
+      "step": 2530
+    },
+    {
+      "epoch": 0.126154763087315,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0007899115923313798,
+      "loss": 0.895,
+      "step": 2540
+    },
+    {
+      "epoch": 0.12665143538293433,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0007898718585477302,
+      "loss": 0.9083,
+      "step": 2550
+    },
+    {
+      "epoch": 0.1271481076785537,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0007898321247640807,
+      "loss": 0.8967,
+      "step": 2560
+    },
+    {
+      "epoch": 0.12764477997417303,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0007897923909804312,
+      "loss": 0.8881,
+      "step": 2570
+    },
+    {
+      "epoch": 0.1281414522697924,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0007897526571967816,
+      "loss": 0.9021,
+      "step": 2580
+    },
+    {
+      "epoch": 0.12863812456541174,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.000789712923413132,
+      "loss": 0.8832,
+      "step": 2590
+    },
+    {
+      "epoch": 0.1291347968610311,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0007896731896294825,
+      "loss": 0.9154,
+      "step": 2600
+    },
+    {
+      "epoch": 0.12963146915665044,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.000789633455845833,
+      "loss": 0.8943,
+      "step": 2610
+    },
+    {
+      "epoch": 0.13012814145226978,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0007895937220621834,
+      "loss": 0.8884,
+      "step": 2620
+    },
+    {
+      "epoch": 0.13062481374788915,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0007895539882785339,
+      "loss": 0.8645,
+      "step": 2630
+    },
+    {
+      "epoch": 0.13112148604350848,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0007895142544948843,
+      "loss": 0.8769,
+      "step": 2640
+    },
+    {
+      "epoch": 0.13161815833912785,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0007894745207112347,
+      "loss": 0.9108,
+      "step": 2650
+    },
+    {
+      "epoch": 0.1321148306347472,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0007894347869275853,
+      "loss": 0.8442,
+      "step": 2660
+    },
+    {
+      "epoch": 0.13261150293036655,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0007893950531439357,
+      "loss": 0.8799,
+      "step": 2670
+    },
+    {
+      "epoch": 0.1331081752259859,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0007893553193602861,
+      "loss": 0.9038,
+      "step": 2680
+    },
+    {
+      "epoch": 0.13360484752160526,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0007893155855766366,
+      "loss": 0.8603,
+      "step": 2690
+    },
+    {
+      "epoch": 0.1341015198172246,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.000789275851792987,
+      "loss": 0.8633,
+      "step": 2700
+    },
+    {
+      "epoch": 0.13459819211284393,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0007892361180093375,
+      "loss": 0.8866,
+      "step": 2710
+    },
+    {
+      "epoch": 0.1350948644084633,
+      "grad_norm": 0.25,
+      "learning_rate": 0.000789196384225688,
+      "loss": 0.873,
+      "step": 2720
+    },
+    {
+      "epoch": 0.13559153670408264,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0007891566504420384,
+      "loss": 0.85,
+      "step": 2730
+    },
+    {
+      "epoch": 0.136088208999702,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0007891169166583888,
+      "loss": 0.8394,
+      "step": 2740
+    },
+    {
+      "epoch": 0.13658488129532134,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0007890771828747392,
+      "loss": 0.8567,
+      "step": 2750
+    },
+    {
+      "epoch": 0.1370815535909407,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0007890374490910898,
+      "loss": 0.9067,
+      "step": 2760
+    },
+    {
+      "epoch": 0.13757822588656005,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0007889977153074402,
+      "loss": 0.8816,
+      "step": 2770
+    },
+    {
+      "epoch": 0.1380748981821794,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0007889579815237906,
+      "loss": 0.8963,
+      "step": 2780
+    },
+    {
+      "epoch": 0.13857157047779875,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0007889182477401411,
+      "loss": 0.838,
+      "step": 2790
+    },
+    {
+      "epoch": 0.1390682427734181,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0007888785139564915,
+      "loss": 0.8586,
+      "step": 2800
+    },
+    {
+      "epoch": 0.13956491506903745,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.000788838780172842,
+      "loss": 0.8659,
+      "step": 2810
+    },
+    {
+      "epoch": 0.1400615873646568,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0007887990463891925,
+      "loss": 0.9058,
+      "step": 2820
+    },
+    {
+      "epoch": 0.14055825966027616,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0007887593126055429,
+      "loss": 0.8714,
+      "step": 2830
+    },
+    {
+      "epoch": 0.1410549319558955,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0007887195788218933,
+      "loss": 0.8725,
+      "step": 2840
+    },
+    {
+      "epoch": 0.14155160425151486,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0007886798450382438,
+      "loss": 0.8797,
+      "step": 2850
+    },
+    {
+      "epoch": 0.1420482765471342,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0007886401112545943,
+      "loss": 0.866,
+      "step": 2860
+    },
+    {
+      "epoch": 0.14254494884275354,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0007886003774709447,
+      "loss": 0.8559,
+      "step": 2870
+    },
+    {
+      "epoch": 0.1430416211383729,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0007885606436872952,
+      "loss": 0.8683,
+      "step": 2880
+    },
+    {
+      "epoch": 0.14353829343399224,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0007885209099036456,
+      "loss": 0.8658,
+      "step": 2890
+    },
+    {
+      "epoch": 0.1440349657296116,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.000788481176119996,
+      "loss": 0.8778,
+      "step": 2900
+    },
+    {
+      "epoch": 0.14453163802523095,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0007884414423363466,
+      "loss": 0.8682,
+      "step": 2910
+    },
+    {
+      "epoch": 0.1450283103208503,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.000788401708552697,
+      "loss": 0.8524,
+      "step": 2920
+    },
+    {
+      "epoch": 0.14552498261646965,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0007883619747690474,
+      "loss": 0.8535,
+      "step": 2930
+    },
+    {
+      "epoch": 0.14602165491208902,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0007883222409853979,
+      "loss": 0.8414,
+      "step": 2940
+    },
+    {
+      "epoch": 0.14651832720770835,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0007882825072017483,
+      "loss": 0.8294,
+      "step": 2950
+    },
+    {
+      "epoch": 0.1470149995033277,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0007882427734180987,
+      "loss": 0.8701,
+      "step": 2960
+    },
+    {
+      "epoch": 0.14751167179894706,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0007882030396344492,
+      "loss": 0.8428,
+      "step": 2970
+    },
+    {
+      "epoch": 0.1480083440945664,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0007881633058507997,
+      "loss": 0.8341,
+      "step": 2980
+    },
+    {
+      "epoch": 0.14850501639018576,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0007881235720671502,
+      "loss": 0.839,
+      "step": 2990
+    },
+    {
+      "epoch": 0.1490016886858051,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0007880838382835005,
+      "loss": 0.8262,
+      "step": 3000
+    },
+    {
+      "epoch": 0.14949836098142447,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.000788044104499851,
+      "loss": 0.8626,
+      "step": 3010
+    },
+    {
+      "epoch": 0.1499950332770438,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0007880043707162015,
+      "loss": 0.8461,
+      "step": 3020
+    },
+    {
+      "epoch": 0.15049170557266317,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0007879646369325519,
+      "loss": 0.8441,
+      "step": 3030
+    },
+    {
+      "epoch": 0.1509883778682825,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0007879249031489024,
+      "loss": 0.8595,
+      "step": 3040
+    },
+    {
+      "epoch": 0.15148505016390185,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0007878851693652528,
+      "loss": 0.877,
+      "step": 3050
+    },
+    {
+      "epoch": 0.1519817224595212,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0007878454355816032,
+      "loss": 0.8823,
+      "step": 3060
+    },
+    {
+      "epoch": 0.15247839475514055,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0007878057017979538,
+      "loss": 0.8481,
+      "step": 3070
+    },
+    {
+      "epoch": 0.15297506705075992,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0007877659680143042,
+      "loss": 0.8838,
+      "step": 3080
+    },
+    {
+      "epoch": 0.15347173934637925,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0007877262342306546,
+      "loss": 0.8341,
+      "step": 3090
+    },
+    {
+      "epoch": 0.15396841164199862,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0007876865004470051,
+      "loss": 0.8893,
+      "step": 3100
+    },
+    {
+      "epoch": 0.15446508393761796,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0007876467666633555,
+      "loss": 0.8442,
+      "step": 3110
+    },
+    {
+      "epoch": 0.1549617562332373,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.000787607032879706,
+      "loss": 0.837,
+      "step": 3120
+    },
+    {
+      "epoch": 0.15545842852885666,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0007875672990960565,
+      "loss": 0.8601,
+      "step": 3130
+    },
+    {
+      "epoch": 0.155955100824476,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0007875275653124069,
+      "loss": 0.8514,
+      "step": 3140
+    },
+    {
+      "epoch": 0.15645177312009537,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0007874878315287574,
+      "loss": 0.886,
+      "step": 3150
+    },
+    {
+      "epoch": 0.1569484454157147,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0007874480977451077,
+      "loss": 0.8572,
+      "step": 3160
+    },
+    {
+      "epoch": 0.15744511771133407,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0007874083639614583,
+      "loss": 0.862,
+      "step": 3170
+    },
+    {
+      "epoch": 0.1579417900069534,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0007873686301778088,
+      "loss": 0.8112,
+      "step": 3180
+    },
+    {
+      "epoch": 0.15843846230257277,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0007873288963941591,
+      "loss": 0.8726,
+      "step": 3190
+    },
+    {
+      "epoch": 0.1589351345981921,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0007872891626105096,
+      "loss": 0.8618,
+      "step": 3200
+    },
+    {
+      "epoch": 0.15943180689381145,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0007872494288268602,
+      "loss": 0.8528,
+      "step": 3210
+    },
+    {
+      "epoch": 0.15992847918943082,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0007872096950432105,
+      "loss": 0.8395,
+      "step": 3220
+    },
+    {
+      "epoch": 0.16042515148505015,
+      "grad_norm": 0.25,
+      "learning_rate": 0.000787169961259561,
+      "loss": 0.827,
+      "step": 3230
+    },
+    {
+      "epoch": 0.16092182378066952,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0007871302274759114,
+      "loss": 0.8478,
+      "step": 3240
+    },
+    {
+      "epoch": 0.16141849607628886,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0007870904936922618,
+      "loss": 0.8247,
+      "step": 3250
+    },
+    {
+      "epoch": 0.16191516837190822,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0007870507599086123,
+      "loss": 0.8557,
+      "step": 3260
+    },
+    {
+      "epoch": 0.16241184066752756,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0007870110261249628,
+      "loss": 0.8532,
+      "step": 3270
+    },
+    {
+      "epoch": 0.16290851296314693,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0007869712923413133,
+      "loss": 0.8545,
+      "step": 3280
+    },
+    {
+      "epoch": 0.16340518525876627,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0007869315585576637,
+      "loss": 0.8675,
+      "step": 3290
+    },
+    {
+      "epoch": 0.1639018575543856,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0007868918247740141,
+      "loss": 0.8209,
+      "step": 3300
+    },
+    {
+      "epoch": 0.16439852985000497,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0007868520909903647,
+      "loss": 0.8332,
+      "step": 3310
+    },
+    {
+      "epoch": 0.1648952021456243,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0007868123572067151,
+      "loss": 0.8373,
+      "step": 3320
+    },
+    {
+      "epoch": 0.16539187444124367,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0007867726234230655,
+      "loss": 0.862,
+      "step": 3330
+    },
+    {
+      "epoch": 0.165888546736863,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.000786732889639416,
+      "loss": 0.8397,
+      "step": 3340
+    },
+    {
+      "epoch": 0.16638521903248238,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0007866931558557663,
+      "loss": 0.8448,
+      "step": 3350
+    },
+    {
+      "epoch": 0.16688189132810172,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0007866534220721168,
+      "loss": 0.8382,
+      "step": 3360
+    },
+    {
+      "epoch": 0.16737856362372106,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0007866136882884674,
+      "loss": 0.837,
+      "step": 3370
+    },
+    {
+      "epoch": 0.16787523591934042,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0007865739545048177,
+      "loss": 0.8321,
+      "step": 3380
+    },
+    {
+      "epoch": 0.16837190821495976,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0007865342207211682,
+      "loss": 0.8347,
+      "step": 3390
+    },
+    {
+      "epoch": 0.16886858051057912,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0007864944869375187,
+      "loss": 0.879,
+      "step": 3400
+    },
+    {
+      "epoch": 0.16936525280619846,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.000786454753153869,
+      "loss": 0.8377,
+      "step": 3410
+    },
+    {
+      "epoch": 0.16986192510181783,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0007864150193702196,
+      "loss": 0.7883,
+      "step": 3420
+    },
+    {
+      "epoch": 0.17035859739743717,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00078637528558657,
+      "loss": 0.8456,
+      "step": 3430
+    },
+    {
+      "epoch": 0.17085526969305653,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0007863355518029205,
+      "loss": 0.8309,
+      "step": 3440
+    },
+    {
+      "epoch": 0.17135194198867587,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0007862958180192709,
+      "loss": 0.8715,
+      "step": 3450
+    },
+    {
+      "epoch": 0.1718486142842952,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0007862560842356213,
+      "loss": 0.8108,
+      "step": 3460
+    },
+    {
+      "epoch": 0.17234528657991458,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0007862163504519719,
+      "loss": 0.8348,
+      "step": 3470
+    },
+    {
+      "epoch": 0.1728419588755339,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0007861766166683223,
+      "loss": 0.8579,
+      "step": 3480
+    },
+    {
+      "epoch": 0.17333863117115328,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0007861368828846727,
+      "loss": 0.8149,
+      "step": 3490
+    },
+    {
+      "epoch": 0.17383530346677262,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0007860971491010232,
+      "loss": 0.8031,
+      "step": 3500
+    },
+    {
+      "epoch": 0.17433197576239198,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0007860574153173736,
+      "loss": 0.8666,
+      "step": 3510
+    },
+    {
+      "epoch": 0.17482864805801132,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0007860176815337241,
+      "loss": 0.8271,
+      "step": 3520
+    },
+    {
+      "epoch": 0.1753253203536307,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0007859779477500746,
+      "loss": 0.8059,
+      "step": 3530
+    },
+    {
+      "epoch": 0.17582199264925003,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.000785938213966425,
+      "loss": 0.8332,
+      "step": 3540
+    },
+    {
+      "epoch": 0.17631866494486936,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0007858984801827754,
+      "loss": 0.827,
+      "step": 3550
+    },
+    {
+      "epoch": 0.17681533724048873,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.000785858746399126,
+      "loss": 0.8217,
+      "step": 3560
+    },
+    {
+      "epoch": 0.17731200953610807,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0007858190126154763,
+      "loss": 0.8426,
+      "step": 3570
+    },
+    {
+      "epoch": 0.17780868183172743,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0007857792788318268,
+      "loss": 0.8383,
+      "step": 3580
+    },
+    {
+      "epoch": 0.17830535412734677,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0007857395450481773,
+      "loss": 0.8361,
+      "step": 3590
+    },
+    {
+      "epoch": 0.17880202642296614,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0007856998112645277,
+      "loss": 0.815,
+      "step": 3600
+    },
+    {
+      "epoch": 0.17929869871858548,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0007856600774808781,
+      "loss": 0.822,
+      "step": 3610
+    },
+    {
+      "epoch": 0.17979537101420484,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0007856203436972286,
+      "loss": 0.8272,
+      "step": 3620
+    },
+    {
+      "epoch": 0.18029204330982418,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0007855806099135791,
+      "loss": 0.8239,
+      "step": 3630
+    },
+    {
+      "epoch": 0.18078871560544352,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0007855408761299295,
+      "loss": 0.8254,
+      "step": 3640
+    },
+    {
+      "epoch": 0.18128538790106288,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0007855011423462799,
+      "loss": 0.8422,
+      "step": 3650
+    },
+    {
+      "epoch": 0.18178206019668222,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0007854614085626304,
+      "loss": 0.8412,
+      "step": 3660
+    },
+    {
+      "epoch": 0.1822787324923016,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0007854216747789809,
+      "loss": 0.8247,
+      "step": 3670
+    },
+    {
+      "epoch": 0.18277540478792093,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0007853819409953313,
+      "loss": 0.8094,
+      "step": 3680
+    },
+    {
+      "epoch": 0.1832720770835403,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0007853422072116818,
+      "loss": 0.8558,
+      "step": 3690
+    },
+    {
+      "epoch": 0.18376874937915963,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0007853024734280322,
+      "loss": 0.8463,
+      "step": 3700
+    },
+    {
+      "epoch": 0.18426542167477897,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0007852627396443826,
+      "loss": 0.834,
+      "step": 3710
+    },
+    {
+      "epoch": 0.18476209397039833,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0007852230058607332,
+      "loss": 0.8033,
+      "step": 3720
+    },
+    {
+      "epoch": 0.18525876626601767,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0007851832720770836,
+      "loss": 0.8285,
+      "step": 3730
+    },
+    {
+      "epoch": 0.18575543856163704,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.000785143538293434,
+      "loss": 0.7968,
+      "step": 3740
+    },
+    {
+      "epoch": 0.18625211085725638,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0007851038045097845,
+      "loss": 0.823,
+      "step": 3750
+    },
+    {
+      "epoch": 0.18674878315287574,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0007850640707261349,
+      "loss": 0.8429,
+      "step": 3760
+    },
+    {
+      "epoch": 0.18724545544849508,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0007850243369424854,
+      "loss": 0.8029,
+      "step": 3770
+    },
+    {
+      "epoch": 0.18774212774411445,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0007849846031588359,
+      "loss": 0.7974,
+      "step": 3780
+    },
+    {
+      "epoch": 0.18823880003973378,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0007849448693751863,
+      "loss": 0.8167,
+      "step": 3790
+    },
+    {
+      "epoch": 0.18873547233535312,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0007849051355915367,
+      "loss": 0.8322,
+      "step": 3800
+    },
+    {
+      "epoch": 0.1892321446309725,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0007848654018078872,
+      "loss": 0.8332,
+      "step": 3810
+    },
+    {
+      "epoch": 0.18972881692659183,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0007848256680242377,
+      "loss": 0.8058,
+      "step": 3820
+    },
+    {
+      "epoch": 0.1902254892222112,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0007847859342405881,
+      "loss": 0.8551,
+      "step": 3830
+    },
+    {
+      "epoch": 0.19072216151783053,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0007847462004569385,
+      "loss": 0.8265,
+      "step": 3840
+    },
+    {
+      "epoch": 0.1912188338134499,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.000784706466673289,
+      "loss": 0.8347,
+      "step": 3850
+    },
+    {
+      "epoch": 0.19171550610906923,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0007846667328896394,
+      "loss": 0.803,
+      "step": 3860
+    },
+    {
+      "epoch": 0.1922121784046886,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0007846269991059899,
+      "loss": 0.8361,
+      "step": 3870
+    },
+    {
+      "epoch": 0.19270885070030794,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0007845872653223404,
+      "loss": 0.8125,
+      "step": 3880
+    },
+    {
+      "epoch": 0.19320552299592728,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0007845475315386908,
+      "loss": 0.8293,
+      "step": 3890
+    },
+    {
+      "epoch": 0.19370219529154664,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0007845077977550412,
+      "loss": 0.8149,
+      "step": 3900
+    },
+    {
+      "epoch": 0.19419886758716598,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0007844680639713917,
+      "loss": 0.8187,
+      "step": 3910
+    },
+    {
+      "epoch": 0.19469553988278535,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0007844283301877422,
+      "loss": 0.8111,
+      "step": 3920
+    },
+    {
+      "epoch": 0.19519221217840468,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0007843885964040926,
+      "loss": 0.8085,
+      "step": 3930
+    },
+    {
+      "epoch": 0.19568888447402405,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0007843488626204431,
+      "loss": 0.8489,
+      "step": 3940
+    },
+    {
+      "epoch": 0.1961855567696434,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0007843091288367935,
+      "loss": 0.8231,
+      "step": 3950
+    },
+    {
+      "epoch": 0.19668222906526273,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0007842693950531439,
+      "loss": 0.8353,
+      "step": 3960
+    },
+    {
+      "epoch": 0.1971789013608821,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0007842296612694945,
+      "loss": 0.8116,
+      "step": 3970
+    },
+    {
+      "epoch": 0.19767557365650143,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0007841899274858449,
+      "loss": 0.8392,
+      "step": 3980
+    },
+    {
+      "epoch": 0.1981722459521208,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0007841501937021953,
+      "loss": 0.8145,
+      "step": 3990
+    },
+    {
+      "epoch": 0.19866891824774013,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0007841104599185458,
+      "loss": 0.8646,
+      "step": 4000
+    },
+    {
+      "epoch": 0.1991655905433595,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0007840707261348962,
+      "loss": 0.792,
+      "step": 4010
+    },
+    {
+      "epoch": 0.19966226283897884,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0007840309923512468,
+      "loss": 0.7947,
+      "step": 4020
+    },
+    {
+      "epoch": 0.2001589351345982,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0007839912585675971,
+      "loss": 0.7848,
+      "step": 4030
+    },
+    {
+      "epoch": 0.20065560743021754,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0007839515247839476,
+      "loss": 0.8112,
+      "step": 4040
+    },
+    {
+      "epoch": 0.20115227972583688,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0007839117910002981,
+      "loss": 0.8311,
+      "step": 4050
+    },
+    {
+      "epoch": 0.20164895202145625,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0007838720572166484,
+      "loss": 0.7896,
+      "step": 4060
+    },
+    {
+      "epoch": 0.20214562431707558,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.000783832323432999,
+      "loss": 0.7951,
+      "step": 4070
+    },
+    {
+      "epoch": 0.20264229661269495,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0007837925896493495,
+      "loss": 0.757,
+      "step": 4080
+    },
+    {
+      "epoch": 0.2031389689083143,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0007837528558656998,
+      "loss": 0.7988,
+      "step": 4090
+    },
+    {
+      "epoch": 0.20363564120393365,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0007837131220820503,
+      "loss": 0.8171,
+      "step": 4100
+    },
+    {
+      "epoch": 0.204132313499553,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.0007836733882984007,
+      "loss": 0.806,
+      "step": 4110
+    },
+    {
+      "epoch": 0.20462898579517236,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0007836336545147512,
+      "loss": 0.8052,
+      "step": 4120
+    },
+    {
+      "epoch": 0.2051256580907917,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0007835939207311017,
+      "loss": 0.814,
+      "step": 4130
+    },
+    {
+      "epoch": 0.20562233038641103,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0007835541869474521,
+      "loss": 0.8323,
+      "step": 4140
+    },
+    {
+      "epoch": 0.2061190026820304,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0007835144531638025,
+      "loss": 0.7967,
+      "step": 4150
+    },
+    {
+      "epoch": 0.20661567497764974,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.000783474719380153,
+      "loss": 0.7966,
+      "step": 4160
+    },
+    {
+      "epoch": 0.2071123472732691,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0007834349855965035,
+      "loss": 0.8208,
+      "step": 4170
+    },
+    {
+      "epoch": 0.20760901956888844,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.000783395251812854,
+      "loss": 0.8258,
+      "step": 4180
+    },
+    {
+      "epoch": 0.2081056918645078,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0007833555180292044,
+      "loss": 0.8061,
+      "step": 4190
+    },
+    {
+      "epoch": 0.20860236416012715,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0007833157842455548,
+      "loss": 0.824,
+      "step": 4200
+    },
+    {
+      "epoch": 0.2090990364557465,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0007832760504619053,
+      "loss": 0.8106,
+      "step": 4210
+    },
+    {
+      "epoch": 0.20959570875136585,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0007832363166782556,
+      "loss": 0.8123,
+      "step": 4220
+    },
+    {
+      "epoch": 0.2100923810469852,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0007831965828946062,
+      "loss": 0.863,
+      "step": 4230
+    },
+    {
+      "epoch": 0.21058905334260455,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0007831568491109567,
+      "loss": 0.8205,
+      "step": 4240
+    },
+    {
+      "epoch": 0.2110857256382239,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.000783117115327307,
+      "loss": 0.8325,
+      "step": 4250
+    },
+    {
+      "epoch": 0.21158239793384326,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0007830773815436575,
+      "loss": 0.8257,
+      "step": 4260
+    },
+    {
+      "epoch": 0.2120790702294626,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0007830376477600081,
+      "loss": 0.8101,
+      "step": 4270
+    },
+    {
+      "epoch": 0.21257574252508196,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0007829979139763584,
+      "loss": 0.8094,
+      "step": 4280
+    },
+    {
+      "epoch": 0.2130724148207013,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0007829581801927089,
+      "loss": 0.8194,
+      "step": 4290
+    },
+    {
+      "epoch": 0.21356908711632064,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0007829184464090593,
+      "loss": 0.8438,
+      "step": 4300
+    },
+    {
+      "epoch": 0.21406575941194,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0007828787126254097,
+      "loss": 0.8385,
+      "step": 4310
+    },
+    {
+      "epoch": 0.21456243170755934,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0007828389788417603,
+      "loss": 0.8205,
+      "step": 4320
+    },
+    {
+      "epoch": 0.2150591040031787,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0007827992450581107,
+      "loss": 0.7999,
+      "step": 4330
+    },
+    {
+      "epoch": 0.21555577629879805,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0007827595112744612,
+      "loss": 0.7847,
+      "step": 4340
+    },
+    {
+      "epoch": 0.2160524485944174,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0007827197774908116,
+      "loss": 0.7887,
+      "step": 4350
+    },
+    {
+      "epoch": 0.21654912089003675,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.000782680043707162,
+      "loss": 0.8135,
+      "step": 4360
+    },
+    {
+      "epoch": 0.21704579318565612,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0007826403099235126,
+      "loss": 0.8124,
+      "step": 4370
+    },
+    {
+      "epoch": 0.21754246548127545,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.000782600576139863,
+      "loss": 0.7823,
+      "step": 4380
+    },
+    {
+      "epoch": 0.2180391377768948,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0007825608423562134,
+      "loss": 0.8115,
+      "step": 4390
+    },
+    {
+      "epoch": 0.21853581007251416,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0007825211085725639,
+      "loss": 0.8017,
+      "step": 4400
+    },
+    {
+      "epoch": 0.2190324823681335,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0007824813747889143,
+      "loss": 0.7821,
+      "step": 4410
+    },
+    {
+      "epoch": 0.21952915466375286,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0007824416410052647,
+      "loss": 0.8001,
+      "step": 4420
+    },
+    {
+      "epoch": 0.2200258269593722,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0007824019072216153,
+      "loss": 0.7993,
+      "step": 4430
+    },
+    {
+      "epoch": 0.22052249925499157,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0007823621734379656,
+      "loss": 0.8269,
+      "step": 4440
+    },
+    {
+      "epoch": 0.2210191715506109,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0007823224396543161,
+      "loss": 0.8173,
+      "step": 4450
+    },
+    {
+      "epoch": 0.22151584384623027,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0007822827058706666,
+      "loss": 0.7898,
+      "step": 4460
+    },
+    {
+      "epoch": 0.2220125161418496,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.000782242972087017,
+      "loss": 0.8034,
+      "step": 4470
+    },
+    {
+      "epoch": 0.22250918843746895,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0007822032383033675,
+      "loss": 0.7929,
+      "step": 4480
+    },
+    {
+      "epoch": 0.2230058607330883,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0007821635045197179,
+      "loss": 0.7875,
+      "step": 4490
+    },
+    {
+      "epoch": 0.22350253302870765,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0007821237707360684,
+      "loss": 0.7699,
+      "step": 4500
+    },
+    {
+      "epoch": 0.22399920532432702,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0007820840369524188,
+      "loss": 0.815,
+      "step": 4510
+    },
+    {
+      "epoch": 0.22449587761994635,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0007820443031687692,
+      "loss": 0.7915,
+      "step": 4520
+    },
+    {
+      "epoch": 0.22499254991556572,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0007820045693851198,
+      "loss": 0.8077,
+      "step": 4530
+    },
+    {
+      "epoch": 0.22548922221118506,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0007819648356014702,
+      "loss": 0.7716,
+      "step": 4540
+    },
+    {
+      "epoch": 0.2259858945068044,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0007819251018178206,
+      "loss": 0.7838,
+      "step": 4550
+    },
+    {
+      "epoch": 0.22648256680242376,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0007818853680341711,
+      "loss": 0.8343,
+      "step": 4560
+    },
+    {
+      "epoch": 0.2269792390980431,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0007818456342505215,
+      "loss": 0.7671,
+      "step": 4570
+    },
+    {
+      "epoch": 0.22747591139366247,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.000781805900466872,
+      "loss": 0.7987,
+      "step": 4580
+    },
+    {
+      "epoch": 0.2279725836892818,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0007817661666832225,
+      "loss": 0.7885,
+      "step": 4590
+    },
+    {
+      "epoch": 0.22846925598490117,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0007817264328995729,
+      "loss": 0.7943,
+      "step": 4600
+    },
+    {
+      "epoch": 0.2289659282805205,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0007816866991159233,
+      "loss": 0.8023,
+      "step": 4610
+    },
+    {
+      "epoch": 0.22946260057613987,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0007816469653322739,
+      "loss": 0.7773,
+      "step": 4620
+    },
+    {
+      "epoch": 0.2299592728717592,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0007816072315486243,
+      "loss": 0.8065,
+      "step": 4630
+    },
+    {
+      "epoch": 0.23045594516737855,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0007815674977649747,
+      "loss": 0.7635,
+      "step": 4640
+    },
+    {
+      "epoch": 0.23095261746299792,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0007815277639813252,
+      "loss": 0.7777,
+      "step": 4650
+    },
+    {
+      "epoch": 0.23144928975861725,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0007814880301976756,
+      "loss": 0.7775,
+      "step": 4660
+    },
+    {
+      "epoch": 0.23194596205423662,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.000781448296414026,
+      "loss": 0.8366,
+      "step": 4670
+    },
+    {
+      "epoch": 0.23244263434985596,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0007814085626303766,
+      "loss": 0.7759,
+      "step": 4680
+    },
+    {
+      "epoch": 0.23293930664547532,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.000781368828846727,
+      "loss": 0.7545,
+      "step": 4690
+    },
+    {
+      "epoch": 0.23343597894109466,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0007813290950630774,
+      "loss": 0.8055,
+      "step": 4700
+    },
+    {
+      "epoch": 0.23393265123671403,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0007812893612794278,
+      "loss": 0.7886,
+      "step": 4710
+    },
+    {
+      "epoch": 0.23442932353233337,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0007812496274957783,
+      "loss": 0.8074,
+      "step": 4720
+    },
+    {
+      "epoch": 0.2349259958279527,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0007812098937121288,
+      "loss": 0.8098,
+      "step": 4730
+    },
+    {
+      "epoch": 0.23542266812357207,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0007811701599284792,
+      "loss": 0.8287,
+      "step": 4740
+    },
+    {
+      "epoch": 0.2359193404191914,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0007811304261448297,
+      "loss": 0.7923,
+      "step": 4750
+    },
+    {
+      "epoch": 0.23641601271481077,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0007810906923611801,
+      "loss": 0.8065,
+      "step": 4760
+    },
+    {
+      "epoch": 0.2369126850104301,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0007810509585775305,
+      "loss": 0.7706,
+      "step": 4770
+    },
+    {
+      "epoch": 0.23740935730604948,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0007810112247938811,
+      "loss": 0.8026,
+      "step": 4780
+    },
+    {
+      "epoch": 0.23790602960166882,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0007809714910102315,
+      "loss": 0.8084,
+      "step": 4790
+    },
+    {
+      "epoch": 0.23840270189728818,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0007809317572265819,
+      "loss": 0.7945,
+      "step": 4800
+    },
+    {
+      "epoch": 0.23889937419290752,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0007808920234429324,
+      "loss": 0.8042,
+      "step": 4810
+    },
+    {
+      "epoch": 0.23939604648852686,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0007808522896592828,
+      "loss": 0.7626,
+      "step": 4820
+    },
+    {
+      "epoch": 0.23989271878414622,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0007808125558756333,
+      "loss": 0.7616,
+      "step": 4830
+    },
+    {
+      "epoch": 0.24038939107976556,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007807728220919838,
+      "loss": 0.7974,
+      "step": 4840
+    },
+    {
+      "epoch": 0.24088606337538493,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0007807330883083342,
+      "loss": 0.8034,
+      "step": 4850
+    },
+    {
+      "epoch": 0.24138273567100427,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0007806933545246846,
+      "loss": 0.8325,
+      "step": 4860
+    },
+    {
+      "epoch": 0.24187940796662363,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007806536207410351,
+      "loss": 0.8055,
+      "step": 4870
+    },
+    {
+      "epoch": 0.24237608026224297,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0007806138869573856,
+      "loss": 0.7872,
+      "step": 4880
+    },
+    {
+      "epoch": 0.2428727525578623,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.000780574153173736,
+      "loss": 0.8107,
+      "step": 4890
+    },
+    {
+      "epoch": 0.24336942485348168,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0007805344193900864,
+      "loss": 0.7817,
+      "step": 4900
+    },
+    {
+      "epoch": 0.243866097149101,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0007804946856064369,
+      "loss": 0.8126,
+      "step": 4910
+    },
+    {
+      "epoch": 0.24436276944472038,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0007804549518227874,
+      "loss": 0.7746,
+      "step": 4920
+    },
+    {
+      "epoch": 0.24485944174033972,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0007804152180391378,
+      "loss": 0.7759,
+      "step": 4930
+    },
+    {
+      "epoch": 0.24535611403595908,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0007803754842554883,
+      "loss": 0.77,
+      "step": 4940
+    },
+    {
+      "epoch": 0.24585278633157842,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0007803357504718388,
+      "loss": 0.7883,
+      "step": 4950
+    },
+    {
+      "epoch": 0.2463494586271978,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0007802960166881891,
+      "loss": 0.7896,
+      "step": 4960
+    },
+    {
+      "epoch": 0.24684613092281713,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0007802562829045396,
+      "loss": 0.796,
+      "step": 4970
+    },
+    {
+      "epoch": 0.24734280321843646,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007802165491208901,
+      "loss": 0.8069,
+      "step": 4980
+    },
+    {
+      "epoch": 0.24783947551405583,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0007801768153372405,
+      "loss": 0.8263,
+      "step": 4990
+    },
+    {
+      "epoch": 0.24833614780967517,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 0.000780137081553591,
+      "loss": 0.7691,
+      "step": 5000
+    },
+    {
+      "epoch": 0.24883282010529453,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0007800973477699414,
+      "loss": 0.7552,
+      "step": 5010
+    },
+    {
+      "epoch": 0.24932949240091387,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0007800576139862918,
+      "loss": 0.7764,
+      "step": 5020
+    },
+    {
+      "epoch": 0.24982616469653324,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0007800178802026424,
+      "loss": 0.7935,
+      "step": 5030
+    },
+    {
+      "epoch": 0.2503228369921526,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0007799781464189928,
+      "loss": 0.7723,
+      "step": 5040
+    },
+    {
+      "epoch": 0.2508195092877719,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0007799384126353432,
+      "loss": 0.7674,
+      "step": 5050
+    },
+    {
+      "epoch": 0.25131618158339125,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007798986788516937,
+      "loss": 0.7987,
+      "step": 5060
+    },
+    {
+      "epoch": 0.25181285387901065,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0007798589450680441,
+      "loss": 0.8071,
+      "step": 5070
+    },
+    {
+      "epoch": 0.25230952617463,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0007798192112843947,
+      "loss": 0.8022,
+      "step": 5080
+    },
+    {
+      "epoch": 0.2528061984702493,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.000779779477500745,
+      "loss": 0.7887,
+      "step": 5090
+    },
+    {
+      "epoch": 0.25330287076586866,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0007797397437170955,
+      "loss": 0.7665,
+      "step": 5100
+    },
+    {
+      "epoch": 0.25379954306148805,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.000779700009933446,
+      "loss": 0.8016,
+      "step": 5110
+    },
+    {
+      "epoch": 0.2542962153571074,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0007796602761497963,
+      "loss": 0.7871,
+      "step": 5120
+    },
+    {
+      "epoch": 0.25479288765272673,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0007796205423661469,
+      "loss": 0.8054,
+      "step": 5130
+    },
+    {
+      "epoch": 0.25528955994834607,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0007795808085824974,
+      "loss": 0.7859,
+      "step": 5140
+    },
+    {
+      "epoch": 0.2557862322439654,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0007795410747988477,
+      "loss": 0.7812,
+      "step": 5150
+    },
+    {
+      "epoch": 0.2562829045395848,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0007795013410151982,
+      "loss": 0.7794,
+      "step": 5160
+    },
+    {
+      "epoch": 0.25677957683520414,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0007794616072315486,
+      "loss": 0.7664,
+      "step": 5170
+    },
+    {
+      "epoch": 0.2572762491308235,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.000779421873447899,
+      "loss": 0.7864,
+      "step": 5180
+    },
+    {
+      "epoch": 0.2577729214264428,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0007793821396642496,
+      "loss": 0.8203,
+      "step": 5190
+    },
+    {
+      "epoch": 0.2582695937220622,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0007793424058806,
+      "loss": 0.7712,
+      "step": 5200
+    },
+    {
+      "epoch": 0.25876626601768155,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0007793026720969504,
+      "loss": 0.7664,
+      "step": 5210
+    },
+    {
+      "epoch": 0.2592629383133009,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0007792629383133009,
+      "loss": 0.7916,
+      "step": 5220
+    },
+    {
+      "epoch": 0.2597596106089202,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0007792232045296514,
+      "loss": 0.7596,
+      "step": 5230
+    },
+    {
+      "epoch": 0.26025628290453956,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0007791834707460019,
+      "loss": 0.7672,
+      "step": 5240
+    },
+    {
+      "epoch": 0.26075295520015895,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0007791437369623523,
+      "loss": 0.7835,
+      "step": 5250
+    },
+    {
+      "epoch": 0.2612496274957783,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0007791040031787027,
+      "loss": 0.7874,
+      "step": 5260
+    },
+    {
+      "epoch": 0.26174629979139763,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0007790642693950532,
+      "loss": 0.7597,
+      "step": 5270
+    },
+    {
+      "epoch": 0.26224297208701697,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0007790245356114037,
+      "loss": 0.7825,
+      "step": 5280
+    },
+    {
+      "epoch": 0.26273964438263636,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0007789848018277541,
+      "loss": 0.8065,
+      "step": 5290
+    },
+    {
+      "epoch": 0.2632363166782557,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0007789450680441046,
+      "loss": 0.7338,
+      "step": 5300
+    },
+    {
+      "epoch": 0.26373298897387504,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0007789053342604549,
+      "loss": 0.769,
+      "step": 5310
+    },
+    {
+      "epoch": 0.2642296612694944,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0007788656004768054,
+      "loss": 0.776,
+      "step": 5320
+    },
+    {
+      "epoch": 0.2647263335651137,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.000778825866693156,
+      "loss": 0.8244,
+      "step": 5330
+    },
+    {
+      "epoch": 0.2652230058607331,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007787861329095063,
+      "loss": 0.7883,
+      "step": 5340
+    },
+    {
+      "epoch": 0.26571967815635245,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0007787463991258568,
+      "loss": 0.7951,
+      "step": 5350
+    },
+    {
+      "epoch": 0.2662163504519718,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0007787066653422072,
+      "loss": 0.8077,
+      "step": 5360
+    },
+    {
+      "epoch": 0.2667130227475911,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0007786669315585577,
+      "loss": 0.7914,
+      "step": 5370
+    },
+    {
+      "epoch": 0.2672096950432105,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0007786271977749082,
+      "loss": 0.7563,
+      "step": 5380
+    },
+    {
+      "epoch": 0.26770636733882985,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0007785874639912586,
+      "loss": 0.7683,
+      "step": 5390
+    },
+    {
+      "epoch": 0.2682030396344492,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0007785477302076091,
+      "loss": 0.7949,
+      "step": 5400
+    },
+    {
+      "epoch": 0.26869971193006853,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0007785079964239595,
+      "loss": 0.8195,
+      "step": 5410
+    },
+    {
+      "epoch": 0.26919638422568787,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007784682626403099,
+      "loss": 0.7995,
+      "step": 5420
+    },
+    {
+      "epoch": 0.26969305652130726,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0007784285288566605,
+      "loss": 0.7776,
+      "step": 5430
+    },
+    {
+      "epoch": 0.2701897288169266,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0007783887950730109,
+      "loss": 0.7899,
+      "step": 5440
+    },
+    {
+      "epoch": 0.27068640111254594,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0007783490612893613,
+      "loss": 0.8031,
+      "step": 5450
+    },
+    {
+      "epoch": 0.2711830734081653,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007783093275057118,
+      "loss": 0.749,
+      "step": 5460
+    },
+    {
+      "epoch": 0.27167974570378467,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0007782695937220622,
+      "loss": 0.7685,
+      "step": 5470
+    },
+    {
+      "epoch": 0.272176417999404,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0007782298599384127,
+      "loss": 0.765,
+      "step": 5480
+    },
+    {
+      "epoch": 0.27267309029502335,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0007781901261547632,
+      "loss": 0.7843,
+      "step": 5490
+    },
+    {
+      "epoch": 0.2731697625906427,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0007781503923711135,
+      "loss": 0.7801,
+      "step": 5500
+    },
+    {
+      "epoch": 0.273666434886262,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.000778110658587464,
+      "loss": 0.7614,
+      "step": 5510
+    },
+    {
+      "epoch": 0.2741631071818814,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0007780709248038145,
+      "loss": 0.8162,
+      "step": 5520
+    },
+    {
+      "epoch": 0.27465977947750075,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.000778031191020165,
+      "loss": 0.7625,
+      "step": 5530
+    },
+    {
+      "epoch": 0.2751564517731201,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0007779914572365154,
+      "loss": 0.7837,
+      "step": 5540
+    },
+    {
+      "epoch": 0.27565312406873943,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0007779517234528659,
+      "loss": 0.7613,
+      "step": 5550
+    },
+    {
+      "epoch": 0.2761497963643588,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0007779119896692163,
+      "loss": 0.7512,
+      "step": 5560
+    },
+    {
+      "epoch": 0.27664646865997816,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0007778722558855667,
+      "loss": 0.7763,
+      "step": 5570
+    },
+    {
+      "epoch": 0.2771431409555975,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0007778325221019171,
+      "loss": 0.7752,
+      "step": 5580
+    },
+    {
+      "epoch": 0.27763981325121684,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0007777927883182677,
+      "loss": 0.8011,
+      "step": 5590
+    },
+    {
+      "epoch": 0.2781364855468362,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0007777530545346181,
+      "loss": 0.7834,
+      "step": 5600
+    },
+    {
+      "epoch": 0.27863315784245557,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0007777133207509685,
+      "loss": 0.8148,
+      "step": 5610
+    },
+    {
+      "epoch": 0.2791298301380749,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.000777673586967319,
+      "loss": 0.7647,
+      "step": 5620
+    },
+    {
+      "epoch": 0.27962650243369425,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0007776338531836695,
+      "loss": 0.7446,
+      "step": 5630
+    },
+    {
+      "epoch": 0.2801231747293136,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0007775941194000199,
+      "loss": 0.7964,
+      "step": 5640
+    },
+    {
+      "epoch": 0.2806198470249329,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0007775543856163704,
+      "loss": 0.7563,
+      "step": 5650
+    },
+    {
+      "epoch": 0.2811165193205523,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0007775146518327208,
+      "loss": 0.7848,
+      "step": 5660
+    },
+    {
+      "epoch": 0.28161319161617165,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0007774749180490712,
+      "loss": 0.7805,
+      "step": 5670
+    },
+    {
+      "epoch": 0.282109863911791,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0007774351842654218,
+      "loss": 0.7383,
+      "step": 5680
+    },
+    {
+      "epoch": 0.28260653620741033,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0007773954504817722,
+      "loss": 0.7549,
+      "step": 5690
+    },
+    {
+      "epoch": 0.2831032085030297,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0007773557166981226,
+      "loss": 0.7672,
+      "step": 5700
+    },
+    {
+      "epoch": 0.28359988079864906,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0007773159829144731,
+      "loss": 0.7659,
+      "step": 5710
+    },
+    {
+      "epoch": 0.2840965530942684,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0007772762491308235,
+      "loss": 0.7696,
+      "step": 5720
+    },
+    {
+      "epoch": 0.28459322538988774,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.000777236515347174,
+      "loss": 0.7976,
+      "step": 5730
+    },
+    {
+      "epoch": 0.2850898976855071,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0007771967815635245,
+      "loss": 0.7625,
+      "step": 5740
+    },
+    {
+      "epoch": 0.28558656998112647,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0007771570477798749,
+      "loss": 0.8234,
+      "step": 5750
+    },
+    {
+      "epoch": 0.2860832422767458,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.0007771173139962253,
+      "loss": 0.7991,
+      "step": 5760
+    },
+    {
+      "epoch": 0.28657991457236515,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0007770775802125757,
+      "loss": 0.7791,
+      "step": 5770
+    },
+    {
+      "epoch": 0.2870765868679845,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0007770378464289263,
+      "loss": 0.8082,
+      "step": 5780
+    },
+    {
+      "epoch": 0.2875732591636039,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0007769981126452767,
+      "loss": 0.7729,
+      "step": 5790
+    },
+    {
+      "epoch": 0.2880699314592232,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0007769583788616271,
+      "loss": 0.7724,
+      "step": 5800
+    },
+    {
+      "epoch": 0.28856660375484255,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0007769186450779776,
+      "loss": 0.7698,
+      "step": 5810
+    },
+    {
+      "epoch": 0.2890632760504619,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.000776878911294328,
+      "loss": 0.7587,
+      "step": 5820
+    },
+    {
+      "epoch": 0.28955994834608123,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0007768391775106784,
+      "loss": 0.7634,
+      "step": 5830
+    },
+    {
+      "epoch": 0.2900566206417006,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.000776799443727029,
+      "loss": 0.7557,
+      "step": 5840
+    },
+    {
+      "epoch": 0.29055329293731996,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0007767597099433794,
+      "loss": 0.7405,
+      "step": 5850
+    },
+    {
+      "epoch": 0.2910499652329393,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0007767199761597298,
+      "loss": 0.7826,
+      "step": 5860
+    },
+    {
+      "epoch": 0.29154663752855864,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0007766802423760803,
+      "loss": 0.7608,
+      "step": 5870
+    },
+    {
+      "epoch": 0.29204330982417803,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0007766405085924307,
+      "loss": 0.739,
+      "step": 5880
+    },
+    {
+      "epoch": 0.29253998211979737,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0007766007748087812,
+      "loss": 0.7819,
+      "step": 5890
+    },
+    {
+      "epoch": 0.2930366544154167,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0007765610410251317,
+      "loss": 0.7584,
+      "step": 5900
+    },
+    {
+      "epoch": 0.29353332671103605,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0007765213072414821,
+      "loss": 0.7396,
+      "step": 5910
+    },
+    {
+      "epoch": 0.2940299990066554,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0007764815734578325,
+      "loss": 0.7617,
+      "step": 5920
+    },
+    {
+      "epoch": 0.2945266713022748,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.000776441839674183,
+      "loss": 0.7464,
+      "step": 5930
+    },
+    {
+      "epoch": 0.2950233435978941,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007764021058905335,
+      "loss": 0.757,
+      "step": 5940
+    },
+    {
+      "epoch": 0.29552001589351345,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0007763623721068839,
+      "loss": 0.7461,
+      "step": 5950
+    },
+    {
+      "epoch": 0.2960166881891328,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0007763226383232343,
+      "loss": 0.77,
+      "step": 5960
+    },
+    {
+      "epoch": 0.2965133604847522,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0007762829045395848,
+      "loss": 0.7571,
+      "step": 5970
+    },
+    {
+      "epoch": 0.2970100327803715,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0007762431707559354,
+      "loss": 0.7764,
+      "step": 5980
+    },
+    {
+      "epoch": 0.29750670507599086,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0007762034369722857,
+      "loss": 0.7419,
+      "step": 5990
+    },
+    {
+      "epoch": 0.2980033773716102,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007761637031886362,
+      "loss": 0.7873,
+      "step": 6000
+    },
+    {
+      "epoch": 0.29850004966722954,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0007761239694049867,
+      "loss": 0.7525,
+      "step": 6010
+    },
+    {
+      "epoch": 0.29899672196284893,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.000776084235621337,
+      "loss": 0.7782,
+      "step": 6020
+    },
+    {
+      "epoch": 0.29949339425846827,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0007760445018376875,
+      "loss": 0.7807,
+      "step": 6030
+    },
+    {
+      "epoch": 0.2999900665540876,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.000776004768054038,
+      "loss": 0.7586,
+      "step": 6040
+    },
+    {
+      "epoch": 0.30048673884970695,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0007759650342703884,
+      "loss": 0.7722,
+      "step": 6050
+    },
+    {
+      "epoch": 0.30098341114532634,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0007759253004867389,
+      "loss": 0.7659,
+      "step": 6060
+    },
+    {
+      "epoch": 0.3014800834409457,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0007758855667030893,
+      "loss": 0.7636,
+      "step": 6070
+    },
+    {
+      "epoch": 0.301976755736565,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0007758458329194397,
+      "loss": 0.7604,
+      "step": 6080
+    },
+    {
+      "epoch": 0.30247342803218435,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0007758060991357903,
+      "loss": 0.7528,
+      "step": 6090
+    },
+    {
+      "epoch": 0.3029701003278037,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0007757663653521407,
+      "loss": 0.7645,
+      "step": 6100
+    },
+    {
+      "epoch": 0.3034667726234231,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0007757266315684912,
+      "loss": 0.7779,
+      "step": 6110
+    },
+    {
+      "epoch": 0.3039634449190424,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0007756868977848416,
+      "loss": 0.758,
+      "step": 6120
+    },
+    {
+      "epoch": 0.30446011721466176,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.000775647164001192,
+      "loss": 0.7451,
+      "step": 6130
+    },
+    {
+      "epoch": 0.3049567895102811,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0007756074302175426,
+      "loss": 0.7465,
+      "step": 6140
+    },
+    {
+      "epoch": 0.30545346180590044,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0007755676964338929,
+      "loss": 0.7755,
+      "step": 6150
+    },
+    {
+      "epoch": 0.30595013410151983,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0007755279626502434,
+      "loss": 0.7573,
+      "step": 6160
+    },
+    {
+      "epoch": 0.30644680639713917,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0007754882288665939,
+      "loss": 0.7536,
+      "step": 6170
+    },
+    {
+      "epoch": 0.3069434786927585,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0007754484950829442,
+      "loss": 0.7855,
+      "step": 6180
+    },
+    {
+      "epoch": 0.30744015098837785,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0007754087612992948,
+      "loss": 0.756,
+      "step": 6190
+    },
+    {
+      "epoch": 0.30793682328399724,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0007753690275156453,
+      "loss": 0.7682,
+      "step": 6200
+    },
+    {
+      "epoch": 0.3084334955796166,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007753292937319956,
+      "loss": 0.7463,
+      "step": 6210
+    },
+    {
+      "epoch": 0.3089301678752359,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0007752895599483461,
+      "loss": 0.7839,
+      "step": 6220
+    },
+    {
+      "epoch": 0.30942684017085526,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0007752498261646965,
+      "loss": 0.7485,
+      "step": 6230
+    },
+    {
+      "epoch": 0.3099235124664746,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.000775210092381047,
+      "loss": 0.7452,
+      "step": 6240
+    },
+    {
+      "epoch": 0.310420184762094,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0007751703585973975,
+      "loss": 0.7546,
+      "step": 6250
+    },
+    {
+      "epoch": 0.3109168570577133,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0007751306248137479,
+      "loss": 0.7653,
+      "step": 6260
+    },
+    {
+      "epoch": 0.31141352935333266,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0007750908910300984,
+      "loss": 0.7725,
+      "step": 6270
+    },
+    {
+      "epoch": 0.311910201648952,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0007750511572464488,
+      "loss": 0.7692,
+      "step": 6280
+    },
+    {
+      "epoch": 0.3124068739445714,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0007750114234627993,
+      "loss": 0.7469,
+      "step": 6290
+    },
+    {
+      "epoch": 0.31290354624019073,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0007749716896791498,
+      "loss": 0.7579,
+      "step": 6300
+    },
+    {
+      "epoch": 0.31340021853581007,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0007749319558955002,
+      "loss": 0.7455,
+      "step": 6310
+    },
+    {
+      "epoch": 0.3138968908314294,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0007748922221118506,
+      "loss": 0.7563,
+      "step": 6320
+    },
+    {
+      "epoch": 0.31439356312704875,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0007748524883282011,
+      "loss": 0.7535,
+      "step": 6330
+    },
+    {
+      "epoch": 0.31489023542266814,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007748127545445516,
+      "loss": 0.7547,
+      "step": 6340
+    },
+    {
+      "epoch": 0.3153869077182875,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.000774773020760902,
+      "loss": 0.7392,
+      "step": 6350
+    },
+    {
+      "epoch": 0.3158835800139068,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0007747332869772525,
+      "loss": 0.7651,
+      "step": 6360
+    },
+    {
+      "epoch": 0.31638025230952616,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0007746935531936028,
+      "loss": 0.7575,
+      "step": 6370
+    },
+    {
+      "epoch": 0.31687692460514555,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0007746538194099533,
+      "loss": 0.7446,
+      "step": 6380
+    },
+    {
+      "epoch": 0.3173735969007649,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0007746140856263039,
+      "loss": 0.7425,
+      "step": 6390
+    },
+    {
+      "epoch": 0.3178702691963842,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0007745743518426542,
+      "loss": 0.7696,
+      "step": 6400
+    },
+    {
+      "epoch": 0.31836694149200356,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0007745346180590047,
+      "loss": 0.7632,
+      "step": 6410
+    },
+    {
+      "epoch": 0.3188636137876229,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0007744948842753551,
+      "loss": 0.7838,
+      "step": 6420
+    },
+    {
+      "epoch": 0.3193602860832423,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0007744551504917056,
+      "loss": 0.7685,
+      "step": 6430
+    },
+    {
+      "epoch": 0.31985695837886163,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0007744154167080561,
+      "loss": 0.7652,
+      "step": 6440
+    },
+    {
+      "epoch": 0.32035363067448097,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0007743756829244065,
+      "loss": 0.755,
+      "step": 6450
+    },
+    {
+      "epoch": 0.3208503029701003,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.000774335949140757,
+      "loss": 0.7277,
+      "step": 6460
+    },
+    {
+      "epoch": 0.3213469752657197,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0007742962153571074,
+      "loss": 0.7414,
+      "step": 6470
+    },
+    {
+      "epoch": 0.32184364756133904,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0007742564815734578,
+      "loss": 0.7466,
+      "step": 6480
+    },
+    {
+      "epoch": 0.3223403198569584,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007742167477898084,
+      "loss": 0.7707,
+      "step": 6490
+    },
+    {
+      "epoch": 0.3228369921525777,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0007741770140061588,
+      "loss": 0.7345,
+      "step": 6500
+    },
+    {
+      "epoch": 0.32333366444819706,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0007741372802225092,
+      "loss": 0.7477,
+      "step": 6510
+    },
+    {
+      "epoch": 0.32383033674381645,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0007740975464388597,
+      "loss": 0.7523,
+      "step": 6520
+    },
+    {
+      "epoch": 0.3243270090394358,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0007740578126552101,
+      "loss": 0.7233,
+      "step": 6530
+    },
+    {
+      "epoch": 0.3248236813350551,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0007740180788715606,
+      "loss": 0.7454,
+      "step": 6540
+    },
+    {
+      "epoch": 0.32532035363067446,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0007739783450879111,
+      "loss": 0.8268,
+      "step": 6550
+    },
+    {
+      "epoch": 0.32581702592629386,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0007739386113042615,
+      "loss": 0.7481,
+      "step": 6560
+    },
+    {
+      "epoch": 0.3263136982219132,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0007738988775206119,
+      "loss": 0.742,
+      "step": 6570
+    },
+    {
+      "epoch": 0.32681037051753253,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0007738591437369624,
+      "loss": 0.7556,
+      "step": 6580
+    },
+    {
+      "epoch": 0.32730704281315187,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0007738194099533129,
+      "loss": 0.7327,
+      "step": 6590
+    },
+    {
+      "epoch": 0.3278037151087712,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0007737796761696633,
+      "loss": 0.7419,
+      "step": 6600
+    },
+    {
+      "epoch": 0.3283003874043906,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0007737399423860138,
+      "loss": 0.7433,
+      "step": 6610
+    },
+    {
+      "epoch": 0.32879705970000994,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0007737002086023642,
+      "loss": 0.7658,
+      "step": 6620
+    },
+    {
+      "epoch": 0.3292937319956293,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0007736604748187146,
+      "loss": 0.7514,
+      "step": 6630
+    },
+    {
+      "epoch": 0.3297904042912486,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.000773620741035065,
+      "loss": 0.7481,
+      "step": 6640
+    },
+    {
+      "epoch": 0.330287076586868,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0007735810072514156,
+      "loss": 0.741,
+      "step": 6650
+    },
+    {
+      "epoch": 0.33078374888248735,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.000773541273467766,
+      "loss": 0.7469,
+      "step": 6660
+    },
+    {
+      "epoch": 0.3312804211781067,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0007735015396841164,
+      "loss": 0.7719,
+      "step": 6670
+    },
+    {
+      "epoch": 0.331777093473726,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0007734618059004669,
+      "loss": 0.7762,
+      "step": 6680
+    },
+    {
+      "epoch": 0.33227376576934536,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0007734220721168174,
+      "loss": 0.7379,
+      "step": 6690
+    },
+    {
+      "epoch": 0.33277043806496476,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0007733823383331678,
+      "loss": 0.7259,
+      "step": 6700
+    },
+    {
+      "epoch": 0.3332671103605841,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0007733426045495183,
+      "loss": 0.7645,
+      "step": 6710
+    },
+    {
+      "epoch": 0.33376378265620343,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0007733028707658687,
+      "loss": 0.7616,
+      "step": 6720
+    },
+    {
+      "epoch": 0.33426045495182277,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0007732631369822191,
+      "loss": 0.7393,
+      "step": 6730
+    },
+    {
+      "epoch": 0.3347571272474421,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0007732234031985697,
+      "loss": 0.7931,
+      "step": 6740
+    },
+    {
+      "epoch": 0.3352537995430615,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0007731836694149201,
+      "loss": 0.7229,
+      "step": 6750
+    },
+    {
+      "epoch": 0.33575047183868084,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0007731439356312705,
+      "loss": 0.7394,
+      "step": 6760
+    },
+    {
+      "epoch": 0.3362471441343002,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.000773104201847621,
+      "loss": 0.7355,
+      "step": 6770
+    },
+    {
+      "epoch": 0.3367438164299195,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0007730644680639714,
+      "loss": 0.7552,
+      "step": 6780
+    },
+    {
+      "epoch": 0.3372404887255389,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0007730247342803219,
+      "loss": 0.7536,
+      "step": 6790
+    },
+    {
+      "epoch": 0.33773716102115825,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0007729850004966724,
+      "loss": 0.7641,
+      "step": 6800
+    },
+    {
+      "epoch": 0.3382338333167776,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0007729452667130228,
+      "loss": 0.7434,
+      "step": 6810
+    },
+    {
+      "epoch": 0.3387305056123969,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007729055329293732,
+      "loss": 0.7578,
+      "step": 6820
+    },
+    {
+      "epoch": 0.33922717790801626,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0007728657991457236,
+      "loss": 0.7551,
+      "step": 6830
+    },
+    {
+      "epoch": 0.33972385020363566,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0007728260653620742,
+      "loss": 0.7743,
+      "step": 6840
+    },
+    {
+      "epoch": 0.340220522499255,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0007727863315784246,
+      "loss": 0.7424,
+      "step": 6850
+    },
+    {
+      "epoch": 0.34071719479487433,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.000772746597794775,
+      "loss": 0.7685,
+      "step": 6860
+    },
+    {
+      "epoch": 0.34121386709049367,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0007727068640111255,
+      "loss": 0.7318,
+      "step": 6870
+    },
+    {
+      "epoch": 0.34171053938611307,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.000772667130227476,
+      "loss": 0.7452,
+      "step": 6880
+    },
+    {
+      "epoch": 0.3422072116817324,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0007726273964438263,
+      "loss": 0.7754,
+      "step": 6890
+    },
+    {
+      "epoch": 0.34270388397735174,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0007725876626601769,
+      "loss": 0.763,
+      "step": 6900
+    },
+    {
+      "epoch": 0.3432005562729711,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007725479288765273,
+      "loss": 0.7346,
+      "step": 6910
+    },
+    {
+      "epoch": 0.3436972285685904,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0007725081950928777,
+      "loss": 0.7633,
+      "step": 6920
+    },
+    {
+      "epoch": 0.3441939008642098,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0007724684613092282,
+      "loss": 0.7258,
+      "step": 6930
+    },
+    {
+      "epoch": 0.34469057315982915,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0007724287275255787,
+      "loss": 0.7323,
+      "step": 6940
+    },
+    {
+      "epoch": 0.3451872454554485,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0007723889937419291,
+      "loss": 0.7417,
+      "step": 6950
+    },
+    {
+      "epoch": 0.3456839177510678,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0007723492599582796,
+      "loss": 0.7511,
+      "step": 6960
+    },
+    {
+      "epoch": 0.3461805900466872,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00077230952617463,
+      "loss": 0.703,
+      "step": 6970
+    },
+    {
+      "epoch": 0.34667726234230656,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0007722697923909804,
+      "loss": 0.7316,
+      "step": 6980
+    },
+    {
+      "epoch": 0.3471739346379259,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.000772230058607331,
+      "loss": 0.7663,
+      "step": 6990
+    },
+    {
+      "epoch": 0.34767060693354523,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0007721903248236814,
+      "loss": 0.7679,
+      "step": 7000
+    },
+    {
+      "epoch": 0.3481672792291646,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0007721505910400319,
+      "loss": 0.7432,
+      "step": 7010
+    },
+    {
+      "epoch": 0.34866395152478397,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0007721108572563822,
+      "loss": 0.7137,
+      "step": 7020
+    },
+    {
+      "epoch": 0.3491606238204033,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0007720711234727327,
+      "loss": 0.7484,
+      "step": 7030
+    },
+    {
+      "epoch": 0.34965729611602264,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0007720313896890833,
+      "loss": 0.8066,
+      "step": 7040
+    },
+    {
+      "epoch": 0.350153968411642,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0007719916559054336,
+      "loss": 0.7591,
+      "step": 7050
+    },
+    {
+      "epoch": 0.3506506407072614,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0007719519221217841,
+      "loss": 0.7204,
+      "step": 7060
+    },
+    {
+      "epoch": 0.3511473130028807,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0007719121883381346,
+      "loss": 0.7619,
+      "step": 7070
+    },
+    {
+      "epoch": 0.35164398529850005,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0007718724545544849,
+      "loss": 0.7494,
+      "step": 7080
+    },
+    {
+      "epoch": 0.3521406575941194,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0007718327207708355,
+      "loss": 0.7119,
+      "step": 7090
+    },
+    {
+      "epoch": 0.3526373298897387,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0007717929869871859,
+      "loss": 0.74,
+      "step": 7100
+    },
+    {
+      "epoch": 0.3531340021853581,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0007717532532035363,
+      "loss": 0.7364,
+      "step": 7110
+    },
+    {
+      "epoch": 0.35363067448097746,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0007717135194198868,
+      "loss": 0.723,
+      "step": 7120
+    },
+    {
+      "epoch": 0.3541273467765968,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0007716737856362372,
+      "loss": 0.7895,
+      "step": 7130
+    },
+    {
+      "epoch": 0.35462401907221613,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0007716340518525876,
+      "loss": 0.7586,
+      "step": 7140
+    },
+    {
+      "epoch": 0.35512069136783553,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0007715943180689382,
+      "loss": 0.7063,
+      "step": 7150
+    },
+    {
+      "epoch": 0.35561736366345487,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0007715545842852886,
+      "loss": 0.7826,
+      "step": 7160
+    },
+    {
+      "epoch": 0.3561140359590742,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0007715148505016391,
+      "loss": 0.7246,
+      "step": 7170
+    },
+    {
+      "epoch": 0.35661070825469354,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0007714751167179895,
+      "loss": 0.7159,
+      "step": 7180
+    },
+    {
+      "epoch": 0.3571073805503129,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00077143538293434,
+      "loss": 0.7204,
+      "step": 7190
+    },
+    {
+      "epoch": 0.3576040528459323,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0007713956491506905,
+      "loss": 0.7538,
+      "step": 7200
+    },
+    {
+      "epoch": 0.3581007251415516,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0007713559153670409,
+      "loss": 0.7469,
+      "step": 7210
+    },
+    {
+      "epoch": 0.35859739743717095,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0007713161815833913,
+      "loss": 0.7408,
+      "step": 7220
+    },
+    {
+      "epoch": 0.3590940697327903,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0007712764477997418,
+      "loss": 0.7432,
+      "step": 7230
+    },
+    {
+      "epoch": 0.3595907420284097,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0007712367140160921,
+      "loss": 0.7366,
+      "step": 7240
+    },
+    {
+      "epoch": 0.360087414324029,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0007711969802324427,
+      "loss": 0.728,
+      "step": 7250
+    },
+    {
+      "epoch": 0.36058408661964836,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0007711572464487932,
+      "loss": 0.7584,
+      "step": 7260
+    },
+    {
+      "epoch": 0.3610807589152677,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0007711175126651435,
+      "loss": 0.6997,
+      "step": 7270
+    },
+    {
+      "epoch": 0.36157743121088703,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.000771077778881494,
+      "loss": 0.7144,
+      "step": 7280
+    },
+    {
+      "epoch": 0.36207410350650643,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0007710380450978444,
+      "loss": 0.728,
+      "step": 7290
+    },
+    {
+      "epoch": 0.36257077580212577,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0007709983113141949,
+      "loss": 0.7011,
+      "step": 7300
+    },
+    {
+      "epoch": 0.3630674480977451,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007709585775305454,
+      "loss": 0.7655,
+      "step": 7310
+    },
+    {
+      "epoch": 0.36356412039336444,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0007709188437468958,
+      "loss": 0.7247,
+      "step": 7320
+    },
+    {
+      "epoch": 0.3640607926889838,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0007708791099632463,
+      "loss": 0.7441,
+      "step": 7330
+    },
+    {
+      "epoch": 0.3645574649846032,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0007708393761795967,
+      "loss": 0.7515,
+      "step": 7340
+    },
+    {
+      "epoch": 0.3650541372802225,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0007707996423959472,
+      "loss": 0.7171,
+      "step": 7350
+    },
+    {
+      "epoch": 0.36555080957584185,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0007707599086122977,
+      "loss": 0.6862,
+      "step": 7360
+    },
+    {
+      "epoch": 0.3660474818714612,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0007707201748286481,
+      "loss": 0.706,
+      "step": 7370
+    },
+    {
+      "epoch": 0.3665441541670806,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0007706804410449985,
+      "loss": 0.7725,
+      "step": 7380
+    },
+    {
+      "epoch": 0.3670408264626999,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.000770640707261349,
+      "loss": 0.7314,
+      "step": 7390
+    },
+    {
+      "epoch": 0.36753749875831926,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0007706009734776995,
+      "loss": 0.7398,
+      "step": 7400
+    },
+    {
+      "epoch": 0.3680341710539386,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0007705612396940499,
+      "loss": 0.7266,
+      "step": 7410
+    },
+    {
+      "epoch": 0.36853084334955793,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0007705215059104004,
+      "loss": 0.7625,
+      "step": 7420
+    },
+    {
+      "epoch": 0.36902751564517733,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0007704817721267507,
+      "loss": 0.716,
+      "step": 7430
+    },
+    {
+      "epoch": 0.36952418794079667,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0007704420383431012,
+      "loss": 0.7418,
+      "step": 7440
+    },
+    {
+      "epoch": 0.370020860236416,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0007704023045594518,
+      "loss": 0.7125,
+      "step": 7450
+    },
+    {
+      "epoch": 0.37051753253203534,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0007703625707758022,
+      "loss": 0.7312,
+      "step": 7460
+    },
+    {
+      "epoch": 0.37101420482765474,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0007703228369921526,
+      "loss": 0.7424,
+      "step": 7470
+    },
+    {
+      "epoch": 0.3715108771232741,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0007702831032085031,
+      "loss": 0.7156,
+      "step": 7480
+    },
+    {
+      "epoch": 0.3720075494188934,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0007702433694248535,
+      "loss": 0.7348,
+      "step": 7490
+    },
+    {
+      "epoch": 0.37250422171451275,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.000770203635641204,
+      "loss": 0.6993,
+      "step": 7500
+    },
+    {
+      "epoch": 0.3730008940101321,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0007701639018575544,
+      "loss": 0.7259,
+      "step": 7510
+    },
+    {
+      "epoch": 0.3734975663057515,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0007701241680739049,
+      "loss": 0.681,
+      "step": 7520
+    },
+    {
+      "epoch": 0.3739942386013708,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0007700844342902553,
+      "loss": 0.734,
+      "step": 7530
+    },
+    {
+      "epoch": 0.37449091089699016,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0007700447005066057,
+      "loss": 0.7473,
+      "step": 7540
+    },
+    {
+      "epoch": 0.3749875831926095,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0007700049667229563,
+      "loss": 0.7197,
+      "step": 7550
+    },
+    {
+      "epoch": 0.3754842554882289,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0007699652329393067,
+      "loss": 0.7397,
+      "step": 7560
+    },
+    {
+      "epoch": 0.37598092778384823,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0007699254991556571,
+      "loss": 0.749,
+      "step": 7570
+    },
+    {
+      "epoch": 0.37647760007946757,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0007698857653720076,
+      "loss": 0.7399,
+      "step": 7580
+    },
+    {
+      "epoch": 0.3769742723750869,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.000769846031588358,
+      "loss": 0.7477,
+      "step": 7590
+    },
+    {
+      "epoch": 0.37747094467070624,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0007698062978047085,
+      "loss": 0.7321,
+      "step": 7600
+    },
+    {
+      "epoch": 0.37796761696632564,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.000769766564021059,
+      "loss": 0.7328,
+      "step": 7610
+    },
+    {
+      "epoch": 0.378464289261945,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0007697268302374094,
+      "loss": 0.7521,
+      "step": 7620
+    },
+    {
+      "epoch": 0.3789609615575643,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0007696870964537598,
+      "loss": 0.762,
+      "step": 7630
+    },
+    {
+      "epoch": 0.37945763385318365,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0007696473626701103,
+      "loss": 0.7221,
+      "step": 7640
+    },
+    {
+      "epoch": 0.37995430614880304,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0007696076288864608,
+      "loss": 0.7237,
+      "step": 7650
+    },
+    {
+      "epoch": 0.3804509784444224,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0007695678951028112,
+      "loss": 0.7303,
+      "step": 7660
+    },
+    {
+      "epoch": 0.3809476507400417,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0007695281613191617,
+      "loss": 0.7715,
+      "step": 7670
+    },
+    {
+      "epoch": 0.38144432303566106,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0007694884275355121,
+      "loss": 0.7481,
+      "step": 7680
+    },
+    {
+      "epoch": 0.3819409953312804,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0007694486937518625,
+      "loss": 0.7587,
+      "step": 7690
+    },
+    {
+      "epoch": 0.3824376676268998,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.000769408959968213,
+      "loss": 0.7155,
+      "step": 7700
+    },
+    {
+      "epoch": 0.38293433992251913,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0007693692261845635,
+      "loss": 0.7326,
+      "step": 7710
+    },
+    {
+      "epoch": 0.38343101221813847,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0007693294924009139,
+      "loss": 0.7087,
+      "step": 7720
+    },
+    {
+      "epoch": 0.3839276845137578,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0007692897586172643,
+      "loss": 0.7131,
+      "step": 7730
+    },
+    {
+      "epoch": 0.3844243568093772,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0007692500248336148,
+      "loss": 0.7382,
+      "step": 7740
+    },
+    {
+      "epoch": 0.38492102910499654,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0007692102910499653,
+      "loss": 0.7186,
+      "step": 7750
+    },
+    {
+      "epoch": 0.3854177014006159,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0007691705572663157,
+      "loss": 0.7351,
+      "step": 7760
+    },
+    {
+      "epoch": 0.3859143736962352,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0007691308234826662,
+      "loss": 0.7372,
+      "step": 7770
+    },
+    {
+      "epoch": 0.38641104599185455,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0007690910896990166,
+      "loss": 0.751,
+      "step": 7780
+    },
+    {
+      "epoch": 0.38690771828747395,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.000769051355915367,
+      "loss": 0.7708,
+      "step": 7790
+    },
+    {
+      "epoch": 0.3874043905830933,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0007690116221317176,
+      "loss": 0.7545,
+      "step": 7800
+    },
+    {
+      "epoch": 0.3879010628787126,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.000768971888348068,
+      "loss": 0.7645,
+      "step": 7810
+    },
+    {
+      "epoch": 0.38839773517433196,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0007689321545644184,
+      "loss": 0.7348,
+      "step": 7820
+    },
+    {
+      "epoch": 0.38889440746995135,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0007688924207807689,
+      "loss": 0.7003,
+      "step": 7830
+    },
+    {
+      "epoch": 0.3893910797655707,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0007688526869971193,
+      "loss": 0.7449,
+      "step": 7840
+    },
+    {
+      "epoch": 0.38988775206119003,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0007688129532134698,
+      "loss": 0.7048,
+      "step": 7850
+    },
+    {
+      "epoch": 0.39038442435680937,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0007687732194298203,
+      "loss": 0.6932,
+      "step": 7860
+    },
+    {
+      "epoch": 0.3908810966524287,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0007687334856461707,
+      "loss": 0.6957,
+      "step": 7870
+    },
+    {
+      "epoch": 0.3913777689480481,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0007686937518625211,
+      "loss": 0.7196,
+      "step": 7880
+    },
+    {
+      "epoch": 0.39187444124366744,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0007686540180788715,
+      "loss": 0.7493,
+      "step": 7890
+    },
+    {
+      "epoch": 0.3923711135392868,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0007686142842952221,
+      "loss": 0.7667,
+      "step": 7900
+    },
+    {
+      "epoch": 0.3928677858349061,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0007685745505115726,
+      "loss": 0.7573,
+      "step": 7910
+    },
+    {
+      "epoch": 0.39336445813052545,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0007685348167279229,
+      "loss": 0.7438,
+      "step": 7920
+    },
+    {
+      "epoch": 0.39386113042614485,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0007684950829442734,
+      "loss": 0.7369,
+      "step": 7930
+    },
+    {
+      "epoch": 0.3943578027217642,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0007684553491606239,
+      "loss": 0.7336,
+      "step": 7940
+    },
+    {
+      "epoch": 0.3948544750173835,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0007684156153769743,
+      "loss": 0.7205,
+      "step": 7950
+    },
+    {
+      "epoch": 0.39535114731300286,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0007683758815933248,
+      "loss": 0.699,
+      "step": 7960
+    },
+    {
+      "epoch": 0.39584781960862225,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0007683361478096752,
+      "loss": 0.7284,
+      "step": 7970
+    },
+    {
+      "epoch": 0.3963444919042416,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0007682964140260256,
+      "loss": 0.727,
+      "step": 7980
+    },
+    {
+      "epoch": 0.39684116419986093,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0007682566802423761,
+      "loss": 0.7521,
+      "step": 7990
+    },
+    {
+      "epoch": 0.39733783649548027,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0007682169464587266,
+      "loss": 0.7427,
+      "step": 8000
+    },
+    {
+      "epoch": 0.3978345087910996,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.000768177212675077,
+      "loss": 0.6822,
+      "step": 8010
+    },
+    {
+      "epoch": 0.398331181086719,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0007681374788914275,
+      "loss": 0.718,
+      "step": 8020
+    },
+    {
+      "epoch": 0.39882785338233834,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0007680977451077779,
+      "loss": 0.7153,
+      "step": 8030
+    },
+    {
+      "epoch": 0.3993245256779577,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0007680580113241283,
+      "loss": 0.7443,
+      "step": 8040
+    },
+    {
+      "epoch": 0.399821197973577,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0007680182775404789,
+      "loss": 0.7849,
+      "step": 8050
+    },
+    {
+      "epoch": 0.4003178702691964,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0007679785437568293,
+      "loss": 0.7288,
+      "step": 8060
+    },
+    {
+      "epoch": 0.40081454256481575,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0007679388099731798,
+      "loss": 0.7566,
+      "step": 8070
+    },
+    {
+      "epoch": 0.4013112148604351,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0007678990761895302,
+      "loss": 0.7008,
+      "step": 8080
+    },
+    {
+      "epoch": 0.4018078871560544,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0007678593424058806,
+      "loss": 0.754,
+      "step": 8090
+    },
+    {
+      "epoch": 0.40230455945167376,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0007678196086222312,
+      "loss": 0.7161,
+      "step": 8100
+    },
+    {
+      "epoch": 0.40280123174729315,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0007677798748385815,
+      "loss": 0.7508,
+      "step": 8110
+    },
+    {
+      "epoch": 0.4032979040429125,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.000767740141054932,
+      "loss": 0.7183,
+      "step": 8120
+    },
+    {
+      "epoch": 0.40379457633853183,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0007677004072712825,
+      "loss": 0.7561,
+      "step": 8130
+    },
+    {
+      "epoch": 0.40429124863415117,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0007676606734876328,
+      "loss": 0.7123,
+      "step": 8140
+    },
+    {
+      "epoch": 0.40478792092977056,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0007676209397039834,
+      "loss": 0.7231,
+      "step": 8150
+    },
+    {
+      "epoch": 0.4052845932253899,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0007675812059203338,
+      "loss": 0.7267,
+      "step": 8160
+    },
+    {
+      "epoch": 0.40578126552100924,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0007675414721366842,
+      "loss": 0.7173,
+      "step": 8170
+    },
+    {
+      "epoch": 0.4062779378166286,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0007675017383530347,
+      "loss": 0.7031,
+      "step": 8180
+    },
+    {
+      "epoch": 0.4067746101122479,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0007674620045693851,
+      "loss": 0.7208,
+      "step": 8190
+    },
+    {
+      "epoch": 0.4072712824078673,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0007674222707857355,
+      "loss": 0.7613,
+      "step": 8200
+    },
+    {
+      "epoch": 0.40776795470348665,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0007673825370020861,
+      "loss": 0.7161,
+      "step": 8210
+    },
+    {
+      "epoch": 0.408264626999106,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0007673428032184365,
+      "loss": 0.7485,
+      "step": 8220
+    },
+    {
+      "epoch": 0.4087612992947253,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.000767303069434787,
+      "loss": 0.7091,
+      "step": 8230
+    },
+    {
+      "epoch": 0.4092579715903447,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0007672633356511374,
+      "loss": 0.6908,
+      "step": 8240
+    },
+    {
+      "epoch": 0.40975464388596405,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0007672236018674879,
+      "loss": 0.7208,
+      "step": 8250
+    },
+    {
+      "epoch": 0.4102513161815834,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0007671838680838384,
+      "loss": 0.735,
+      "step": 8260
+    },
+    {
+      "epoch": 0.41074798847720273,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0007671441343001888,
+      "loss": 0.7397,
+      "step": 8270
+    },
+    {
+      "epoch": 0.41124466077282207,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0007671044005165392,
+      "loss": 0.7125,
+      "step": 8280
+    },
+    {
+      "epoch": 0.41174133306844146,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0007670646667328897,
+      "loss": 0.7101,
+      "step": 8290
+    },
+    {
+      "epoch": 0.4122380053640608,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00076702493294924,
+      "loss": 0.74,
+      "step": 8300
+    },
+    {
+      "epoch": 0.41273467765968014,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0007669851991655906,
+      "loss": 0.7184,
+      "step": 8310
+    },
+    {
+      "epoch": 0.4132313499552995,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0007669454653819411,
+      "loss": 0.7223,
+      "step": 8320
+    },
+    {
+      "epoch": 0.41372802225091887,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0007669057315982914,
+      "loss": 0.7655,
+      "step": 8330
+    },
+    {
+      "epoch": 0.4142246945465382,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0007668659978146419,
+      "loss": 0.7088,
+      "step": 8340
+    },
+    {
+      "epoch": 0.41472136684215755,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0007668262640309925,
+      "loss": 0.7325,
+      "step": 8350
+    },
+    {
+      "epoch": 0.4152180391377769,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0007667865302473429,
+      "loss": 0.7189,
+      "step": 8360
+    },
+    {
+      "epoch": 0.4157147114333962,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0007667467964636933,
+      "loss": 0.7222,
+      "step": 8370
+    },
+    {
+      "epoch": 0.4162113837290156,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0007667070626800437,
+      "loss": 0.7308,
+      "step": 8380
+    },
+    {
+      "epoch": 0.41670805602463495,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0007666673288963942,
+      "loss": 0.6973,
+      "step": 8390
+    },
+    {
+      "epoch": 0.4172047283202543,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0007666275951127447,
+      "loss": 0.6914,
+      "step": 8400
+    },
+    {
+      "epoch": 0.41770140061587363,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0007665878613290951,
+      "loss": 0.72,
+      "step": 8410
+    },
+    {
+      "epoch": 0.418198072911493,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0007665481275454456,
+      "loss": 0.6967,
+      "step": 8420
+    },
+    {
+      "epoch": 0.41869474520711236,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.000766508393761796,
+      "loss": 0.7139,
+      "step": 8430
+    },
+    {
+      "epoch": 0.4191914175027317,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0007664686599781464,
+      "loss": 0.7003,
+      "step": 8440
+    },
+    {
+      "epoch": 0.41968808979835104,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.000766428926194497,
+      "loss": 0.7177,
+      "step": 8450
+    },
+    {
+      "epoch": 0.4201847620939704,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0007663891924108474,
+      "loss": 0.7192,
+      "step": 8460
+    },
+    {
+      "epoch": 0.42068143438958977,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0007663494586271978,
+      "loss": 0.7029,
+      "step": 8470
+    },
+    {
+      "epoch": 0.4211781066852091,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0007663097248435483,
+      "loss": 0.6977,
+      "step": 8480
+    },
+    {
+      "epoch": 0.42167477898082845,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0007662699910598986,
+      "loss": 0.7426,
+      "step": 8490
+    },
+    {
+      "epoch": 0.4221714512764478,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0007662302572762491,
+      "loss": 0.7655,
+      "step": 8500
+    },
+    {
+      "epoch": 0.4226681235720671,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0007661905234925997,
+      "loss": 0.7211,
+      "step": 8510
+    },
+    {
+      "epoch": 0.4231647958676865,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0007661507897089501,
+      "loss": 0.7069,
+      "step": 8520
+    },
+    {
+      "epoch": 0.42366146816330585,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0007661110559253005,
+      "loss": 0.7189,
+      "step": 8530
+    },
+    {
+      "epoch": 0.4241581404589252,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.000766071322141651,
+      "loss": 0.7167,
+      "step": 8540
+    },
+    {
+      "epoch": 0.42465481275454453,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0007660315883580014,
+      "loss": 0.7307,
+      "step": 8550
+    },
+    {
+      "epoch": 0.4251514850501639,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0007659918545743519,
+      "loss": 0.7289,
+      "step": 8560
+    },
+    {
+      "epoch": 0.42564815734578326,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0007659521207907023,
+      "loss": 0.7157,
+      "step": 8570
+    },
+    {
+      "epoch": 0.4261448296414026,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0007659123870070528,
+      "loss": 0.7354,
+      "step": 8580
+    },
+    {
+      "epoch": 0.42664150193702194,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0007658726532234032,
+      "loss": 0.7431,
+      "step": 8590
+    },
+    {
+      "epoch": 0.4271381742326413,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0007658329194397536,
+      "loss": 0.7369,
+      "step": 8600
+    },
+    {
+      "epoch": 0.42763484652826067,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0007657931856561042,
+      "loss": 0.7348,
+      "step": 8610
+    },
+    {
+      "epoch": 0.42813151882388,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0007657534518724546,
+      "loss": 0.7231,
+      "step": 8620
+    },
+    {
+      "epoch": 0.42862819111949935,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.000765713718088805,
+      "loss": 0.7217,
+      "step": 8630
+    },
+    {
+      "epoch": 0.4291248634151187,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0007656739843051555,
+      "loss": 0.7138,
+      "step": 8640
+    },
+    {
+      "epoch": 0.4296215357107381,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.000765634250521506,
+      "loss": 0.7112,
+      "step": 8650
+    },
+    {
+      "epoch": 0.4301182080063574,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0007655945167378564,
+      "loss": 0.7388,
+      "step": 8660
+    },
+    {
+      "epoch": 0.43061488030197675,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0007655547829542069,
+      "loss": 0.7088,
+      "step": 8670
+    },
+    {
+      "epoch": 0.4311115525975961,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0007655150491705573,
+      "loss": 0.7148,
+      "step": 8680
+    },
+    {
+      "epoch": 0.43160822489321543,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0007654753153869077,
+      "loss": 0.7137,
+      "step": 8690
+    },
+    {
+      "epoch": 0.4321048971888348,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0007654355816032582,
+      "loss": 0.6875,
+      "step": 8700
+    },
+    {
+      "epoch": 0.43260156948445416,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0007653958478196087,
+      "loss": 0.6891,
+      "step": 8710
+    },
+    {
+      "epoch": 0.4330982417800735,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0007653561140359591,
+      "loss": 0.7064,
+      "step": 8720
+    },
+    {
+      "epoch": 0.43359491407569284,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0007653163802523096,
+      "loss": 0.7231,
+      "step": 8730
+    },
+    {
+      "epoch": 0.43409158637131223,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00076527664646866,
+      "loss": 0.7066,
+      "step": 8740
+    },
+    {
+      "epoch": 0.43458825866693157,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0007652369126850104,
+      "loss": 0.7,
+      "step": 8750
+    },
+    {
+      "epoch": 0.4350849309625509,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0007651971789013609,
+      "loss": 0.7378,
+      "step": 8760
+    },
+    {
+      "epoch": 0.43558160325817025,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0007651574451177114,
+      "loss": 0.7336,
+      "step": 8770
+    },
+    {
+      "epoch": 0.4360782755537896,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0007651177113340618,
+      "loss": 0.7575,
+      "step": 8780
+    },
+    {
+      "epoch": 0.436574947849409,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0007650779775504122,
+      "loss": 0.7354,
+      "step": 8790
+    },
+    {
+      "epoch": 0.4370716201450283,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0007650382437667627,
+      "loss": 0.7177,
+      "step": 8800
+    },
+    {
+      "epoch": 0.43756829244064765,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0007649985099831133,
+      "loss": 0.7363,
+      "step": 8810
+    },
+    {
+      "epoch": 0.438064964736267,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0007649587761994636,
+      "loss": 0.7019,
+      "step": 8820
+    },
+    {
+      "epoch": 0.4385616370318864,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0007649190424158141,
+      "loss": 0.6998,
+      "step": 8830
+    },
+    {
+      "epoch": 0.4390583093275057,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0007648793086321645,
+      "loss": 0.7236,
+      "step": 8840
+    },
+    {
+      "epoch": 0.43955498162312506,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0007648395748485149,
+      "loss": 0.7097,
+      "step": 8850
+    },
+    {
+      "epoch": 0.4400516539187444,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0007647998410648655,
+      "loss": 0.6748,
+      "step": 8860
+    },
+    {
+      "epoch": 0.44054832621436374,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0007647601072812159,
+      "loss": 0.6921,
+      "step": 8870
+    },
+    {
+      "epoch": 0.44104499850998313,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0007647203734975663,
+      "loss": 0.7079,
+      "step": 8880
+    },
+    {
+      "epoch": 0.44154167080560247,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0007646806397139168,
+      "loss": 0.7137,
+      "step": 8890
+    },
+    {
+      "epoch": 0.4420383431012218,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0007646409059302672,
+      "loss": 0.7328,
+      "step": 8900
+    },
+    {
+      "epoch": 0.44253501539684115,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0007646011721466177,
+      "loss": 0.7409,
+      "step": 8910
+    },
+    {
+      "epoch": 0.44303168769246054,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0007645614383629682,
+      "loss": 0.7505,
+      "step": 8920
+    },
+    {
+      "epoch": 0.4435283599880799,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0007645217045793186,
+      "loss": 0.7348,
+      "step": 8930
+    },
+    {
+      "epoch": 0.4440250322836992,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.000764481970795669,
+      "loss": 0.7477,
+      "step": 8940
+    },
+    {
+      "epoch": 0.44452170457931856,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0007644422370120194,
+      "loss": 0.71,
+      "step": 8950
+    },
+    {
+      "epoch": 0.4450183768749379,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00076440250322837,
+      "loss": 0.7137,
+      "step": 8960
+    },
+    {
+      "epoch": 0.4455150491705573,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0007643627694447205,
+      "loss": 0.7641,
+      "step": 8970
+    },
+    {
+      "epoch": 0.4460117214661766,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0007643230356610708,
+      "loss": 0.699,
+      "step": 8980
+    },
+    {
+      "epoch": 0.44650839376179596,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0007642833018774213,
+      "loss": 0.6906,
+      "step": 8990
+    },
+    {
+      "epoch": 0.4470050660574153,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0007642435680937718,
+      "loss": 0.7058,
+      "step": 9000
+    },
+    {
+      "epoch": 0.4475017383530347,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0007642038343101222,
+      "loss": 0.7377,
+      "step": 9010
+    },
+    {
+      "epoch": 0.44799841064865403,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0007641641005264727,
+      "loss": 0.6808,
+      "step": 9020
+    },
+    {
+      "epoch": 0.44849508294427337,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0007641243667428231,
+      "loss": 0.7154,
+      "step": 9030
+    },
+    {
+      "epoch": 0.4489917552398927,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0007640846329591735,
+      "loss": 0.7077,
+      "step": 9040
+    },
+    {
+      "epoch": 0.44948842753551205,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.000764044899175524,
+      "loss": 0.6977,
+      "step": 9050
+    },
+    {
+      "epoch": 0.44998509983113144,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0007640051653918745,
+      "loss": 0.7314,
+      "step": 9060
+    },
+    {
+      "epoch": 0.4504817721267508,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0007639654316082249,
+      "loss": 0.7332,
+      "step": 9070
+    },
+    {
+      "epoch": 0.4509784444223701,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0007639256978245754,
+      "loss": 0.7256,
+      "step": 9080
+    },
+    {
+      "epoch": 0.45147511671798946,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0007638859640409258,
+      "loss": 0.6945,
+      "step": 9090
+    },
+    {
+      "epoch": 0.4519717890136088,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0007638462302572763,
+      "loss": 0.7337,
+      "step": 9100
+    },
+    {
+      "epoch": 0.4524684613092282,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0007638064964736268,
+      "loss": 0.7143,
+      "step": 9110
+    },
+    {
+      "epoch": 0.4529651336048475,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0007637667626899772,
+      "loss": 0.7248,
+      "step": 9120
+    },
+    {
+      "epoch": 0.45346180590046686,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0007637270289063277,
+      "loss": 0.7323,
+      "step": 9130
+    },
+    {
+      "epoch": 0.4539584781960862,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0007636872951226781,
+      "loss": 0.7023,
+      "step": 9140
+    },
+    {
+      "epoch": 0.4544551504917056,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0007636475613390285,
+      "loss": 0.7105,
+      "step": 9150
+    },
+    {
+      "epoch": 0.45495182278732493,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0007636078275553791,
+      "loss": 0.7187,
+      "step": 9160
+    },
+    {
+      "epoch": 0.45544849508294427,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0007635680937717294,
+      "loss": 0.7122,
+      "step": 9170
+    },
+    {
+      "epoch": 0.4559451673785636,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0007635283599880799,
+      "loss": 0.7229,
+      "step": 9180
+    },
+    {
+      "epoch": 0.45644183967418295,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0007634886262044304,
+      "loss": 0.7008,
+      "step": 9190
+    },
+    {
+      "epoch": 0.45693851196980234,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0007634488924207807,
+      "loss": 0.6972,
+      "step": 9200
+    },
+    {
+      "epoch": 0.4574351842654217,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0007634091586371313,
+      "loss": 0.6998,
+      "step": 9210
+    },
+    {
+      "epoch": 0.457931856561041,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0007633694248534817,
+      "loss": 0.7409,
+      "step": 9220
+    },
+    {
+      "epoch": 0.45842852885666036,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0007633296910698321,
+      "loss": 0.7144,
+      "step": 9230
+    },
+    {
+      "epoch": 0.45892520115227975,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0007632899572861826,
+      "loss": 0.7273,
+      "step": 9240
+    },
+    {
+      "epoch": 0.4594218734478991,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.000763250223502533,
+      "loss": 0.7022,
+      "step": 9250
+    },
+    {
+      "epoch": 0.4599185457435184,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0007632104897188836,
+      "loss": 0.7172,
+      "step": 9260
+    },
+    {
+      "epoch": 0.46041521803913776,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.000763170755935234,
+      "loss": 0.7561,
+      "step": 9270
+    },
+    {
+      "epoch": 0.4609118903347571,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0007631310221515844,
+      "loss": 0.7032,
+      "step": 9280
+    },
+    {
+      "epoch": 0.4614085626303765,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0007630912883679349,
+      "loss": 0.7463,
+      "step": 9290
+    },
+    {
+      "epoch": 0.46190523492599583,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0007630515545842853,
+      "loss": 0.7138,
+      "step": 9300
+    },
+    {
+      "epoch": 0.46240190722161517,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0007630118208006358,
+      "loss": 0.7258,
+      "step": 9310
+    },
+    {
+      "epoch": 0.4628985795172345,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0007629720870169863,
+      "loss": 0.7383,
+      "step": 9320
+    },
+    {
+      "epoch": 0.4633952518128539,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0007629323532333367,
+      "loss": 0.7449,
+      "step": 9330
+    },
+    {
+      "epoch": 0.46389192410847324,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0007628926194496871,
+      "loss": 0.7348,
+      "step": 9340
+    },
+    {
+      "epoch": 0.4643885964040926,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0007628528856660376,
+      "loss": 0.7234,
+      "step": 9350
+    },
+    {
+      "epoch": 0.4648852686997119,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.000762813151882388,
+      "loss": 0.6861,
+      "step": 9360
+    },
+    {
+      "epoch": 0.46538194099533126,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0007627734180987385,
+      "loss": 0.6995,
+      "step": 9370
+    },
+    {
+      "epoch": 0.46587861329095065,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.000762733684315089,
+      "loss": 0.6967,
+      "step": 9380
+    },
+    {
+      "epoch": 0.46637528558657,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0007626939505314393,
+      "loss": 0.6964,
+      "step": 9390
+    },
+    {
+      "epoch": 0.4668719578821893,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0007626542167477898,
+      "loss": 0.7155,
+      "step": 9400
+    },
+    {
+      "epoch": 0.46736863017780866,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0007626144829641404,
+      "loss": 0.7281,
+      "step": 9410
+    },
+    {
+      "epoch": 0.46786530247342806,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0007625747491804908,
+      "loss": 0.7683,
+      "step": 9420
+    },
+    {
+      "epoch": 0.4683619747690474,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0007625350153968412,
+      "loss": 0.6822,
+      "step": 9430
+    },
+    {
+      "epoch": 0.46885864706466673,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0007624952816131916,
+      "loss": 0.7145,
+      "step": 9440
+    },
+    {
+      "epoch": 0.46935531936028607,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0007624555478295421,
+      "loss": 0.7266,
+      "step": 9450
+    },
+    {
+      "epoch": 0.4698519916559054,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0007624158140458926,
+      "loss": 0.7053,
+      "step": 9460
+    },
+    {
+      "epoch": 0.4703486639515248,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.000762376080262243,
+      "loss": 0.6892,
+      "step": 9470
+    },
+    {
+      "epoch": 0.47084533624714414,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0007623363464785935,
+      "loss": 0.7253,
+      "step": 9480
+    },
+    {
+      "epoch": 0.4713420085427635,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0007622966126949439,
+      "loss": 0.7233,
+      "step": 9490
+    },
+    {
+      "epoch": 0.4718386808383828,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0007622568789112943,
+      "loss": 0.7241,
+      "step": 9500
+    },
+    {
+      "epoch": 0.4723353531340022,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0007622171451276449,
+      "loss": 0.7205,
+      "step": 9510
+    },
+    {
+      "epoch": 0.47283202542962155,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0007621774113439953,
+      "loss": 0.7085,
+      "step": 9520
+    },
+    {
+      "epoch": 0.4733286977252409,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0007621376775603457,
+      "loss": 0.6921,
+      "step": 9530
+    },
+    {
+      "epoch": 0.4738253700208602,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0007620979437766962,
+      "loss": 0.7269,
+      "step": 9540
+    },
+    {
+      "epoch": 0.47432204231647956,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0007620582099930466,
+      "loss": 0.7659,
+      "step": 9550
+    },
+    {
+      "epoch": 0.47481871461209896,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.000762018476209397,
+      "loss": 0.7237,
+      "step": 9560
+    },
+    {
+      "epoch": 0.4753153869077183,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0007619787424257476,
+      "loss": 0.7036,
+      "step": 9570
+    },
+    {
+      "epoch": 0.47581205920333763,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.000761939008642098,
+      "loss": 0.7232,
+      "step": 9580
+    },
+    {
+      "epoch": 0.47630873149895697,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0007618992748584484,
+      "loss": 0.6916,
+      "step": 9590
+    },
+    {
+      "epoch": 0.47680540379457637,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0007618595410747989,
+      "loss": 0.6853,
+      "step": 9600
+    },
+    {
+      "epoch": 0.4773020760901957,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0007618198072911494,
+      "loss": 0.6997,
+      "step": 9610
+    },
+    {
+      "epoch": 0.47779874838581504,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0007617800735074998,
+      "loss": 0.7772,
+      "step": 9620
+    },
+    {
+      "epoch": 0.4782954206814344,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0007617403397238502,
+      "loss": 0.7037,
+      "step": 9630
+    },
+    {
+      "epoch": 0.4787920929770537,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0007617006059402007,
+      "loss": 0.7249,
+      "step": 9640
+    },
+    {
+      "epoch": 0.4792887652726731,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0007616608721565511,
+      "loss": 0.7063,
+      "step": 9650
+    },
+    {
+      "epoch": 0.47978543756829245,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0007616211383729015,
+      "loss": 0.6947,
+      "step": 9660
+    },
+    {
+      "epoch": 0.4802821098639118,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0007615814045892521,
+      "loss": 0.6985,
+      "step": 9670
+    },
+    {
+      "epoch": 0.4807787821595311,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0007615416708056025,
+      "loss": 0.7313,
+      "step": 9680
+    },
+    {
+      "epoch": 0.48127545445515046,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0007615019370219529,
+      "loss": 0.708,
+      "step": 9690
+    },
+    {
+      "epoch": 0.48177212675076986,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0007614622032383034,
+      "loss": 0.7357,
+      "step": 9700
+    },
+    {
+      "epoch": 0.4822687990463892,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0007614224694546539,
+      "loss": 0.7357,
+      "step": 9710
+    },
+    {
+      "epoch": 0.48276547134200853,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0007613827356710043,
+      "loss": 0.6737,
+      "step": 9720
+    },
+    {
+      "epoch": 0.4832621436376279,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0007613430018873548,
+      "loss": 0.6925,
+      "step": 9730
+    },
+    {
+      "epoch": 0.48375881593324727,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0007613032681037052,
+      "loss": 0.6754,
+      "step": 9740
+    },
+    {
+      "epoch": 0.4842554882288666,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0007612635343200556,
+      "loss": 0.7491,
+      "step": 9750
+    },
+    {
+      "epoch": 0.48475216052448594,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0007612238005364062,
+      "loss": 0.7246,
+      "step": 9760
+    },
+    {
+      "epoch": 0.4852488328201053,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0007611840667527566,
+      "loss": 0.6921,
+      "step": 9770
+    },
+    {
+      "epoch": 0.4857455051157246,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.000761144332969107,
+      "loss": 0.6853,
+      "step": 9780
+    },
+    {
+      "epoch": 0.486242177411344,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0007611045991854575,
+      "loss": 0.6736,
+      "step": 9790
+    },
+    {
+      "epoch": 0.48673884970696335,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0007610648654018079,
+      "loss": 0.7246,
+      "step": 9800
+    },
+    {
+      "epoch": 0.4872355220025827,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0007610251316181583,
+      "loss": 0.7048,
+      "step": 9810
+    },
+    {
+      "epoch": 0.487732194298202,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0007609853978345088,
+      "loss": 0.688,
+      "step": 9820
+    },
+    {
+      "epoch": 0.4882288665938214,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0007609456640508593,
+      "loss": 0.7032,
+      "step": 9830
+    },
+    {
+      "epoch": 0.48872553888944076,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0007609059302672097,
+      "loss": 0.6997,
+      "step": 9840
+    },
+    {
+      "epoch": 0.4892222111850601,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0007608661964835601,
+      "loss": 0.6674,
+      "step": 9850
+    },
+    {
+      "epoch": 0.48971888348067943,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0007608264626999106,
+      "loss": 0.7512,
+      "step": 9860
+    },
+    {
+      "epoch": 0.4902155557762988,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0007607867289162612,
+      "loss": 0.7171,
+      "step": 9870
+    },
+    {
+      "epoch": 0.49071222807191817,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0007607469951326115,
+      "loss": 0.7149,
+      "step": 9880
+    },
+    {
+      "epoch": 0.4912089003675375,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.000760707261348962,
+      "loss": 0.6987,
+      "step": 9890
+    },
+    {
+      "epoch": 0.49170557266315684,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0007606675275653124,
+      "loss": 0.728,
+      "step": 9900
+    },
+    {
+      "epoch": 0.4922022449587762,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0007606277937816628,
+      "loss": 0.669,
+      "step": 9910
+    },
+    {
+      "epoch": 0.4926989172543956,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0007605880599980134,
+      "loss": 0.6901,
+      "step": 9920
+    },
+    {
+      "epoch": 0.4931955895500149,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0007605483262143638,
+      "loss": 0.7053,
+      "step": 9930
+    },
+    {
+      "epoch": 0.49369226184563425,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0007605085924307142,
+      "loss": 0.7132,
+      "step": 9940
+    },
+    {
+      "epoch": 0.4941889341412536,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0007604688586470647,
+      "loss": 0.6702,
+      "step": 9950
+    },
+    {
+      "epoch": 0.4946856064368729,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0007604291248634151,
+      "loss": 0.7449,
+      "step": 9960
+    },
+    {
+      "epoch": 0.4951822787324923,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0007603893910797656,
+      "loss": 0.6888,
+      "step": 9970
+    },
+    {
+      "epoch": 0.49567895102811166,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0007603496572961161,
+      "loss": 0.7041,
+      "step": 9980
+    },
+    {
+      "epoch": 0.496175623323731,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0007603099235124665,
+      "loss": 0.7158,
+      "step": 9990
+    },
+    {
+      "epoch": 0.49667229561935033,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.000760270189728817,
+      "loss": 0.7205,
+      "step": 10000
+    },
+    {
+      "epoch": 0.49716896791496973,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0007602304559451674,
+      "loss": 0.7384,
+      "step": 10010
+    },
+    {
+      "epoch": 0.49766564021058907,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0007601907221615179,
+      "loss": 0.7441,
+      "step": 10020
+    },
+    {
+      "epoch": 0.4981623125062084,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0007601509883778684,
+      "loss": 0.6916,
+      "step": 10030
+    },
+    {
+      "epoch": 0.49865898480182774,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0007601112545942187,
+      "loss": 0.7187,
+      "step": 10040
+    },
+    {
+      "epoch": 0.4991556570974471,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007600715208105692,
+      "loss": 0.7011,
+      "step": 10050
+    },
+    {
+      "epoch": 0.4996523293930665,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0007600317870269198,
+      "loss": 0.6988,
+      "step": 10060
+    },
+    {
+      "epoch": 0.5001490016886858,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0007599920532432701,
+      "loss": 0.6829,
+      "step": 10070
+    },
+    {
+      "epoch": 0.5006456739843052,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0007599523194596206,
+      "loss": 0.7256,
+      "step": 10080
+    },
+    {
+      "epoch": 0.5011423462799245,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.000759912585675971,
+      "loss": 0.741,
+      "step": 10090
+    },
+    {
+      "epoch": 0.5016390185755438,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0007598728518923214,
+      "loss": 0.6929,
+      "step": 10100
+    },
+    {
+      "epoch": 0.5021356908711632,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0007598331181086719,
+      "loss": 0.7239,
+      "step": 10110
+    },
+    {
+      "epoch": 0.5026323631667825,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0007597933843250224,
+      "loss": 0.7027,
+      "step": 10120
+    },
+    {
+      "epoch": 0.503129035462402,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007597536505413728,
+      "loss": 0.6563,
+      "step": 10130
+    },
+    {
+      "epoch": 0.5036257077580213,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0007597139167577233,
+      "loss": 0.727,
+      "step": 10140
+    },
+    {
+      "epoch": 0.5041223800536406,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.0007596741829740737,
+      "loss": 0.7071,
+      "step": 10150
+    },
+    {
+      "epoch": 0.50461905234926,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0007596344491904242,
+      "loss": 0.7103,
+      "step": 10160
+    },
+    {
+      "epoch": 0.5051157246448793,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0007595947154067747,
+      "loss": 0.6813,
+      "step": 10170
+    },
+    {
+      "epoch": 0.5056123969404986,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0007595549816231251,
+      "loss": 0.7,
+      "step": 10180
+    },
+    {
+      "epoch": 0.506109069236118,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0007595152478394756,
+      "loss": 0.7448,
+      "step": 10190
+    },
+    {
+      "epoch": 0.5066057415317373,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.000759475514055826,
+      "loss": 0.6848,
+      "step": 10200
+    },
+    {
+      "epoch": 0.5071024138273567,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0007594357802721764,
+      "loss": 0.7022,
+      "step": 10210
+    },
+    {
+      "epoch": 0.5075990861229761,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.000759396046488527,
+      "loss": 0.6825,
+      "step": 10220
+    },
+    {
+      "epoch": 0.5080957584185954,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0007593563127048773,
+      "loss": 0.6914,
+      "step": 10230
+    },
+    {
+      "epoch": 0.5085924307142148,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0007593165789212278,
+      "loss": 0.737,
+      "step": 10240
+    },
+    {
+      "epoch": 0.5090891030098341,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0007592768451375783,
+      "loss": 0.6901,
+      "step": 10250
+    },
+    {
+      "epoch": 0.5095857753054535,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0007592371113539286,
+      "loss": 0.7328,
+      "step": 10260
+    },
+    {
+      "epoch": 0.5100824476010728,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0007591973775702792,
+      "loss": 0.7049,
+      "step": 10270
+    },
+    {
+      "epoch": 0.5105791198966921,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0007591576437866297,
+      "loss": 0.6783,
+      "step": 10280
+    },
+    {
+      "epoch": 0.5110757921923115,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00075911791000298,
+      "loss": 0.7185,
+      "step": 10290
+    },
+    {
+      "epoch": 0.5115724644879308,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0007590781762193305,
+      "loss": 0.7185,
+      "step": 10300
+    },
+    {
+      "epoch": 0.5120691367835503,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0007590384424356809,
+      "loss": 0.7252,
+      "step": 10310
+    },
+    {
+      "epoch": 0.5125658090791696,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0007589987086520315,
+      "loss": 0.7113,
+      "step": 10320
+    },
+    {
+      "epoch": 0.5130624813747889,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007589589748683819,
+      "loss": 0.6794,
+      "step": 10330
+    },
+    {
+      "epoch": 0.5135591536704083,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0007589192410847323,
+      "loss": 0.7233,
+      "step": 10340
+    },
+    {
+      "epoch": 0.5140558259660276,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007588795073010828,
+      "loss": 0.6812,
+      "step": 10350
+    },
+    {
+      "epoch": 0.514552498261647,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0007588397735174332,
+      "loss": 0.7091,
+      "step": 10360
+    },
+    {
+      "epoch": 0.5150491705572663,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0007588000397337837,
+      "loss": 0.734,
+      "step": 10370
+    },
+    {
+      "epoch": 0.5155458428528856,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007587603059501342,
+      "loss": 0.7,
+      "step": 10380
+    },
+    {
+      "epoch": 0.516042515148505,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0007587205721664846,
+      "loss": 0.6981,
+      "step": 10390
+    },
+    {
+      "epoch": 0.5165391874441244,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.000758680838382835,
+      "loss": 0.7074,
+      "step": 10400
+    },
+    {
+      "epoch": 0.5170358597397438,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0007586411045991855,
+      "loss": 0.722,
+      "step": 10410
+    },
+    {
+      "epoch": 0.5175325320353631,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0007586013708155359,
+      "loss": 0.7269,
+      "step": 10420
+    },
+    {
+      "epoch": 0.5180292043309824,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007585616370318864,
+      "loss": 0.7256,
+      "step": 10430
+    },
+    {
+      "epoch": 0.5185258766266018,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0007585219032482369,
+      "loss": 0.711,
+      "step": 10440
+    },
+    {
+      "epoch": 0.5190225489222211,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0007584821694645873,
+      "loss": 0.6722,
+      "step": 10450
+    },
+    {
+      "epoch": 0.5195192212178404,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0007584424356809377,
+      "loss": 0.6736,
+      "step": 10460
+    },
+    {
+      "epoch": 0.5200158935134598,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0007584027018972883,
+      "loss": 0.6942,
+      "step": 10470
+    },
+    {
+      "epoch": 0.5205125658090791,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0007583629681136387,
+      "loss": 0.6936,
+      "step": 10480
+    },
+    {
+      "epoch": 0.5210092381046986,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0007583232343299891,
+      "loss": 0.7083,
+      "step": 10490
+    },
+    {
+      "epoch": 0.5215059104003179,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0007582835005463395,
+      "loss": 0.7019,
+      "step": 10500
+    },
+    {
+      "epoch": 0.5220025826959372,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00075824376676269,
+      "loss": 0.7379,
+      "step": 10510
+    },
+    {
+      "epoch": 0.5224992549915566,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0007582040329790405,
+      "loss": 0.6855,
+      "step": 10520
+    },
+    {
+      "epoch": 0.5229959272871759,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0007581642991953909,
+      "loss": 0.7362,
+      "step": 10530
+    },
+    {
+      "epoch": 0.5234925995827953,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0007581245654117414,
+      "loss": 0.6975,
+      "step": 10540
+    },
+    {
+      "epoch": 0.5239892718784146,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0007580848316280918,
+      "loss": 0.668,
+      "step": 10550
+    },
+    {
+      "epoch": 0.5244859441740339,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007580450978444422,
+      "loss": 0.7077,
+      "step": 10560
+    },
+    {
+      "epoch": 0.5249826164696533,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0007580053640607928,
+      "loss": 0.7183,
+      "step": 10570
+    },
+    {
+      "epoch": 0.5254792887652727,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0007579656302771432,
+      "loss": 0.6935,
+      "step": 10580
+    },
+    {
+      "epoch": 0.5259759610608921,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0007579258964934936,
+      "loss": 0.7131,
+      "step": 10590
+    },
+    {
+      "epoch": 0.5264726333565114,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0007578861627098441,
+      "loss": 0.6672,
+      "step": 10600
+    },
+    {
+      "epoch": 0.5269693056521307,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0007578464289261945,
+      "loss": 0.6699,
+      "step": 10610
+    },
+    {
+      "epoch": 0.5274659779477501,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000757806695142545,
+      "loss": 0.6993,
+      "step": 10620
+    },
+    {
+      "epoch": 0.5279626502433694,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0007577669613588955,
+      "loss": 0.6946,
+      "step": 10630
+    },
+    {
+      "epoch": 0.5284593225389888,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0007577272275752459,
+      "loss": 0.6683,
+      "step": 10640
+    },
+    {
+      "epoch": 0.5289559948346081,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007576874937915963,
+      "loss": 0.7298,
+      "step": 10650
+    },
+    {
+      "epoch": 0.5294526671302274,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0007576477600079468,
+      "loss": 0.6767,
+      "step": 10660
+    },
+    {
+      "epoch": 0.5299493394258469,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0007576080262242973,
+      "loss": 0.7152,
+      "step": 10670
+    },
+    {
+      "epoch": 0.5304460117214662,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007575682924406477,
+      "loss": 0.7154,
+      "step": 10680
+    },
+    {
+      "epoch": 0.5309426840170856,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0007575285586569981,
+      "loss": 0.6851,
+      "step": 10690
+    },
+    {
+      "epoch": 0.5314393563127049,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0007574888248733486,
+      "loss": 0.7228,
+      "step": 10700
+    },
+    {
+      "epoch": 0.5319360286083242,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.000757449091089699,
+      "loss": 0.6955,
+      "step": 10710
+    },
+    {
+      "epoch": 0.5324327009039436,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007574093573060495,
+      "loss": 0.6805,
+      "step": 10720
+    },
+    {
+      "epoch": 0.5329293731995629,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0007573696235224,
+      "loss": 0.6945,
+      "step": 10730
+    },
+    {
+      "epoch": 0.5334260454951822,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0007573298897387504,
+      "loss": 0.6913,
+      "step": 10740
+    },
+    {
+      "epoch": 0.5339227177908016,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0007572901559551008,
+      "loss": 0.6735,
+      "step": 10750
+    },
+    {
+      "epoch": 0.534419390086421,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0007572504221714513,
+      "loss": 0.7134,
+      "step": 10760
+    },
+    {
+      "epoch": 0.5349160623820404,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0007572106883878018,
+      "loss": 0.6831,
+      "step": 10770
+    },
+    {
+      "epoch": 0.5354127346776597,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0007571709546041522,
+      "loss": 0.7085,
+      "step": 10780
+    },
+    {
+      "epoch": 0.535909406973279,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0007571312208205027,
+      "loss": 0.6748,
+      "step": 10790
+    },
+    {
+      "epoch": 0.5364060792688984,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0007570914870368531,
+      "loss": 0.7158,
+      "step": 10800
+    },
+    {
+      "epoch": 0.5369027515645177,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0007570517532532035,
+      "loss": 0.7052,
+      "step": 10810
+    },
+    {
+      "epoch": 0.5373994238601371,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0007570120194695541,
+      "loss": 0.7261,
+      "step": 10820
+    },
+    {
+      "epoch": 0.5378960961557564,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007569722856859045,
+      "loss": 0.7455,
+      "step": 10830
+    },
+    {
+      "epoch": 0.5383927684513757,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0007569325519022549,
+      "loss": 0.6688,
+      "step": 10840
+    },
+    {
+      "epoch": 0.5388894407469952,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0007568928181186054,
+      "loss": 0.6693,
+      "step": 10850
+    },
+    {
+      "epoch": 0.5393861130426145,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007568530843349558,
+      "loss": 0.7121,
+      "step": 10860
+    },
+    {
+      "epoch": 0.5398827853382339,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0007568133505513063,
+      "loss": 0.6963,
+      "step": 10870
+    },
+    {
+      "epoch": 0.5403794576338532,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0007567736167676568,
+      "loss": 0.7068,
+      "step": 10880
+    },
+    {
+      "epoch": 0.5408761299294725,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0007567338829840072,
+      "loss": 0.6676,
+      "step": 10890
+    },
+    {
+      "epoch": 0.5413728022250919,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0007566941492003577,
+      "loss": 0.6815,
+      "step": 10900
+    },
+    {
+      "epoch": 0.5418694745207112,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.000756654415416708,
+      "loss": 0.7308,
+      "step": 10910
+    },
+    {
+      "epoch": 0.5423661468163306,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0007566146816330586,
+      "loss": 0.6691,
+      "step": 10920
+    },
+    {
+      "epoch": 0.5428628191119499,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0007565749478494091,
+      "loss": 0.7001,
+      "step": 10930
+    },
+    {
+      "epoch": 0.5433594914075693,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0007565352140657594,
+      "loss": 0.6703,
+      "step": 10940
+    },
+    {
+      "epoch": 0.5438561637031887,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0007564954802821099,
+      "loss": 0.6909,
+      "step": 10950
+    },
+    {
+      "epoch": 0.544352835998808,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0007564557464984603,
+      "loss": 0.6848,
+      "step": 10960
+    },
+    {
+      "epoch": 0.5448495082944274,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0007564160127148107,
+      "loss": 0.6682,
+      "step": 10970
+    },
+    {
+      "epoch": 0.5453461805900467,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0007563762789311613,
+      "loss": 0.7165,
+      "step": 10980
+    },
+    {
+      "epoch": 0.545842852885666,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0007563365451475117,
+      "loss": 0.6916,
+      "step": 10990
+    },
+    {
+      "epoch": 0.5463395251812854,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0007562968113638621,
+      "loss": 0.7004,
+      "step": 11000
+    },
+    {
+      "epoch": 0.5468361974769047,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0007562570775802126,
+      "loss": 0.7482,
+      "step": 11010
+    },
+    {
+      "epoch": 0.547332869772524,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.000756217343796563,
+      "loss": 0.6794,
+      "step": 11020
+    },
+    {
+      "epoch": 0.5478295420681435,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0007561776100129135,
+      "loss": 0.7104,
+      "step": 11030
+    },
+    {
+      "epoch": 0.5483262143637628,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.000756137876229264,
+      "loss": 0.681,
+      "step": 11040
+    },
+    {
+      "epoch": 0.5488228866593822,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0007560981424456144,
+      "loss": 0.6743,
+      "step": 11050
+    },
+    {
+      "epoch": 0.5493195589550015,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0007560584086619649,
+      "loss": 0.7129,
+      "step": 11060
+    },
+    {
+      "epoch": 0.5498162312506208,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0007560186748783154,
+      "loss": 0.7309,
+      "step": 11070
+    },
+    {
+      "epoch": 0.5503129035462402,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0007559789410946658,
+      "loss": 0.6973,
+      "step": 11080
+    },
+    {
+      "epoch": 0.5508095758418595,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0007559392073110163,
+      "loss": 0.6736,
+      "step": 11090
+    },
+    {
+      "epoch": 0.5513062481374789,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0007558994735273666,
+      "loss": 0.6728,
+      "step": 11100
+    },
+    {
+      "epoch": 0.5518029204330982,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0007558597397437171,
+      "loss": 0.6845,
+      "step": 11110
+    },
+    {
+      "epoch": 0.5522995927287176,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0007558200059600677,
+      "loss": 0.6922,
+      "step": 11120
+    },
+    {
+      "epoch": 0.552796265024337,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.000755780272176418,
+      "loss": 0.6936,
+      "step": 11130
+    },
+    {
+      "epoch": 0.5532929373199563,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007557405383927685,
+      "loss": 0.6968,
+      "step": 11140
+    },
+    {
+      "epoch": 0.5537896096155757,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.000755700804609119,
+      "loss": 0.6866,
+      "step": 11150
+    },
+    {
+      "epoch": 0.554286281911195,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007556610708254693,
+      "loss": 0.6958,
+      "step": 11160
+    },
+    {
+      "epoch": 0.5547829542068143,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0007556213370418198,
+      "loss": 0.6782,
+      "step": 11170
+    },
+    {
+      "epoch": 0.5552796265024337,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0007555816032581703,
+      "loss": 0.6872,
+      "step": 11180
+    },
+    {
+      "epoch": 0.555776298798053,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007555418694745208,
+      "loss": 0.6499,
+      "step": 11190
+    },
+    {
+      "epoch": 0.5562729710936724,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0007555021356908712,
+      "loss": 0.6933,
+      "step": 11200
+    },
+    {
+      "epoch": 0.5567696433892917,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0007554624019072216,
+      "loss": 0.7007,
+      "step": 11210
+    },
+    {
+      "epoch": 0.5572663156849111,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007554226681235722,
+      "loss": 0.7212,
+      "step": 11220
+    },
+    {
+      "epoch": 0.5577629879805305,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0007553829343399226,
+      "loss": 0.6869,
+      "step": 11230
+    },
+    {
+      "epoch": 0.5582596602761498,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.000755343200556273,
+      "loss": 0.7108,
+      "step": 11240
+    },
+    {
+      "epoch": 0.5587563325717692,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0007553034667726235,
+      "loss": 0.7133,
+      "step": 11250
+    },
+    {
+      "epoch": 0.5592530048673885,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0007552637329889739,
+      "loss": 0.6634,
+      "step": 11260
+    },
+    {
+      "epoch": 0.5597496771630078,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0007552239992053243,
+      "loss": 0.7046,
+      "step": 11270
+    },
+    {
+      "epoch": 0.5602463494586272,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0007551842654216749,
+      "loss": 0.6988,
+      "step": 11280
+    },
+    {
+      "epoch": 0.5607430217542465,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007551445316380252,
+      "loss": 0.681,
+      "step": 11290
+    },
+    {
+      "epoch": 0.5612396940498658,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0007551047978543757,
+      "loss": 0.7067,
+      "step": 11300
+    },
+    {
+      "epoch": 0.5617363663454853,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007550650640707262,
+      "loss": 0.6846,
+      "step": 11310
+    },
+    {
+      "epoch": 0.5622330386411046,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0007550253302870765,
+      "loss": 0.6974,
+      "step": 11320
+    },
+    {
+      "epoch": 0.562729710936724,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0007549855965034271,
+      "loss": 0.6877,
+      "step": 11330
+    },
+    {
+      "epoch": 0.5632263832323433,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007549458627197776,
+      "loss": 0.6674,
+      "step": 11340
+    },
+    {
+      "epoch": 0.5637230555279626,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.000754906128936128,
+      "loss": 0.6544,
+      "step": 11350
+    },
+    {
+      "epoch": 0.564219727823582,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0007548663951524784,
+      "loss": 0.6775,
+      "step": 11360
+    },
+    {
+      "epoch": 0.5647164001192013,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0007548266613688288,
+      "loss": 0.6838,
+      "step": 11370
+    },
+    {
+      "epoch": 0.5652130724148207,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0007547869275851794,
+      "loss": 0.6775,
+      "step": 11380
+    },
+    {
+      "epoch": 0.56570974471044,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0007547471938015298,
+      "loss": 0.6907,
+      "step": 11390
+    },
+    {
+      "epoch": 0.5662064170060594,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0007547074600178802,
+      "loss": 0.669,
+      "step": 11400
+    },
+    {
+      "epoch": 0.5667030893016788,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0007546677262342307,
+      "loss": 0.681,
+      "step": 11410
+    },
+    {
+      "epoch": 0.5671997615972981,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007546279924505811,
+      "loss": 0.685,
+      "step": 11420
+    },
+    {
+      "epoch": 0.5676964338929175,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007545882586669316,
+      "loss": 0.7101,
+      "step": 11430
+    },
+    {
+      "epoch": 0.5681931061885368,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0007545485248832821,
+      "loss": 0.7297,
+      "step": 11440
+    },
+    {
+      "epoch": 0.5686897784841561,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007545087910996325,
+      "loss": 0.7116,
+      "step": 11450
+    },
+    {
+      "epoch": 0.5691864507797755,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0007544690573159829,
+      "loss": 0.6823,
+      "step": 11460
+    },
+    {
+      "epoch": 0.5696831230753948,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0007544293235323334,
+      "loss": 0.7039,
+      "step": 11470
+    },
+    {
+      "epoch": 0.5701797953710142,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007543895897486839,
+      "loss": 0.6877,
+      "step": 11480
+    },
+    {
+      "epoch": 0.5706764676666336,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007543498559650343,
+      "loss": 0.7188,
+      "step": 11490
+    },
+    {
+      "epoch": 0.5711731399622529,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007543101221813848,
+      "loss": 0.6998,
+      "step": 11500
+    },
+    {
+      "epoch": 0.5716698122578723,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0007542703883977352,
+      "loss": 0.6943,
+      "step": 11510
+    },
+    {
+      "epoch": 0.5721664845534916,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0007542306546140856,
+      "loss": 0.6909,
+      "step": 11520
+    },
+    {
+      "epoch": 0.572663156849111,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007541909208304362,
+      "loss": 0.6551,
+      "step": 11530
+    },
+    {
+      "epoch": 0.5731598291447303,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0007541511870467866,
+      "loss": 0.6907,
+      "step": 11540
+    },
+    {
+      "epoch": 0.5736565014403496,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.000754111453263137,
+      "loss": 0.6967,
+      "step": 11550
+    },
+    {
+      "epoch": 0.574153173735969,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0007540717194794874,
+      "loss": 0.6866,
+      "step": 11560
+    },
+    {
+      "epoch": 0.5746498460315883,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007540319856958379,
+      "loss": 0.6793,
+      "step": 11570
+    },
+    {
+      "epoch": 0.5751465183272078,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0007539922519121884,
+      "loss": 0.71,
+      "step": 11580
+    },
+    {
+      "epoch": 0.5756431906228271,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0007539525181285388,
+      "loss": 0.6951,
+      "step": 11590
+    },
+    {
+      "epoch": 0.5761398629184464,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0007539127843448893,
+      "loss": 0.6879,
+      "step": 11600
+    },
+    {
+      "epoch": 0.5766365352140658,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0007538730505612397,
+      "loss": 0.6827,
+      "step": 11610
+    },
+    {
+      "epoch": 0.5771332075096851,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007538333167775901,
+      "loss": 0.666,
+      "step": 11620
+    },
+    {
+      "epoch": 0.5776298798053044,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0007537935829939407,
+      "loss": 0.6948,
+      "step": 11630
+    },
+    {
+      "epoch": 0.5781265521009238,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007537538492102911,
+      "loss": 0.6813,
+      "step": 11640
+    },
+    {
+      "epoch": 0.5786232243965431,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0007537141154266415,
+      "loss": 0.6757,
+      "step": 11650
+    },
+    {
+      "epoch": 0.5791198966921625,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.000753674381642992,
+      "loss": 0.7123,
+      "step": 11660
+    },
+    {
+      "epoch": 0.5796165689877819,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0007536346478593424,
+      "loss": 0.7048,
+      "step": 11670
+    },
+    {
+      "epoch": 0.5801132412834012,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0007535949140756929,
+      "loss": 0.6927,
+      "step": 11680
+    },
+    {
+      "epoch": 0.5806099135790206,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0007535551802920434,
+      "loss": 0.6679,
+      "step": 11690
+    },
+    {
+      "epoch": 0.5811065858746399,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0007535154465083938,
+      "loss": 0.729,
+      "step": 11700
+    },
+    {
+      "epoch": 0.5816032581702593,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007534757127247442,
+      "loss": 0.7099,
+      "step": 11710
+    },
+    {
+      "epoch": 0.5820999304658786,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0007534359789410947,
+      "loss": 0.6524,
+      "step": 11720
+    },
+    {
+      "epoch": 0.5825966027614979,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007533962451574452,
+      "loss": 0.7253,
+      "step": 11730
+    },
+    {
+      "epoch": 0.5830932750571173,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007533565113737956,
+      "loss": 0.6756,
+      "step": 11740
+    },
+    {
+      "epoch": 0.5835899473527366,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0007533167775901461,
+      "loss": 0.6726,
+      "step": 11750
+    },
+    {
+      "epoch": 0.5840866196483561,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0007532770438064965,
+      "loss": 0.6854,
+      "step": 11760
+    },
+    {
+      "epoch": 0.5845832919439754,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0007532373100228469,
+      "loss": 0.6881,
+      "step": 11770
+    },
+    {
+      "epoch": 0.5850799642395947,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0007531975762391974,
+      "loss": 0.7094,
+      "step": 11780
+    },
+    {
+      "epoch": 0.5855766365352141,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0007531578424555479,
+      "loss": 0.6962,
+      "step": 11790
+    },
+    {
+      "epoch": 0.5860733088308334,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007531181086718984,
+      "loss": 0.7148,
+      "step": 11800
+    },
+    {
+      "epoch": 0.5865699811264528,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007530783748882487,
+      "loss": 0.6397,
+      "step": 11810
+    },
+    {
+      "epoch": 0.5870666534220721,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0007530386411045992,
+      "loss": 0.6775,
+      "step": 11820
+    },
+    {
+      "epoch": 0.5875633257176914,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0007529989073209497,
+      "loss": 0.7079,
+      "step": 11830
+    },
+    {
+      "epoch": 0.5880599980133108,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0007529591735373001,
+      "loss": 0.7033,
+      "step": 11840
+    },
+    {
+      "epoch": 0.5885566703089302,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0007529194397536506,
+      "loss": 0.6891,
+      "step": 11850
+    },
+    {
+      "epoch": 0.5890533426045496,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.000752879705970001,
+      "loss": 0.6991,
+      "step": 11860
+    },
+    {
+      "epoch": 0.5895500149001689,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0007528399721863514,
+      "loss": 0.7138,
+      "step": 11870
+    },
+    {
+      "epoch": 0.5900466871957882,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.000752800238402702,
+      "loss": 0.694,
+      "step": 11880
+    },
+    {
+      "epoch": 0.5905433594914076,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0007527605046190524,
+      "loss": 0.6693,
+      "step": 11890
+    },
+    {
+      "epoch": 0.5910400317870269,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0007527207708354028,
+      "loss": 0.6626,
+      "step": 11900
+    },
+    {
+      "epoch": 0.5915367040826462,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0007526810370517533,
+      "loss": 0.6696,
+      "step": 11910
+    },
+    {
+      "epoch": 0.5920333763782656,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0007526413032681037,
+      "loss": 0.7056,
+      "step": 11920
+    },
+    {
+      "epoch": 0.5925300486738849,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0007526015694844542,
+      "loss": 0.7229,
+      "step": 11930
+    },
+    {
+      "epoch": 0.5930267209695044,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0007525618357008047,
+      "loss": 0.6533,
+      "step": 11940
+    },
+    {
+      "epoch": 0.5935233932651237,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0007525221019171551,
+      "loss": 0.6976,
+      "step": 11950
+    },
+    {
+      "epoch": 0.594020065560743,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0007524823681335056,
+      "loss": 0.6907,
+      "step": 11960
+    },
+    {
+      "epoch": 0.5945167378563624,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0007524426343498559,
+      "loss": 0.6898,
+      "step": 11970
+    },
+    {
+      "epoch": 0.5950134101519817,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0007524029005662065,
+      "loss": 0.6755,
+      "step": 11980
+    },
+    {
+      "epoch": 0.5955100824476011,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.000752363166782557,
+      "loss": 0.7117,
+      "step": 11990
+    },
+    {
+      "epoch": 0.5960067547432204,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0007523234329989073,
+      "loss": 0.6984,
+      "step": 12000
+    },
+    {
+      "epoch": 0.5965034270388397,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0007522836992152578,
+      "loss": 0.693,
+      "step": 12010
+    },
+    {
+      "epoch": 0.5970000993344591,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0007522439654316083,
+      "loss": 0.6968,
+      "step": 12020
+    },
+    {
+      "epoch": 0.5974967716300785,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0007522042316479587,
+      "loss": 0.726,
+      "step": 12030
+    },
+    {
+      "epoch": 0.5979934439256979,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007521644978643092,
+      "loss": 0.6597,
+      "step": 12040
+    },
+    {
+      "epoch": 0.5984901162213172,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0007521247640806596,
+      "loss": 0.6991,
+      "step": 12050
+    },
+    {
+      "epoch": 0.5989867885169365,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00075208503029701,
+      "loss": 0.69,
+      "step": 12060
+    },
+    {
+      "epoch": 0.5994834608125559,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0007520452965133605,
+      "loss": 0.6338,
+      "step": 12070
+    },
+    {
+      "epoch": 0.5999801331081752,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.000752005562729711,
+      "loss": 0.6749,
+      "step": 12080
+    },
+    {
+      "epoch": 0.6004768054037946,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007519658289460615,
+      "loss": 0.6509,
+      "step": 12090
+    },
+    {
+      "epoch": 0.6009734776994139,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007519260951624119,
+      "loss": 0.6835,
+      "step": 12100
+    },
+    {
+      "epoch": 0.6014701499950332,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0007518863613787623,
+      "loss": 0.6922,
+      "step": 12110
+    },
+    {
+      "epoch": 0.6019668222906527,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0007518466275951128,
+      "loss": 0.6965,
+      "step": 12120
+    },
+    {
+      "epoch": 0.602463494586272,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0007518068938114633,
+      "loss": 0.6965,
+      "step": 12130
+    },
+    {
+      "epoch": 0.6029601668818914,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0007517671600278137,
+      "loss": 0.6884,
+      "step": 12140
+    },
+    {
+      "epoch": 0.6034568391775107,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0007517274262441642,
+      "loss": 0.6822,
+      "step": 12150
+    },
+    {
+      "epoch": 0.60395351147313,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007516876924605145,
+      "loss": 0.6637,
+      "step": 12160
+    },
+    {
+      "epoch": 0.6044501837687494,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.000751647958676865,
+      "loss": 0.6784,
+      "step": 12170
+    },
+    {
+      "epoch": 0.6049468560643687,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0007516082248932156,
+      "loss": 0.6741,
+      "step": 12180
+    },
+    {
+      "epoch": 0.605443528359988,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0007515684911095659,
+      "loss": 0.6529,
+      "step": 12190
+    },
+    {
+      "epoch": 0.6059402006556074,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007515287573259164,
+      "loss": 0.6806,
+      "step": 12200
+    },
+    {
+      "epoch": 0.6064368729512268,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007514890235422669,
+      "loss": 0.6758,
+      "step": 12210
+    },
+    {
+      "epoch": 0.6069335452468462,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0007514492897586172,
+      "loss": 0.7012,
+      "step": 12220
+    },
+    {
+      "epoch": 0.6074302175424655,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007514095559749678,
+      "loss": 0.7264,
+      "step": 12230
+    },
+    {
+      "epoch": 0.6079268898380848,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0007513698221913182,
+      "loss": 0.6513,
+      "step": 12240
+    },
+    {
+      "epoch": 0.6084235621337042,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007513300884076687,
+      "loss": 0.6646,
+      "step": 12250
+    },
+    {
+      "epoch": 0.6089202344293235,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0007512903546240191,
+      "loss": 0.6786,
+      "step": 12260
+    },
+    {
+      "epoch": 0.6094169067249429,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007512506208403695,
+      "loss": 0.6541,
+      "step": 12270
+    },
+    {
+      "epoch": 0.6099135790205622,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007512108870567201,
+      "loss": 0.6807,
+      "step": 12280
+    },
+    {
+      "epoch": 0.6104102513161815,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0007511711532730705,
+      "loss": 0.6737,
+      "step": 12290
+    },
+    {
+      "epoch": 0.6109069236118009,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0007511314194894209,
+      "loss": 0.6765,
+      "step": 12300
+    },
+    {
+      "epoch": 0.6114035959074203,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007510916857057714,
+      "loss": 0.7203,
+      "step": 12310
+    },
+    {
+      "epoch": 0.6119002682030397,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0007510519519221218,
+      "loss": 0.665,
+      "step": 12320
+    },
+    {
+      "epoch": 0.612396940498659,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0007510122181384722,
+      "loss": 0.6752,
+      "step": 12330
+    },
+    {
+      "epoch": 0.6128936127942783,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0007509724843548228,
+      "loss": 0.6806,
+      "step": 12340
+    },
+    {
+      "epoch": 0.6133902850898977,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0007509327505711731,
+      "loss": 0.6697,
+      "step": 12350
+    },
+    {
+      "epoch": 0.613886957385517,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007508930167875236,
+      "loss": 0.6807,
+      "step": 12360
+    },
+    {
+      "epoch": 0.6143836296811364,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007508532830038741,
+      "loss": 0.6599,
+      "step": 12370
+    },
+    {
+      "epoch": 0.6148803019767557,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0007508135492202244,
+      "loss": 0.6782,
+      "step": 12380
+    },
+    {
+      "epoch": 0.615376974272375,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.000750773815436575,
+      "loss": 0.6866,
+      "step": 12390
+    },
+    {
+      "epoch": 0.6158736465679945,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007507340816529255,
+      "loss": 0.6832,
+      "step": 12400
+    },
+    {
+      "epoch": 0.6163703188636138,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007506943478692759,
+      "loss": 0.6741,
+      "step": 12410
+    },
+    {
+      "epoch": 0.6168669911592332,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0007506546140856263,
+      "loss": 0.6638,
+      "step": 12420
+    },
+    {
+      "epoch": 0.6173636634548525,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0007506148803019767,
+      "loss": 0.6734,
+      "step": 12430
+    },
+    {
+      "epoch": 0.6178603357504718,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007505751465183273,
+      "loss": 0.6718,
+      "step": 12440
+    },
+    {
+      "epoch": 0.6183570080460912,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007505354127346777,
+      "loss": 0.7069,
+      "step": 12450
+    },
+    {
+      "epoch": 0.6188536803417105,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007504956789510281,
+      "loss": 0.7089,
+      "step": 12460
+    },
+    {
+      "epoch": 0.6193503526373298,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007504559451673786,
+      "loss": 0.6918,
+      "step": 12470
+    },
+    {
+      "epoch": 0.6198470249329492,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.000750416211383729,
+      "loss": 0.695,
+      "step": 12480
+    },
+    {
+      "epoch": 0.6203436972285686,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0007503764776000795,
+      "loss": 0.6542,
+      "step": 12490
+    },
+    {
+      "epoch": 0.620840369524188,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00075033674381643,
+      "loss": 0.7126,
+      "step": 12500
+    },
+    {
+      "epoch": 0.6213370418198073,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0007502970100327804,
+      "loss": 0.6595,
+      "step": 12510
+    },
+    {
+      "epoch": 0.6218337141154267,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007502572762491308,
+      "loss": 0.6853,
+      "step": 12520
+    },
+    {
+      "epoch": 0.622330386411046,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0007502175424654814,
+      "loss": 0.6536,
+      "step": 12530
+    },
+    {
+      "epoch": 0.6228270587066653,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0007501778086818318,
+      "loss": 0.6682,
+      "step": 12540
+    },
+    {
+      "epoch": 0.6233237310022847,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0007501380748981822,
+      "loss": 0.6557,
+      "step": 12550
+    },
+    {
+      "epoch": 0.623820403297904,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007500983411145327,
+      "loss": 0.6871,
+      "step": 12560
+    },
+    {
+      "epoch": 0.6243170755935233,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0007500586073308831,
+      "loss": 0.6744,
+      "step": 12570
+    },
+    {
+      "epoch": 0.6248137478891428,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007500188735472335,
+      "loss": 0.6408,
+      "step": 12580
+    },
+    {
+      "epoch": 0.6253104201847621,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007499791397635841,
+      "loss": 0.6815,
+      "step": 12590
+    },
+    {
+      "epoch": 0.6258070924803815,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0007499394059799345,
+      "loss": 0.6647,
+      "step": 12600
+    },
+    {
+      "epoch": 0.6263037647760008,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0007498996721962849,
+      "loss": 0.666,
+      "step": 12610
+    },
+    {
+      "epoch": 0.6268004370716201,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007498599384126353,
+      "loss": 0.6886,
+      "step": 12620
+    },
+    {
+      "epoch": 0.6272971093672395,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0007498202046289858,
+      "loss": 0.7055,
+      "step": 12630
+    },
+    {
+      "epoch": 0.6277937816628588,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0007497804708453363,
+      "loss": 0.6681,
+      "step": 12640
+    },
+    {
+      "epoch": 0.6282904539584782,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007497407370616867,
+      "loss": 0.6827,
+      "step": 12650
+    },
+    {
+      "epoch": 0.6287871262540975,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0007497010032780372,
+      "loss": 0.6637,
+      "step": 12660
+    },
+    {
+      "epoch": 0.6292837985497169,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0007496612694943876,
+      "loss": 0.6761,
+      "step": 12670
+    },
+    {
+      "epoch": 0.6297804708453363,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.000749621535710738,
+      "loss": 0.6713,
+      "step": 12680
+    },
+    {
+      "epoch": 0.6302771431409556,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007495818019270886,
+      "loss": 0.6797,
+      "step": 12690
+    },
+    {
+      "epoch": 0.630773815436575,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.000749542068143439,
+      "loss": 0.6841,
+      "step": 12700
+    },
+    {
+      "epoch": 0.6312704877321943,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007495023343597894,
+      "loss": 0.6869,
+      "step": 12710
+    },
+    {
+      "epoch": 0.6317671600278136,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0007494626005761399,
+      "loss": 0.6889,
+      "step": 12720
+    },
+    {
+      "epoch": 0.632263832323433,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007494228667924903,
+      "loss": 0.7024,
+      "step": 12730
+    },
+    {
+      "epoch": 0.6327605046190523,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007493831330088408,
+      "loss": 0.6904,
+      "step": 12740
+    },
+    {
+      "epoch": 0.6332571769146716,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0007493433992251913,
+      "loss": 0.6811,
+      "step": 12750
+    },
+    {
+      "epoch": 0.6337538492102911,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0007493036654415417,
+      "loss": 0.6713,
+      "step": 12760
+    },
+    {
+      "epoch": 0.6342505215059104,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0007492639316578921,
+      "loss": 0.6732,
+      "step": 12770
+    },
+    {
+      "epoch": 0.6347471938015298,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0007492241978742426,
+      "loss": 0.6961,
+      "step": 12780
+    },
+    {
+      "epoch": 0.6352438660971491,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007491844640905931,
+      "loss": 0.672,
+      "step": 12790
+    },
+    {
+      "epoch": 0.6357405383927685,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0007491447303069435,
+      "loss": 0.7121,
+      "step": 12800
+    },
+    {
+      "epoch": 0.6362372106883878,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.000749104996523294,
+      "loss": 0.6957,
+      "step": 12810
+    },
+    {
+      "epoch": 0.6367338829840071,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007490652627396444,
+      "loss": 0.6893,
+      "step": 12820
+    },
+    {
+      "epoch": 0.6372305552796265,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007490255289559948,
+      "loss": 0.6501,
+      "step": 12830
+    },
+    {
+      "epoch": 0.6377272275752458,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007489857951723453,
+      "loss": 0.6912,
+      "step": 12840
+    },
+    {
+      "epoch": 0.6382238998708653,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0007489460613886958,
+      "loss": 0.6945,
+      "step": 12850
+    },
+    {
+      "epoch": 0.6387205721664846,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0007489063276050463,
+      "loss": 0.66,
+      "step": 12860
+    },
+    {
+      "epoch": 0.6392172444621039,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0007488665938213966,
+      "loss": 0.7147,
+      "step": 12870
+    },
+    {
+      "epoch": 0.6397139167577233,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0007488268600377471,
+      "loss": 0.668,
+      "step": 12880
+    },
+    {
+      "epoch": 0.6402105890533426,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007487871262540976,
+      "loss": 0.6798,
+      "step": 12890
+    },
+    {
+      "epoch": 0.6407072613489619,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.000748747392470448,
+      "loss": 0.6702,
+      "step": 12900
+    },
+    {
+      "epoch": 0.6412039336445813,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007487076586867985,
+      "loss": 0.661,
+      "step": 12910
+    },
+    {
+      "epoch": 0.6417006059402006,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0007486679249031489,
+      "loss": 0.6679,
+      "step": 12920
+    },
+    {
+      "epoch": 0.64219727823582,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007486281911194993,
+      "loss": 0.659,
+      "step": 12930
+    },
+    {
+      "epoch": 0.6426939505314394,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0007485884573358499,
+      "loss": 0.6966,
+      "step": 12940
+    },
+    {
+      "epoch": 0.6431906228270587,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0007485487235522003,
+      "loss": 0.7063,
+      "step": 12950
+    },
+    {
+      "epoch": 0.6436872951226781,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007485089897685507,
+      "loss": 0.6778,
+      "step": 12960
+    },
+    {
+      "epoch": 0.6441839674182974,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007484692559849012,
+      "loss": 0.7112,
+      "step": 12970
+    },
+    {
+      "epoch": 0.6446806397139168,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007484295222012516,
+      "loss": 0.7005,
+      "step": 12980
+    },
+    {
+      "epoch": 0.6451773120095361,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007483897884176022,
+      "loss": 0.657,
+      "step": 12990
+    },
+    {
+      "epoch": 0.6456739843051554,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0007483500546339526,
+      "loss": 0.6676,
+      "step": 13000
+    },
+    {
+      "epoch": 0.6461706566007748,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.000748310320850303,
+      "loss": 0.6943,
+      "step": 13010
+    },
+    {
+      "epoch": 0.6466673288963941,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007482705870666535,
+      "loss": 0.6727,
+      "step": 13020
+    },
+    {
+      "epoch": 0.6471640011920136,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0007482308532830038,
+      "loss": 0.6818,
+      "step": 13030
+    },
+    {
+      "epoch": 0.6476606734876329,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0007481911194993544,
+      "loss": 0.6715,
+      "step": 13040
+    },
+    {
+      "epoch": 0.6481573457832522,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0007481513857157049,
+      "loss": 0.6993,
+      "step": 13050
+    },
+    {
+      "epoch": 0.6486540180788716,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007481116519320552,
+      "loss": 0.6996,
+      "step": 13060
+    },
+    {
+      "epoch": 0.6491506903744909,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0007480719181484057,
+      "loss": 0.6714,
+      "step": 13070
+    },
+    {
+      "epoch": 0.6496473626701103,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007480321843647562,
+      "loss": 0.702,
+      "step": 13080
+    },
+    {
+      "epoch": 0.6501440349657296,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007479924505811066,
+      "loss": 0.6975,
+      "step": 13090
+    },
+    {
+      "epoch": 0.6506407072613489,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0007479527167974571,
+      "loss": 0.7115,
+      "step": 13100
+    },
+    {
+      "epoch": 0.6511373795569683,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0007479129830138075,
+      "loss": 0.7013,
+      "step": 13110
+    },
+    {
+      "epoch": 0.6516340518525877,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007478732492301579,
+      "loss": 0.6469,
+      "step": 13120
+    },
+    {
+      "epoch": 0.652130724148207,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0007478335154465084,
+      "loss": 0.7094,
+      "step": 13130
+    },
+    {
+      "epoch": 0.6526273964438264,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0007477937816628589,
+      "loss": 0.6609,
+      "step": 13140
+    },
+    {
+      "epoch": 0.6531240687394457,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007477540478792094,
+      "loss": 0.684,
+      "step": 13150
+    },
+    {
+      "epoch": 0.6536207410350651,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007477143140955598,
+      "loss": 0.6833,
+      "step": 13160
+    },
+    {
+      "epoch": 0.6541174133306844,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007476745803119102,
+      "loss": 0.6783,
+      "step": 13170
+    },
+    {
+      "epoch": 0.6546140856263037,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0007476348465282607,
+      "loss": 0.6597,
+      "step": 13180
+    },
+    {
+      "epoch": 0.6551107579219231,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0007475951127446112,
+      "loss": 0.6951,
+      "step": 13190
+    },
+    {
+      "epoch": 0.6556074302175424,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0007475553789609616,
+      "loss": 0.6596,
+      "step": 13200
+    },
+    {
+      "epoch": 0.6561041025131619,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0007475156451773121,
+      "loss": 0.6579,
+      "step": 13210
+    },
+    {
+      "epoch": 0.6566007748087812,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0007474759113936624,
+      "loss": 0.669,
+      "step": 13220
+    },
+    {
+      "epoch": 0.6570974471044005,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0007474361776100129,
+      "loss": 0.6524,
+      "step": 13230
+    },
+    {
+      "epoch": 0.6575941194000199,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007473964438263635,
+      "loss": 0.6781,
+      "step": 13240
+    },
+    {
+      "epoch": 0.6580907916956392,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007473567100427138,
+      "loss": 0.6868,
+      "step": 13250
+    },
+    {
+      "epoch": 0.6585874639912586,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0007473169762590643,
+      "loss": 0.6969,
+      "step": 13260
+    },
+    {
+      "epoch": 0.6590841362868779,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007472772424754148,
+      "loss": 0.7009,
+      "step": 13270
+    },
+    {
+      "epoch": 0.6595808085824972,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007472375086917652,
+      "loss": 0.681,
+      "step": 13280
+    },
+    {
+      "epoch": 0.6600774808781166,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0007471977749081157,
+      "loss": 0.6342,
+      "step": 13290
+    },
+    {
+      "epoch": 0.660574153173736,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007471580411244661,
+      "loss": 0.6394,
+      "step": 13300
+    },
+    {
+      "epoch": 0.6610708254693554,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007471183073408166,
+      "loss": 0.7037,
+      "step": 13310
+    },
+    {
+      "epoch": 0.6615674977649747,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.000747078573557167,
+      "loss": 0.6478,
+      "step": 13320
+    },
+    {
+      "epoch": 0.662064170060594,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0007470388397735174,
+      "loss": 0.6727,
+      "step": 13330
+    },
+    {
+      "epoch": 0.6625608423562134,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.000746999105989868,
+      "loss": 0.6829,
+      "step": 13340
+    },
+    {
+      "epoch": 0.6630575146518327,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0007469593722062184,
+      "loss": 0.6971,
+      "step": 13350
+    },
+    {
+      "epoch": 0.663554186947452,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007469196384225688,
+      "loss": 0.6622,
+      "step": 13360
+    },
+    {
+      "epoch": 0.6640508592430714,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007468799046389193,
+      "loss": 0.6789,
+      "step": 13370
+    },
+    {
+      "epoch": 0.6645475315386907,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007468401708552697,
+      "loss": 0.6986,
+      "step": 13380
+    },
+    {
+      "epoch": 0.6650442038343102,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0007468004370716202,
+      "loss": 0.7204,
+      "step": 13390
+    },
+    {
+      "epoch": 0.6655408761299295,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007467607032879707,
+      "loss": 0.6522,
+      "step": 13400
+    },
+    {
+      "epoch": 0.6660375484255489,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0007467209695043211,
+      "loss": 0.6546,
+      "step": 13410
+    },
+    {
+      "epoch": 0.6665342207211682,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0007466812357206715,
+      "loss": 0.6535,
+      "step": 13420
+    },
+    {
+      "epoch": 0.6670308930167875,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.000746641501937022,
+      "loss": 0.6729,
+      "step": 13430
+    },
+    {
+      "epoch": 0.6675275653124069,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007466017681533725,
+      "loss": 0.6886,
+      "step": 13440
+    },
+    {
+      "epoch": 0.6680242376080262,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0007465620343697229,
+      "loss": 0.6564,
+      "step": 13450
+    },
+    {
+      "epoch": 0.6685209099036455,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007465223005860734,
+      "loss": 0.6938,
+      "step": 13460
+    },
+    {
+      "epoch": 0.6690175821992649,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0007464825668024238,
+      "loss": 0.6715,
+      "step": 13470
+    },
+    {
+      "epoch": 0.6695142544948842,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0007464428330187742,
+      "loss": 0.6789,
+      "step": 13480
+    },
+    {
+      "epoch": 0.6700109267905037,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0007464030992351247,
+      "loss": 0.6537,
+      "step": 13490
+    },
+    {
+      "epoch": 0.670507599086123,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007463633654514752,
+      "loss": 0.6739,
+      "step": 13500
+    },
+    {
+      "epoch": 0.6710042713817423,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007463236316678256,
+      "loss": 0.6662,
+      "step": 13510
+    },
+    {
+      "epoch": 0.6715009436773617,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.000746283897884176,
+      "loss": 0.6451,
+      "step": 13520
+    },
+    {
+      "epoch": 0.671997615972981,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0007462441641005265,
+      "loss": 0.6754,
+      "step": 13530
+    },
+    {
+      "epoch": 0.6724942882686004,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.000746204430316877,
+      "loss": 0.6632,
+      "step": 13540
+    },
+    {
+      "epoch": 0.6729909605642197,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0007461646965332274,
+      "loss": 0.7027,
+      "step": 13550
+    },
+    {
+      "epoch": 0.673487632859839,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007461249627495779,
+      "loss": 0.6755,
+      "step": 13560
+    },
+    {
+      "epoch": 0.6739843051554584,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007460852289659283,
+      "loss": 0.688,
+      "step": 13570
+    },
+    {
+      "epoch": 0.6744809774510778,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0007460454951822787,
+      "loss": 0.69,
+      "step": 13580
+    },
+    {
+      "epoch": 0.6749776497466972,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007460057613986293,
+      "loss": 0.7026,
+      "step": 13590
+    },
+    {
+      "epoch": 0.6754743220423165,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0007459660276149797,
+      "loss": 0.6592,
+      "step": 13600
+    },
+    {
+      "epoch": 0.6759709943379358,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007459262938313301,
+      "loss": 0.6609,
+      "step": 13610
+    },
+    {
+      "epoch": 0.6764676666335552,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0007458865600476806,
+      "loss": 0.6837,
+      "step": 13620
+    },
+    {
+      "epoch": 0.6769643389291745,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000745846826264031,
+      "loss": 0.6865,
+      "step": 13630
+    },
+    {
+      "epoch": 0.6774610112247939,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0007458070924803814,
+      "loss": 0.6354,
+      "step": 13640
+    },
+    {
+      "epoch": 0.6779576835204132,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.000745767358696732,
+      "loss": 0.7152,
+      "step": 13650
+    },
+    {
+      "epoch": 0.6784543558160325,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0007457276249130824,
+      "loss": 0.6739,
+      "step": 13660
+    },
+    {
+      "epoch": 0.678951028111652,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0007456878911294328,
+      "loss": 0.6947,
+      "step": 13670
+    },
+    {
+      "epoch": 0.6794477004072713,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0007456481573457833,
+      "loss": 0.6693,
+      "step": 13680
+    },
+    {
+      "epoch": 0.6799443727028907,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0007456084235621338,
+      "loss": 0.6785,
+      "step": 13690
+    },
+    {
+      "epoch": 0.68044104499851,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007455686897784842,
+      "loss": 0.6698,
+      "step": 13700
+    },
+    {
+      "epoch": 0.6809377172941293,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0007455289559948346,
+      "loss": 0.6377,
+      "step": 13710
+    },
+    {
+      "epoch": 0.6814343895897487,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0007454892222111851,
+      "loss": 0.6667,
+      "step": 13720
+    },
+    {
+      "epoch": 0.681931061885368,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007454494884275356,
+      "loss": 0.7018,
+      "step": 13730
+    },
+    {
+      "epoch": 0.6824277341809873,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.000745409754643886,
+      "loss": 0.6803,
+      "step": 13740
+    },
+    {
+      "epoch": 0.6829244064766067,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007453700208602365,
+      "loss": 0.6521,
+      "step": 13750
+    },
+    {
+      "epoch": 0.6834210787722261,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007453302870765869,
+      "loss": 0.696,
+      "step": 13760
+    },
+    {
+      "epoch": 0.6839177510678455,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0007452905532929373,
+      "loss": 0.6783,
+      "step": 13770
+    },
+    {
+      "epoch": 0.6844144233634648,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0007452508195092878,
+      "loss": 0.6695,
+      "step": 13780
+    },
+    {
+      "epoch": 0.6849110956590841,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007452110857256382,
+      "loss": 0.694,
+      "step": 13790
+    },
+    {
+      "epoch": 0.6854077679547035,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0007451713519419887,
+      "loss": 0.6664,
+      "step": 13800
+    },
+    {
+      "epoch": 0.6859044402503228,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007451316181583392,
+      "loss": 0.6734,
+      "step": 13810
+    },
+    {
+      "epoch": 0.6864011125459422,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007450918843746896,
+      "loss": 0.6455,
+      "step": 13820
+    },
+    {
+      "epoch": 0.6868977848415615,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00074505215059104,
+      "loss": 0.6509,
+      "step": 13830
+    },
+    {
+      "epoch": 0.6873944571371808,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007450124168073906,
+      "loss": 0.6705,
+      "step": 13840
+    },
+    {
+      "epoch": 0.6878911294328003,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.000744972683023741,
+      "loss": 0.6952,
+      "step": 13850
+    },
+    {
+      "epoch": 0.6883878017284196,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0007449329492400914,
+      "loss": 0.6787,
+      "step": 13860
+    },
+    {
+      "epoch": 0.688884474024039,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007448932154564419,
+      "loss": 0.6715,
+      "step": 13870
+    },
+    {
+      "epoch": 0.6893811463196583,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0007448534816727923,
+      "loss": 0.6722,
+      "step": 13880
+    },
+    {
+      "epoch": 0.6898778186152776,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0007448137478891429,
+      "loss": 0.6656,
+      "step": 13890
+    },
+    {
+      "epoch": 0.690374490910897,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007447740141054932,
+      "loss": 0.6341,
+      "step": 13900
+    },
+    {
+      "epoch": 0.6908711632065163,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0007447342803218437,
+      "loss": 0.687,
+      "step": 13910
+    },
+    {
+      "epoch": 0.6913678355021357,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0007446945465381942,
+      "loss": 0.7099,
+      "step": 13920
+    },
+    {
+      "epoch": 0.691864507797755,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0007446548127545445,
+      "loss": 0.6738,
+      "step": 13930
+    },
+    {
+      "epoch": 0.6923611800933744,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.000744615078970895,
+      "loss": 0.6398,
+      "step": 13940
+    },
+    {
+      "epoch": 0.6928578523889938,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0007445753451872456,
+      "loss": 0.66,
+      "step": 13950
+    },
+    {
+      "epoch": 0.6933545246846131,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0007445356114035959,
+      "loss": 0.6886,
+      "step": 13960
+    },
+    {
+      "epoch": 0.6938511969802325,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007444958776199464,
+      "loss": 0.6649,
+      "step": 13970
+    },
+    {
+      "epoch": 0.6943478692758518,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007444561438362968,
+      "loss": 0.6863,
+      "step": 13980
+    },
+    {
+      "epoch": 0.6948445415714711,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0007444164100526472,
+      "loss": 0.6929,
+      "step": 13990
+    },
+    {
+      "epoch": 0.6953412138670905,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0007443766762689978,
+      "loss": 0.6631,
+      "step": 14000
+    },
+    {
+      "epoch": 0.6958378861627098,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0007443369424853482,
+      "loss": 0.6998,
+      "step": 14010
+    },
+    {
+      "epoch": 0.6963345584583291,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0007442972087016986,
+      "loss": 0.6604,
+      "step": 14020
+    },
+    {
+      "epoch": 0.6968312307539486,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007442574749180491,
+      "loss": 0.6781,
+      "step": 14030
+    },
+    {
+      "epoch": 0.6973279030495679,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007442177411343995,
+      "loss": 0.653,
+      "step": 14040
+    },
+    {
+      "epoch": 0.6978245753451873,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0007441780073507501,
+      "loss": 0.6524,
+      "step": 14050
+    },
+    {
+      "epoch": 0.6983212476408066,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0007441382735671005,
+      "loss": 0.6796,
+      "step": 14060
+    },
+    {
+      "epoch": 0.698817919936426,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007440985397834509,
+      "loss": 0.6614,
+      "step": 14070
+    },
+    {
+      "epoch": 0.6993145922320453,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007440588059998014,
+      "loss": 0.6559,
+      "step": 14080
+    },
+    {
+      "epoch": 0.6998112645276646,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007440190722161517,
+      "loss": 0.6349,
+      "step": 14090
+    },
+    {
+      "epoch": 0.700307936823284,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0007439793384325023,
+      "loss": 0.6741,
+      "step": 14100
+    },
+    {
+      "epoch": 0.7008046091189033,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0007439396046488528,
+      "loss": 0.6455,
+      "step": 14110
+    },
+    {
+      "epoch": 0.7013012814145227,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007438998708652031,
+      "loss": 0.6997,
+      "step": 14120
+    },
+    {
+      "epoch": 0.7017979537101421,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007438601370815536,
+      "loss": 0.661,
+      "step": 14130
+    },
+    {
+      "epoch": 0.7022946260057614,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0007438204032979041,
+      "loss": 0.6776,
+      "step": 14140
+    },
+    {
+      "epoch": 0.7027912983013808,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0007437806695142545,
+      "loss": 0.6656,
+      "step": 14150
+    },
+    {
+      "epoch": 0.7032879705970001,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.000743740935730605,
+      "loss": 0.6608,
+      "step": 14160
+    },
+    {
+      "epoch": 0.7037846428926194,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0007437012019469554,
+      "loss": 0.6501,
+      "step": 14170
+    },
+    {
+      "epoch": 0.7042813151882388,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007436614681633059,
+      "loss": 0.6619,
+      "step": 14180
+    },
+    {
+      "epoch": 0.7047779874838581,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0007436217343796563,
+      "loss": 0.6832,
+      "step": 14190
+    },
+    {
+      "epoch": 0.7052746597794775,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007435820005960068,
+      "loss": 0.66,
+      "step": 14200
+    },
+    {
+      "epoch": 0.7057713320750969,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0007435422668123573,
+      "loss": 0.6392,
+      "step": 14210
+    },
+    {
+      "epoch": 0.7062680043707162,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0007435025330287077,
+      "loss": 0.6357,
+      "step": 14220
+    },
+    {
+      "epoch": 0.7067646766663356,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0007434627992450581,
+      "loss": 0.6774,
+      "step": 14230
+    },
+    {
+      "epoch": 0.7072613489619549,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0007434230654614086,
+      "loss": 0.63,
+      "step": 14240
+    },
+    {
+      "epoch": 0.7077580212575743,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0007433833316777591,
+      "loss": 0.6643,
+      "step": 14250
+    },
+    {
+      "epoch": 0.7082546935531936,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0007433435978941095,
+      "loss": 0.6845,
+      "step": 14260
+    },
+    {
+      "epoch": 0.7087513658488129,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00074330386411046,
+      "loss": 0.6639,
+      "step": 14270
+    },
+    {
+      "epoch": 0.7092480381444323,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007432641303268104,
+      "loss": 0.7032,
+      "step": 14280
+    },
+    {
+      "epoch": 0.7097447104400516,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007432243965431608,
+      "loss": 0.6655,
+      "step": 14290
+    },
+    {
+      "epoch": 0.7102413827356711,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0007431846627595114,
+      "loss": 0.6785,
+      "step": 14300
+    },
+    {
+      "epoch": 0.7107380550312904,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007431449289758617,
+      "loss": 0.673,
+      "step": 14310
+    },
+    {
+      "epoch": 0.7112347273269097,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0007431051951922122,
+      "loss": 0.6534,
+      "step": 14320
+    },
+    {
+      "epoch": 0.7117313996225291,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007430654614085627,
+      "loss": 0.6963,
+      "step": 14330
+    },
+    {
+      "epoch": 0.7122280719181484,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007430257276249131,
+      "loss": 0.6584,
+      "step": 14340
+    },
+    {
+      "epoch": 0.7127247442137677,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007429859938412636,
+      "loss": 0.6742,
+      "step": 14350
+    },
+    {
+      "epoch": 0.7132214165093871,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.000742946260057614,
+      "loss": 0.6555,
+      "step": 14360
+    },
+    {
+      "epoch": 0.7137180888050064,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0007429065262739645,
+      "loss": 0.6705,
+      "step": 14370
+    },
+    {
+      "epoch": 0.7142147611006258,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0007428667924903149,
+      "loss": 0.661,
+      "step": 14380
+    },
+    {
+      "epoch": 0.7147114333962452,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007428270587066653,
+      "loss": 0.6886,
+      "step": 14390
+    },
+    {
+      "epoch": 0.7152081056918645,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0007427873249230159,
+      "loss": 0.6865,
+      "step": 14400
+    },
+    {
+      "epoch": 0.7157047779874839,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0007427475911393663,
+      "loss": 0.6566,
+      "step": 14410
+    },
+    {
+      "epoch": 0.7162014502831032,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0007427078573557167,
+      "loss": 0.6595,
+      "step": 14420
+    },
+    {
+      "epoch": 0.7166981225787226,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0007426681235720672,
+      "loss": 0.6846,
+      "step": 14430
+    },
+    {
+      "epoch": 0.7171947948743419,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007426283897884176,
+      "loss": 0.6815,
+      "step": 14440
+    },
+    {
+      "epoch": 0.7176914671699612,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007425886560047681,
+      "loss": 0.6479,
+      "step": 14450
+    },
+    {
+      "epoch": 0.7181881394655806,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0007425489222211186,
+      "loss": 0.6771,
+      "step": 14460
+    },
+    {
+      "epoch": 0.7186848117611999,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.000742509188437469,
+      "loss": 0.6768,
+      "step": 14470
+    },
+    {
+      "epoch": 0.7191814840568194,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0007424694546538194,
+      "loss": 0.6603,
+      "step": 14480
+    },
+    {
+      "epoch": 0.7196781563524387,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007424297208701699,
+      "loss": 0.6645,
+      "step": 14490
+    },
+    {
+      "epoch": 0.720174828648058,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007423899870865204,
+      "loss": 0.6885,
+      "step": 14500
+    },
+    {
+      "epoch": 0.7206715009436774,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0007423502533028708,
+      "loss": 0.6573,
+      "step": 14510
+    },
+    {
+      "epoch": 0.7211681732392967,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007423105195192213,
+      "loss": 0.6914,
+      "step": 14520
+    },
+    {
+      "epoch": 0.721664845534916,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007422707857355717,
+      "loss": 0.6392,
+      "step": 14530
+    },
+    {
+      "epoch": 0.7221615178305354,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007422310519519221,
+      "loss": 0.6606,
+      "step": 14540
+    },
+    {
+      "epoch": 0.7226581901261547,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007421913181682727,
+      "loss": 0.7015,
+      "step": 14550
+    },
+    {
+      "epoch": 0.7231548624217741,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007421515843846231,
+      "loss": 0.6693,
+      "step": 14560
+    },
+    {
+      "epoch": 0.7236515347173935,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007421118506009735,
+      "loss": 0.6702,
+      "step": 14570
+    },
+    {
+      "epoch": 0.7241482070130129,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007420721168173239,
+      "loss": 0.7158,
+      "step": 14580
+    },
+    {
+      "epoch": 0.7246448793086322,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007420323830336744,
+      "loss": 0.6439,
+      "step": 14590
+    },
+    {
+      "epoch": 0.7251415516042515,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007419926492500249,
+      "loss": 0.6677,
+      "step": 14600
+    },
+    {
+      "epoch": 0.7256382238998709,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0007419529154663753,
+      "loss": 0.6892,
+      "step": 14610
+    },
+    {
+      "epoch": 0.7261348961954902,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007419131816827258,
+      "loss": 0.6827,
+      "step": 14620
+    },
+    {
+      "epoch": 0.7266315684911095,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0007418734478990762,
+      "loss": 0.6628,
+      "step": 14630
+    },
+    {
+      "epoch": 0.7271282407867289,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007418337141154266,
+      "loss": 0.6387,
+      "step": 14640
+    },
+    {
+      "epoch": 0.7276249130823482,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0007417939803317772,
+      "loss": 0.6581,
+      "step": 14650
+    },
+    {
+      "epoch": 0.7281215853779676,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0007417542465481276,
+      "loss": 0.6713,
+      "step": 14660
+    },
+    {
+      "epoch": 0.728618257673587,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.000741714512764478,
+      "loss": 0.6754,
+      "step": 14670
+    },
+    {
+      "epoch": 0.7291149299692063,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007416747789808285,
+      "loss": 0.6784,
+      "step": 14680
+    },
+    {
+      "epoch": 0.7296116022648257,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0007416350451971789,
+      "loss": 0.6548,
+      "step": 14690
+    },
+    {
+      "epoch": 0.730108274560445,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007415953114135294,
+      "loss": 0.6828,
+      "step": 14700
+    },
+    {
+      "epoch": 0.7306049468560644,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0007415555776298799,
+      "loss": 0.6802,
+      "step": 14710
+    },
+    {
+      "epoch": 0.7311016191516837,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007415158438462303,
+      "loss": 0.657,
+      "step": 14720
+    },
+    {
+      "epoch": 0.731598291447303,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007414761100625807,
+      "loss": 0.6715,
+      "step": 14730
+    },
+    {
+      "epoch": 0.7320949637429224,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007414363762789312,
+      "loss": 0.633,
+      "step": 14740
+    },
+    {
+      "epoch": 0.7325916360385417,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0007413966424952817,
+      "loss": 0.643,
+      "step": 14750
+    },
+    {
+      "epoch": 0.7330883083341612,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0007413569087116321,
+      "loss": 0.6803,
+      "step": 14760
+    },
+    {
+      "epoch": 0.7335849806297805,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0007413171749279825,
+      "loss": 0.6491,
+      "step": 14770
+    },
+    {
+      "epoch": 0.7340816529253998,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.000741277441144333,
+      "loss": 0.6822,
+      "step": 14780
+    },
+    {
+      "epoch": 0.7345783252210192,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0007412377073606835,
+      "loss": 0.6758,
+      "step": 14790
+    },
+    {
+      "epoch": 0.7350749975166385,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0007411979735770338,
+      "loss": 0.6977,
+      "step": 14800
+    },
+    {
+      "epoch": 0.7355716698122579,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0007411582397933844,
+      "loss": 0.6655,
+      "step": 14810
+    },
+    {
+      "epoch": 0.7360683421078772,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007411185060097349,
+      "loss": 0.6554,
+      "step": 14820
+    },
+    {
+      "epoch": 0.7365650144034965,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0007410787722260852,
+      "loss": 0.623,
+      "step": 14830
+    },
+    {
+      "epoch": 0.7370616866991159,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007410390384424357,
+      "loss": 0.6476,
+      "step": 14840
+    },
+    {
+      "epoch": 0.7375583589947353,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0007409993046587862,
+      "loss": 0.6634,
+      "step": 14850
+    },
+    {
+      "epoch": 0.7380550312903547,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007409595708751366,
+      "loss": 0.6542,
+      "step": 14860
+    },
+    {
+      "epoch": 0.738551703585974,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0007409198370914871,
+      "loss": 0.6161,
+      "step": 14870
+    },
+    {
+      "epoch": 0.7390483758815933,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007408801033078375,
+      "loss": 0.6548,
+      "step": 14880
+    },
+    {
+      "epoch": 0.7395450481772127,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0007408403695241879,
+      "loss": 0.6417,
+      "step": 14890
+    },
+    {
+      "epoch": 0.740041720472832,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0007408006357405385,
+      "loss": 0.6701,
+      "step": 14900
+    },
+    {
+      "epoch": 0.7405383927684513,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007407609019568889,
+      "loss": 0.6672,
+      "step": 14910
+    },
+    {
+      "epoch": 0.7410350650640707,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0007407211681732393,
+      "loss": 0.6359,
+      "step": 14920
+    },
+    {
+      "epoch": 0.74153173735969,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007406814343895898,
+      "loss": 0.668,
+      "step": 14930
+    },
+    {
+      "epoch": 0.7420284096553095,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007406417006059402,
+      "loss": 0.6952,
+      "step": 14940
+    },
+    {
+      "epoch": 0.7425250819509288,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0007406019668222908,
+      "loss": 0.6725,
+      "step": 14950
+    },
+    {
+      "epoch": 0.7430217542465481,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0007405622330386411,
+      "loss": 0.6643,
+      "step": 14960
+    },
+    {
+      "epoch": 0.7435184265421675,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0007405224992549916,
+      "loss": 0.6796,
+      "step": 14970
+    },
+    {
+      "epoch": 0.7440150988377868,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007404827654713421,
+      "loss": 0.6462,
+      "step": 14980
+    },
+    {
+      "epoch": 0.7445117711334062,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0007404430316876924,
+      "loss": 0.6689,
+      "step": 14990
+    },
+    {
+      "epoch": 0.7450084434290255,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.000740403297904043,
+      "loss": 0.7093,
+      "step": 15000
+    },
+    {
+      "epoch": 0.7455051157246448,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0007403635641203935,
+      "loss": 0.6595,
+      "step": 15010
+    },
+    {
+      "epoch": 0.7460017880202642,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007403238303367438,
+      "loss": 0.7084,
+      "step": 15020
+    },
+    {
+      "epoch": 0.7464984603158836,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0007402840965530943,
+      "loss": 0.6587,
+      "step": 15030
+    },
+    {
+      "epoch": 0.746995132611503,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0007402443627694447,
+      "loss": 0.6821,
+      "step": 15040
+    },
+    {
+      "epoch": 0.7474918049071223,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007402046289857951,
+      "loss": 0.6574,
+      "step": 15050
+    },
+    {
+      "epoch": 0.7479884772027416,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0007401648952021457,
+      "loss": 0.6512,
+      "step": 15060
+    },
+    {
+      "epoch": 0.748485149498361,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0007401251614184961,
+      "loss": 0.6612,
+      "step": 15070
+    },
+    {
+      "epoch": 0.7489818217939803,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0007400854276348466,
+      "loss": 0.6923,
+      "step": 15080
+    },
+    {
+      "epoch": 0.7494784940895997,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.000740045693851197,
+      "loss": 0.6737,
+      "step": 15090
+    },
+    {
+      "epoch": 0.749975166385219,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0007400059600675474,
+      "loss": 0.6633,
+      "step": 15100
+    },
+    {
+      "epoch": 0.7504718386808383,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.000739966226283898,
+      "loss": 0.6599,
+      "step": 15110
+    },
+    {
+      "epoch": 0.7509685109764578,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007399264925002484,
+      "loss": 0.6532,
+      "step": 15120
+    },
+    {
+      "epoch": 0.7514651832720771,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007398867587165988,
+      "loss": 0.6475,
+      "step": 15130
+    },
+    {
+      "epoch": 0.7519618555676965,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0007398470249329493,
+      "loss": 0.627,
+      "step": 15140
+    },
+    {
+      "epoch": 0.7524585278633158,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0007398072911492998,
+      "loss": 0.6591,
+      "step": 15150
+    },
+    {
+      "epoch": 0.7529552001589351,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0007397675573656502,
+      "loss": 0.678,
+      "step": 15160
+    },
+    {
+      "epoch": 0.7534518724545545,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007397278235820007,
+      "loss": 0.6714,
+      "step": 15170
+    },
+    {
+      "epoch": 0.7539485447501738,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.000739688089798351,
+      "loss": 0.6759,
+      "step": 15180
+    },
+    {
+      "epoch": 0.7544452170457931,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0007396483560147015,
+      "loss": 0.7192,
+      "step": 15190
+    },
+    {
+      "epoch": 0.7549418893414125,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.000739608622231052,
+      "loss": 0.6535,
+      "step": 15200
+    },
+    {
+      "epoch": 0.7554385616370319,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007395688884474024,
+      "loss": 0.6535,
+      "step": 15210
+    },
+    {
+      "epoch": 0.7559352339326513,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0007395291546637529,
+      "loss": 0.6493,
+      "step": 15220
+    },
+    {
+      "epoch": 0.7564319062282706,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0007394894208801033,
+      "loss": 0.6432,
+      "step": 15230
+    },
+    {
+      "epoch": 0.75692857852389,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0007394496870964538,
+      "loss": 0.6387,
+      "step": 15240
+    },
+    {
+      "epoch": 0.7574252508195093,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0007394099533128042,
+      "loss": 0.6892,
+      "step": 15250
+    },
+    {
+      "epoch": 0.7579219231151286,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0007393702195291547,
+      "loss": 0.6835,
+      "step": 15260
+    },
+    {
+      "epoch": 0.758418595410748,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007393304857455052,
+      "loss": 0.6625,
+      "step": 15270
+    },
+    {
+      "epoch": 0.7589152677063673,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0007392907519618556,
+      "loss": 0.687,
+      "step": 15280
+    },
+    {
+      "epoch": 0.7594119400019866,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.000739251018178206,
+      "loss": 0.63,
+      "step": 15290
+    },
+    {
+      "epoch": 0.7599086122976061,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0007392112843945566,
+      "loss": 0.6379,
+      "step": 15300
+    },
+    {
+      "epoch": 0.7604052845932254,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.000739171550610907,
+      "loss": 0.6866,
+      "step": 15310
+    },
+    {
+      "epoch": 0.7609019568888448,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007391318168272574,
+      "loss": 0.651,
+      "step": 15320
+    },
+    {
+      "epoch": 0.7613986291844641,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007390920830436079,
+      "loss": 0.6841,
+      "step": 15330
+    },
+    {
+      "epoch": 0.7618953014800834,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0007390523492599583,
+      "loss": 0.6586,
+      "step": 15340
+    },
+    {
+      "epoch": 0.7623919737757028,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007390126154763087,
+      "loss": 0.6854,
+      "step": 15350
+    },
+    {
+      "epoch": 0.7628886460713221,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0007389728816926593,
+      "loss": 0.668,
+      "step": 15360
+    },
+    {
+      "epoch": 0.7633853183669415,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0007389331479090096,
+      "loss": 0.6417,
+      "step": 15370
+    },
+    {
+      "epoch": 0.7638819906625608,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0007388934141253601,
+      "loss": 0.6731,
+      "step": 15380
+    },
+    {
+      "epoch": 0.7643786629581802,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007388536803417106,
+      "loss": 0.6381,
+      "step": 15390
+    },
+    {
+      "epoch": 0.7648753352537996,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.000738813946558061,
+      "loss": 0.6457,
+      "step": 15400
+    },
+    {
+      "epoch": 0.7653720075494189,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0007387742127744115,
+      "loss": 0.6486,
+      "step": 15410
+    },
+    {
+      "epoch": 0.7658686798450383,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.000738734478990762,
+      "loss": 0.6659,
+      "step": 15420
+    },
+    {
+      "epoch": 0.7663653521406576,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007386947452071124,
+      "loss": 0.6527,
+      "step": 15430
+    },
+    {
+      "epoch": 0.7668620244362769,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0007386550114234628,
+      "loss": 0.6608,
+      "step": 15440
+    },
+    {
+      "epoch": 0.7673586967318963,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0007386152776398132,
+      "loss": 0.6618,
+      "step": 15450
+    },
+    {
+      "epoch": 0.7678553690275156,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007385755438561638,
+      "loss": 0.6524,
+      "step": 15460
+    },
+    {
+      "epoch": 0.768352041323135,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0007385358100725142,
+      "loss": 0.6715,
+      "step": 15470
+    },
+    {
+      "epoch": 0.7688487136187544,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007384960762888646,
+      "loss": 0.6431,
+      "step": 15480
+    },
+    {
+      "epoch": 0.7693453859143737,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007384563425052151,
+      "loss": 0.6363,
+      "step": 15490
+    },
+    {
+      "epoch": 0.7698420582099931,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0007384166087215655,
+      "loss": 0.6678,
+      "step": 15500
+    },
+    {
+      "epoch": 0.7703387305056124,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.000738376874937916,
+      "loss": 0.6862,
+      "step": 15510
+    },
+    {
+      "epoch": 0.7708354028012318,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0007383371411542665,
+      "loss": 0.6488,
+      "step": 15520
+    },
+    {
+      "epoch": 0.7713320750968511,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007382974073706169,
+      "loss": 0.6608,
+      "step": 15530
+    },
+    {
+      "epoch": 0.7718287473924704,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0007382576735869673,
+      "loss": 0.6887,
+      "step": 15540
+    },
+    {
+      "epoch": 0.7723254196880898,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007382179398033178,
+      "loss": 0.65,
+      "step": 15550
+    },
+    {
+      "epoch": 0.7728220919837091,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0007381782060196683,
+      "loss": 0.6542,
+      "step": 15560
+    },
+    {
+      "epoch": 0.7733187642793286,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0007381384722360187,
+      "loss": 0.6324,
+      "step": 15570
+    },
+    {
+      "epoch": 0.7738154365749479,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007380987384523692,
+      "loss": 0.6682,
+      "step": 15580
+    },
+    {
+      "epoch": 0.7743121088705672,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007380590046687196,
+      "loss": 0.6871,
+      "step": 15590
+    },
+    {
+      "epoch": 0.7748087811661866,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00073801927088507,
+      "loss": 0.6528,
+      "step": 15600
+    },
+    {
+      "epoch": 0.7753054534618059,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0007379795371014206,
+      "loss": 0.6642,
+      "step": 15610
+    },
+    {
+      "epoch": 0.7758021257574252,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.000737939803317771,
+      "loss": 0.657,
+      "step": 15620
+    },
+    {
+      "epoch": 0.7762987980530446,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0007379000695341214,
+      "loss": 0.672,
+      "step": 15630
+    },
+    {
+      "epoch": 0.7767954703486639,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007378603357504718,
+      "loss": 0.6719,
+      "step": 15640
+    },
+    {
+      "epoch": 0.7772921426442833,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0007378206019668223,
+      "loss": 0.6735,
+      "step": 15650
+    },
+    {
+      "epoch": 0.7777888149399027,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007377808681831728,
+      "loss": 0.6648,
+      "step": 15660
+    },
+    {
+      "epoch": 0.778285487235522,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0007377411343995232,
+      "loss": 0.6676,
+      "step": 15670
+    },
+    {
+      "epoch": 0.7787821595311414,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007377014006158737,
+      "loss": 0.685,
+      "step": 15680
+    },
+    {
+      "epoch": 0.7792788318267607,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0007376616668322241,
+      "loss": 0.6626,
+      "step": 15690
+    },
+    {
+      "epoch": 0.7797755041223801,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0007376219330485745,
+      "loss": 0.6375,
+      "step": 15700
+    },
+    {
+      "epoch": 0.7802721764179994,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0007375821992649251,
+      "loss": 0.6498,
+      "step": 15710
+    },
+    {
+      "epoch": 0.7807688487136187,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0007375424654812755,
+      "loss": 0.6627,
+      "step": 15720
+    },
+    {
+      "epoch": 0.7812655210092381,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0007375027316976259,
+      "loss": 0.6476,
+      "step": 15730
+    },
+    {
+      "epoch": 0.7817621933048574,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0007374629979139764,
+      "loss": 0.6621,
+      "step": 15740
+    },
+    {
+      "epoch": 0.7822588656004769,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0007374232641303268,
+      "loss": 0.6777,
+      "step": 15750
+    },
+    {
+      "epoch": 0.7827555378960962,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007373835303466773,
+      "loss": 0.6634,
+      "step": 15760
+    },
+    {
+      "epoch": 0.7832522101917155,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0007373437965630278,
+      "loss": 0.6365,
+      "step": 15770
+    },
+    {
+      "epoch": 0.7837488824873349,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007373040627793782,
+      "loss": 0.6644,
+      "step": 15780
+    },
+    {
+      "epoch": 0.7842455547829542,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0007372643289957286,
+      "loss": 0.6768,
+      "step": 15790
+    },
+    {
+      "epoch": 0.7847422270785736,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007372245952120791,
+      "loss": 0.6666,
+      "step": 15800
+    },
+    {
+      "epoch": 0.7852388993741929,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007371848614284296,
+      "loss": 0.6616,
+      "step": 15810
+    },
+    {
+      "epoch": 0.7857355716698122,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0007371451276447801,
+      "loss": 0.6727,
+      "step": 15820
+    },
+    {
+      "epoch": 0.7862322439654316,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007371053938611304,
+      "loss": 0.6297,
+      "step": 15830
+    },
+    {
+      "epoch": 0.7867289162610509,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0007370656600774809,
+      "loss": 0.661,
+      "step": 15840
+    },
+    {
+      "epoch": 0.7872255885566704,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0007370259262938314,
+      "loss": 0.6471,
+      "step": 15850
+    },
+    {
+      "epoch": 0.7877222608522897,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0007369861925101818,
+      "loss": 0.6412,
+      "step": 15860
+    },
+    {
+      "epoch": 0.788218933147909,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0007369464587265323,
+      "loss": 0.647,
+      "step": 15870
+    },
+    {
+      "epoch": 0.7887156054435284,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0007369067249428828,
+      "loss": 0.6365,
+      "step": 15880
+    },
+    {
+      "epoch": 0.7892122777391477,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0007368669911592331,
+      "loss": 0.6588,
+      "step": 15890
+    },
+    {
+      "epoch": 0.789708950034767,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0007368272573755836,
+      "loss": 0.6499,
+      "step": 15900
+    },
+    {
+      "epoch": 0.7902056223303864,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007367875235919341,
+      "loss": 0.6606,
+      "step": 15910
+    },
+    {
+      "epoch": 0.7907022946260057,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007367477898082845,
+      "loss": 0.6713,
+      "step": 15920
+    },
+    {
+      "epoch": 0.7911989669216251,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.000736708056024635,
+      "loss": 0.633,
+      "step": 15930
+    },
+    {
+      "epoch": 0.7916956392172445,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007366683222409854,
+      "loss": 0.6714,
+      "step": 15940
+    },
+    {
+      "epoch": 0.7921923115128638,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007366285884573358,
+      "loss": 0.6649,
+      "step": 15950
+    },
+    {
+      "epoch": 0.7926889838084832,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007365888546736864,
+      "loss": 0.68,
+      "step": 15960
+    },
+    {
+      "epoch": 0.7931856561041025,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007365491208900368,
+      "loss": 0.6256,
+      "step": 15970
+    },
+    {
+      "epoch": 0.7936823283997219,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007365093871063873,
+      "loss": 0.6611,
+      "step": 15980
+    },
+    {
+      "epoch": 0.7941790006953412,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007364696533227377,
+      "loss": 0.6605,
+      "step": 15990
+    },
+    {
+      "epoch": 0.7946756729909605,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007364299195390881,
+      "loss": 0.6494,
+      "step": 16000
+    },
+    {
+      "epoch": 0.7951723452865799,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0007363901857554387,
+      "loss": 0.6674,
+      "step": 16010
+    },
+    {
+      "epoch": 0.7956690175821992,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.000736350451971789,
+      "loss": 0.6483,
+      "step": 16020
+    },
+    {
+      "epoch": 0.7961656898778187,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0007363107181881395,
+      "loss": 0.6441,
+      "step": 16030
+    },
+    {
+      "epoch": 0.796662362173438,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00073627098440449,
+      "loss": 0.6642,
+      "step": 16040
+    },
+    {
+      "epoch": 0.7971590344690573,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0007362312506208403,
+      "loss": 0.6259,
+      "step": 16050
+    },
+    {
+      "epoch": 0.7976557067646767,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0007361915168371909,
+      "loss": 0.6641,
+      "step": 16060
+    },
+    {
+      "epoch": 0.798152379060296,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007361517830535414,
+      "loss": 0.6431,
+      "step": 16070
+    },
+    {
+      "epoch": 0.7986490513559154,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0007361120492698917,
+      "loss": 0.6418,
+      "step": 16080
+    },
+    {
+      "epoch": 0.7991457236515347,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007360723154862422,
+      "loss": 0.6362,
+      "step": 16090
+    },
+    {
+      "epoch": 0.799642395947154,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007360325817025926,
+      "loss": 0.6457,
+      "step": 16100
+    },
+    {
+      "epoch": 0.8001390682427734,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.000735992847918943,
+      "loss": 0.6353,
+      "step": 16110
+    },
+    {
+      "epoch": 0.8006357405383928,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007359531141352936,
+      "loss": 0.6309,
+      "step": 16120
+    },
+    {
+      "epoch": 0.8011324128340122,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.000735913380351644,
+      "loss": 0.6537,
+      "step": 16130
+    },
+    {
+      "epoch": 0.8016290851296315,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0007358736465679945,
+      "loss": 0.6506,
+      "step": 16140
+    },
+    {
+      "epoch": 0.8021257574252508,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007358339127843449,
+      "loss": 0.6716,
+      "step": 16150
+    },
+    {
+      "epoch": 0.8026224297208702,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0007357941790006954,
+      "loss": 0.6404,
+      "step": 16160
+    },
+    {
+      "epoch": 0.8031191020164895,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007357544452170459,
+      "loss": 0.6493,
+      "step": 16170
+    },
+    {
+      "epoch": 0.8036157743121088,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0007357147114333963,
+      "loss": 0.6635,
+      "step": 16180
+    },
+    {
+      "epoch": 0.8041124466077282,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007356749776497467,
+      "loss": 0.619,
+      "step": 16190
+    },
+    {
+      "epoch": 0.8046091189033475,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007356352438660972,
+      "loss": 0.6638,
+      "step": 16200
+    },
+    {
+      "epoch": 0.805105791198967,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007355955100824477,
+      "loss": 0.665,
+      "step": 16210
+    },
+    {
+      "epoch": 0.8056024634945863,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0007355557762987981,
+      "loss": 0.6213,
+      "step": 16220
+    },
+    {
+      "epoch": 0.8060991357902056,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0007355160425151486,
+      "loss": 0.6451,
+      "step": 16230
+    },
+    {
+      "epoch": 0.806595808085825,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0007354763087314989,
+      "loss": 0.6398,
+      "step": 16240
+    },
+    {
+      "epoch": 0.8070924803814443,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0007354365749478494,
+      "loss": 0.6553,
+      "step": 16250
+    },
+    {
+      "epoch": 0.8075891526770637,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007353968411642,
+      "loss": 0.691,
+      "step": 16260
+    },
+    {
+      "epoch": 0.808085824972683,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007353571073805504,
+      "loss": 0.6825,
+      "step": 16270
+    },
+    {
+      "epoch": 0.8085824972683023,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007353173735969008,
+      "loss": 0.6685,
+      "step": 16280
+    },
+    {
+      "epoch": 0.8090791695639217,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0007352776398132512,
+      "loss": 0.6356,
+      "step": 16290
+    },
+    {
+      "epoch": 0.8095758418595411,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0007352379060296017,
+      "loss": 0.6552,
+      "step": 16300
+    },
+    {
+      "epoch": 0.8100725141551605,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007351981722459522,
+      "loss": 0.6721,
+      "step": 16310
+    },
+    {
+      "epoch": 0.8105691864507798,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007351584384623026,
+      "loss": 0.6464,
+      "step": 16320
+    },
+    {
+      "epoch": 0.8110658587463991,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0007351187046786531,
+      "loss": 0.6584,
+      "step": 16330
+    },
+    {
+      "epoch": 0.8115625310420185,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007350789708950035,
+      "loss": 0.6583,
+      "step": 16340
+    },
+    {
+      "epoch": 0.8120592033376378,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0007350392371113539,
+      "loss": 0.6785,
+      "step": 16350
+    },
+    {
+      "epoch": 0.8125558756332572,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007349995033277045,
+      "loss": 0.6511,
+      "step": 16360
+    },
+    {
+      "epoch": 0.8130525479288765,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0007349597695440549,
+      "loss": 0.6634,
+      "step": 16370
+    },
+    {
+      "epoch": 0.8135492202244958,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007349200357604053,
+      "loss": 0.7017,
+      "step": 16380
+    },
+    {
+      "epoch": 0.8140458925201153,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0007348803019767558,
+      "loss": 0.6716,
+      "step": 16390
+    },
+    {
+      "epoch": 0.8145425648157346,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0007348405681931062,
+      "loss": 0.6769,
+      "step": 16400
+    },
+    {
+      "epoch": 0.815039237111354,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0007348008344094566,
+      "loss": 0.6457,
+      "step": 16410
+    },
+    {
+      "epoch": 0.8155359094069733,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0007347611006258072,
+      "loss": 0.6589,
+      "step": 16420
+    },
+    {
+      "epoch": 0.8160325817025926,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0007347213668421576,
+      "loss": 0.6282,
+      "step": 16430
+    },
+    {
+      "epoch": 0.816529253998212,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.000734681633058508,
+      "loss": 0.6914,
+      "step": 16440
+    },
+    {
+      "epoch": 0.8170259262938313,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0007346418992748585,
+      "loss": 0.6693,
+      "step": 16450
+    },
+    {
+      "epoch": 0.8175225985894506,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.000734602165491209,
+      "loss": 0.6435,
+      "step": 16460
+    },
+    {
+      "epoch": 0.81801927088507,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0007345624317075594,
+      "loss": 0.6676,
+      "step": 16470
+    },
+    {
+      "epoch": 0.8185159431806894,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007345226979239099,
+      "loss": 0.6642,
+      "step": 16480
+    },
+    {
+      "epoch": 0.8190126154763088,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0007344829641402603,
+      "loss": 0.677,
+      "step": 16490
+    },
+    {
+      "epoch": 0.8195092877719281,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0007344432303566107,
+      "loss": 0.6594,
+      "step": 16500
+    },
+    {
+      "epoch": 0.8200059600675474,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007344034965729611,
+      "loss": 0.6555,
+      "step": 16510
+    },
+    {
+      "epoch": 0.8205026323631668,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0007343637627893117,
+      "loss": 0.6422,
+      "step": 16520
+    },
+    {
+      "epoch": 0.8209993046587861,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007343240290056621,
+      "loss": 0.6481,
+      "step": 16530
+    },
+    {
+      "epoch": 0.8214959769544055,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007342842952220125,
+      "loss": 0.678,
+      "step": 16540
+    },
+    {
+      "epoch": 0.8219926492500248,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.000734244561438363,
+      "loss": 0.638,
+      "step": 16550
+    },
+    {
+      "epoch": 0.8224893215456441,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0007342048276547134,
+      "loss": 0.6303,
+      "step": 16560
+    },
+    {
+      "epoch": 0.8229859938412636,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0007341650938710639,
+      "loss": 0.6615,
+      "step": 16570
+    },
+    {
+      "epoch": 0.8234826661368829,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0007341253600874144,
+      "loss": 0.6814,
+      "step": 16580
+    },
+    {
+      "epoch": 0.8239793384325023,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007340856263037648,
+      "loss": 0.6632,
+      "step": 16590
+    },
+    {
+      "epoch": 0.8244760107281216,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007340458925201152,
+      "loss": 0.6387,
+      "step": 16600
+    },
+    {
+      "epoch": 0.8249726830237409,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0007340061587364658,
+      "loss": 0.6394,
+      "step": 16610
+    },
+    {
+      "epoch": 0.8254693553193603,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007339664249528162,
+      "loss": 0.6571,
+      "step": 16620
+    },
+    {
+      "epoch": 0.8259660276149796,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007339266911691666,
+      "loss": 0.662,
+      "step": 16630
+    },
+    {
+      "epoch": 0.826462699910599,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007338869573855171,
+      "loss": 0.6662,
+      "step": 16640
+    },
+    {
+      "epoch": 0.8269593722062183,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007338472236018675,
+      "loss": 0.6464,
+      "step": 16650
+    },
+    {
+      "epoch": 0.8274560445018377,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007338074898182179,
+      "loss": 0.6808,
+      "step": 16660
+    },
+    {
+      "epoch": 0.8279527167974571,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007337677560345685,
+      "loss": 0.6563,
+      "step": 16670
+    },
+    {
+      "epoch": 0.8284493890930764,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007337280222509189,
+      "loss": 0.6661,
+      "step": 16680
+    },
+    {
+      "epoch": 0.8289460613886958,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007336882884672693,
+      "loss": 0.6458,
+      "step": 16690
+    },
+    {
+      "epoch": 0.8294427336843151,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007336485546836197,
+      "loss": 0.6569,
+      "step": 16700
+    },
+    {
+      "epoch": 0.8299394059799344,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007336088208999702,
+      "loss": 0.6661,
+      "step": 16710
+    },
+    {
+      "epoch": 0.8304360782755538,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0007335690871163208,
+      "loss": 0.6804,
+      "step": 16720
+    },
+    {
+      "epoch": 0.8309327505711731,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007335293533326711,
+      "loss": 0.6887,
+      "step": 16730
+    },
+    {
+      "epoch": 0.8314294228667924,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007334896195490216,
+      "loss": 0.6213,
+      "step": 16740
+    },
+    {
+      "epoch": 0.8319260951624119,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0007334498857653721,
+      "loss": 0.649,
+      "step": 16750
+    },
+    {
+      "epoch": 0.8324227674580312,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007334101519817224,
+      "loss": 0.6543,
+      "step": 16760
+    },
+    {
+      "epoch": 0.8329194397536506,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.000733370418198073,
+      "loss": 0.6634,
+      "step": 16770
+    },
+    {
+      "epoch": 0.8334161120492699,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0007333306844144234,
+      "loss": 0.6675,
+      "step": 16780
+    },
+    {
+      "epoch": 0.8339127843448892,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007332909506307738,
+      "loss": 0.6521,
+      "step": 16790
+    },
+    {
+      "epoch": 0.8344094566405086,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0007332512168471243,
+      "loss": 0.6378,
+      "step": 16800
+    },
+    {
+      "epoch": 0.8349061289361279,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0007332114830634747,
+      "loss": 0.6496,
+      "step": 16810
+    },
+    {
+      "epoch": 0.8354028012317473,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0007331717492798252,
+      "loss": 0.6638,
+      "step": 16820
+    },
+    {
+      "epoch": 0.8358994735273666,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007331320154961757,
+      "loss": 0.6344,
+      "step": 16830
+    },
+    {
+      "epoch": 0.836396145822986,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0007330922817125261,
+      "loss": 0.6616,
+      "step": 16840
+    },
+    {
+      "epoch": 0.8368928181186054,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0007330525479288765,
+      "loss": 0.6469,
+      "step": 16850
+    },
+    {
+      "epoch": 0.8373894904142247,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.000733012814145227,
+      "loss": 0.6803,
+      "step": 16860
+    },
+    {
+      "epoch": 0.8378861627098441,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0007329730803615775,
+      "loss": 0.6452,
+      "step": 16870
+    },
+    {
+      "epoch": 0.8383828350054634,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.000732933346577928,
+      "loss": 0.6699,
+      "step": 16880
+    },
+    {
+      "epoch": 0.8388795073010827,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0007328936127942783,
+      "loss": 0.6362,
+      "step": 16890
+    },
+    {
+      "epoch": 0.8393761795967021,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0007328538790106288,
+      "loss": 0.6622,
+      "step": 16900
+    },
+    {
+      "epoch": 0.8398728518923214,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0007328141452269793,
+      "loss": 0.6484,
+      "step": 16910
+    },
+    {
+      "epoch": 0.8403695241879408,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0007327744114433297,
+      "loss": 0.6497,
+      "step": 16920
+    },
+    {
+      "epoch": 0.8408661964835602,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0007327346776596802,
+      "loss": 0.6783,
+      "step": 16930
+    },
+    {
+      "epoch": 0.8413628687791795,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0007326949438760307,
+      "loss": 0.6665,
+      "step": 16940
+    },
+    {
+      "epoch": 0.8418595410747989,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.000732655210092381,
+      "loss": 0.6902,
+      "step": 16950
+    },
+    {
+      "epoch": 0.8423562133704182,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0007326154763087315,
+      "loss": 0.6522,
+      "step": 16960
+    },
+    {
+      "epoch": 0.8428528856660376,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.000732575742525082,
+      "loss": 0.6571,
+      "step": 16970
+    },
+    {
+      "epoch": 0.8433495579616569,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0007325360087414324,
+      "loss": 0.657,
+      "step": 16980
+    },
+    {
+      "epoch": 0.8438462302572762,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007324962749577829,
+      "loss": 0.648,
+      "step": 16990
+    },
+    {
+      "epoch": 0.8443429025528956,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007324565411741333,
+      "loss": 0.6394,
+      "step": 17000
+    },
+    {
+      "epoch": 0.8448395748485149,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0007324168073904837,
+      "loss": 0.6504,
+      "step": 17010
+    },
+    {
+      "epoch": 0.8453362471441342,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0007323770736068343,
+      "loss": 0.6483,
+      "step": 17020
+    },
+    {
+      "epoch": 0.8458329194397537,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0007323373398231847,
+      "loss": 0.6492,
+      "step": 17030
+    },
+    {
+      "epoch": 0.846329591735373,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007322976060395352,
+      "loss": 0.6662,
+      "step": 17040
+    },
+    {
+      "epoch": 0.8468262640309924,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0007322578722558856,
+      "loss": 0.6364,
+      "step": 17050
+    },
+    {
+      "epoch": 0.8473229363266117,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.000732218138472236,
+      "loss": 0.6674,
+      "step": 17060
+    },
+    {
+      "epoch": 0.847819608622231,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007321784046885866,
+      "loss": 0.6588,
+      "step": 17070
+    },
+    {
+      "epoch": 0.8483162809178504,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000732138670904937,
+      "loss": 0.6432,
+      "step": 17080
+    },
+    {
+      "epoch": 0.8488129532134697,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0007320989371212874,
+      "loss": 0.6654,
+      "step": 17090
+    },
+    {
+      "epoch": 0.8493096255090891,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0007320592033376379,
+      "loss": 0.6632,
+      "step": 17100
+    },
+    {
+      "epoch": 0.8498062978047084,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0007320194695539882,
+      "loss": 0.6595,
+      "step": 17110
+    },
+    {
+      "epoch": 0.8503029701003278,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0007319797357703388,
+      "loss": 0.6614,
+      "step": 17120
+    },
+    {
+      "epoch": 0.8507996423959472,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0007319400019866893,
+      "loss": 0.6409,
+      "step": 17130
+    },
+    {
+      "epoch": 0.8512963146915665,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007319002682030396,
+      "loss": 0.6245,
+      "step": 17140
+    },
+    {
+      "epoch": 0.8517929869871859,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0007318605344193901,
+      "loss": 0.6708,
+      "step": 17150
+    },
+    {
+      "epoch": 0.8522896592828052,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007318208006357405,
+      "loss": 0.6608,
+      "step": 17160
+    },
+    {
+      "epoch": 0.8527863315784245,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0007317810668520911,
+      "loss": 0.6411,
+      "step": 17170
+    },
+    {
+      "epoch": 0.8532830038740439,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0007317413330684415,
+      "loss": 0.6536,
+      "step": 17180
+    },
+    {
+      "epoch": 0.8537796761696632,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007317015992847919,
+      "loss": 0.6427,
+      "step": 17190
+    },
+    {
+      "epoch": 0.8542763484652826,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0007316618655011424,
+      "loss": 0.6913,
+      "step": 17200
+    },
+    {
+      "epoch": 0.854773020760902,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007316221317174928,
+      "loss": 0.6436,
+      "step": 17210
+    },
+    {
+      "epoch": 0.8552696930565213,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0007315823979338433,
+      "loss": 0.6381,
+      "step": 17220
+    },
+    {
+      "epoch": 0.8557663653521407,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0007315426641501938,
+      "loss": 0.66,
+      "step": 17230
+    },
+    {
+      "epoch": 0.85626303764776,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0007315029303665442,
+      "loss": 0.6569,
+      "step": 17240
+    },
+    {
+      "epoch": 0.8567597099433794,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0007314631965828946,
+      "loss": 0.6972,
+      "step": 17250
+    },
+    {
+      "epoch": 0.8572563822389987,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0007314234627992451,
+      "loss": 0.6668,
+      "step": 17260
+    },
+    {
+      "epoch": 0.857753054534618,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0007313837290155956,
+      "loss": 0.6675,
+      "step": 17270
+    },
+    {
+      "epoch": 0.8582497268302374,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.000731343995231946,
+      "loss": 0.6405,
+      "step": 17280
+    },
+    {
+      "epoch": 0.8587463991258567,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0007313042614482965,
+      "loss": 0.657,
+      "step": 17290
+    },
+    {
+      "epoch": 0.8592430714214762,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007312645276646468,
+      "loss": 0.6507,
+      "step": 17300
+    },
+    {
+      "epoch": 0.8597397437170955,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0007312247938809973,
+      "loss": 0.6625,
+      "step": 17310
+    },
+    {
+      "epoch": 0.8602364160127148,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007311850600973479,
+      "loss": 0.6755,
+      "step": 17320
+    },
+    {
+      "epoch": 0.8607330883083342,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007311453263136983,
+      "loss": 0.6671,
+      "step": 17330
+    },
+    {
+      "epoch": 0.8612297606039535,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0007311055925300487,
+      "loss": 0.6828,
+      "step": 17340
+    },
+    {
+      "epoch": 0.8617264328995728,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0007310658587463992,
+      "loss": 0.6169,
+      "step": 17350
+    },
+    {
+      "epoch": 0.8622231051951922,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007310261249627496,
+      "loss": 0.6387,
+      "step": 17360
+    },
+    {
+      "epoch": 0.8627197774908115,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007309863911791001,
+      "loss": 0.7025,
+      "step": 17370
+    },
+    {
+      "epoch": 0.8632164497864309,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0007309466573954505,
+      "loss": 0.6599,
+      "step": 17380
+    },
+    {
+      "epoch": 0.8637131220820503,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.000730906923611801,
+      "loss": 0.6288,
+      "step": 17390
+    },
+    {
+      "epoch": 0.8642097943776696,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0007308671898281514,
+      "loss": 0.6718,
+      "step": 17400
+    },
+    {
+      "epoch": 0.864706466673289,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007308274560445018,
+      "loss": 0.6607,
+      "step": 17410
+    },
+    {
+      "epoch": 0.8652031389689083,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0007307877222608524,
+      "loss": 0.6593,
+      "step": 17420
+    },
+    {
+      "epoch": 0.8656998112645277,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007307479884772028,
+      "loss": 0.6401,
+      "step": 17430
+    },
+    {
+      "epoch": 0.866196483560147,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007307082546935532,
+      "loss": 0.6316,
+      "step": 17440
+    },
+    {
+      "epoch": 0.8666931558557663,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0007306685209099037,
+      "loss": 0.6319,
+      "step": 17450
+    },
+    {
+      "epoch": 0.8671898281513857,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007306287871262541,
+      "loss": 0.6675,
+      "step": 17460
+    },
+    {
+      "epoch": 0.867686500447005,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0007305890533426046,
+      "loss": 0.6245,
+      "step": 17470
+    },
+    {
+      "epoch": 0.8681831727426245,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007305493195589551,
+      "loss": 0.6518,
+      "step": 17480
+    },
+    {
+      "epoch": 0.8686798450382438,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0007305095857753055,
+      "loss": 0.6574,
+      "step": 17490
+    },
+    {
+      "epoch": 0.8691765173338631,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007304698519916559,
+      "loss": 0.6386,
+      "step": 17500
+    },
+    {
+      "epoch": 0.8696731896294825,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0007304301182080064,
+      "loss": 0.6485,
+      "step": 17510
+    },
+    {
+      "epoch": 0.8701698619251018,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007303903844243569,
+      "loss": 0.653,
+      "step": 17520
+    },
+    {
+      "epoch": 0.8706665342207212,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007303506506407073,
+      "loss": 0.6777,
+      "step": 17530
+    },
+    {
+      "epoch": 0.8711632065163405,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0007303109168570578,
+      "loss": 0.6541,
+      "step": 17540
+    },
+    {
+      "epoch": 0.8716598788119598,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0007302711830734082,
+      "loss": 0.6746,
+      "step": 17550
+    },
+    {
+      "epoch": 0.8721565511075792,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007302314492897586,
+      "loss": 0.6732,
+      "step": 17560
+    },
+    {
+      "epoch": 0.8726532234031986,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.000730191715506109,
+      "loss": 0.6948,
+      "step": 17570
+    },
+    {
+      "epoch": 0.873149895698818,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007301519817224596,
+      "loss": 0.6498,
+      "step": 17580
+    },
+    {
+      "epoch": 0.8736465679944373,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00073011224793881,
+      "loss": 0.6505,
+      "step": 17590
+    },
+    {
+      "epoch": 0.8741432402900566,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0007300725141551604,
+      "loss": 0.6503,
+      "step": 17600
+    },
+    {
+      "epoch": 0.874639912585676,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007300327803715109,
+      "loss": 0.6586,
+      "step": 17610
+    },
+    {
+      "epoch": 0.8751365848812953,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0007299930465878615,
+      "loss": 0.6732,
+      "step": 17620
+    },
+    {
+      "epoch": 0.8756332571769146,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0007299533128042118,
+      "loss": 0.6564,
+      "step": 17630
+    },
+    {
+      "epoch": 0.876129929472534,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007299135790205623,
+      "loss": 0.6678,
+      "step": 17640
+    },
+    {
+      "epoch": 0.8766266017681533,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007298738452369127,
+      "loss": 0.6443,
+      "step": 17650
+    },
+    {
+      "epoch": 0.8771232740637728,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0007298341114532631,
+      "loss": 0.643,
+      "step": 17660
+    },
+    {
+      "epoch": 0.8776199463593921,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007297943776696137,
+      "loss": 0.6385,
+      "step": 17670
+    },
+    {
+      "epoch": 0.8781166186550114,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0007297546438859641,
+      "loss": 0.6425,
+      "step": 17680
+    },
+    {
+      "epoch": 0.8786132909506308,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0007297149101023145,
+      "loss": 0.6334,
+      "step": 17690
+    },
+    {
+      "epoch": 0.8791099632462501,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.000729675176318665,
+      "loss": 0.6628,
+      "step": 17700
+    },
+    {
+      "epoch": 0.8796066355418695,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0007296354425350154,
+      "loss": 0.674,
+      "step": 17710
+    },
+    {
+      "epoch": 0.8801033078374888,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0007295957087513658,
+      "loss": 0.6799,
+      "step": 17720
+    },
+    {
+      "epoch": 0.8805999801331081,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007295559749677164,
+      "loss": 0.6566,
+      "step": 17730
+    },
+    {
+      "epoch": 0.8810966524287275,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0007295162411840668,
+      "loss": 0.6363,
+      "step": 17740
+    },
+    {
+      "epoch": 0.8815933247243469,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007294765074004172,
+      "loss": 0.681,
+      "step": 17750
+    },
+    {
+      "epoch": 0.8820899970199663,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007294367736167676,
+      "loss": 0.6795,
+      "step": 17760
+    },
+    {
+      "epoch": 0.8825866693155856,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007293970398331182,
+      "loss": 0.668,
+      "step": 17770
+    },
+    {
+      "epoch": 0.8830833416112049,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0007293573060494687,
+      "loss": 0.6346,
+      "step": 17780
+    },
+    {
+      "epoch": 0.8835800139068243,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.000729317572265819,
+      "loss": 0.6488,
+      "step": 17790
+    },
+    {
+      "epoch": 0.8840766862024436,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007292778384821695,
+      "loss": 0.6297,
+      "step": 17800
+    },
+    {
+      "epoch": 0.884573358498063,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.00072923810469852,
+      "loss": 0.6498,
+      "step": 17810
+    },
+    {
+      "epoch": 0.8850700307936823,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007291983709148703,
+      "loss": 0.6379,
+      "step": 17820
+    },
+    {
+      "epoch": 0.8855667030893016,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0007291586371312209,
+      "loss": 0.6485,
+      "step": 17830
+    },
+    {
+      "epoch": 0.8860633753849211,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0007291189033475713,
+      "loss": 0.6255,
+      "step": 17840
+    },
+    {
+      "epoch": 0.8865600476805404,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0007290791695639217,
+      "loss": 0.6269,
+      "step": 17850
+    },
+    {
+      "epoch": 0.8870567199761598,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007290394357802722,
+      "loss": 0.658,
+      "step": 17860
+    },
+    {
+      "epoch": 0.8875533922717791,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0007289997019966226,
+      "loss": 0.6499,
+      "step": 17870
+    },
+    {
+      "epoch": 0.8880500645673984,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0007289599682129731,
+      "loss": 0.6215,
+      "step": 17880
+    },
+    {
+      "epoch": 0.8885467368630178,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007289202344293236,
+      "loss": 0.638,
+      "step": 17890
+    },
+    {
+      "epoch": 0.8890434091586371,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.000728880500645674,
+      "loss": 0.6508,
+      "step": 17900
+    },
+    {
+      "epoch": 0.8895400814542564,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0007288407668620245,
+      "loss": 0.6515,
+      "step": 17910
+    },
+    {
+      "epoch": 0.8900367537498758,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.000728801033078375,
+      "loss": 0.6798,
+      "step": 17920
+    },
+    {
+      "epoch": 0.8905334260454952,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0007287612992947254,
+      "loss": 0.6558,
+      "step": 17930
+    },
+    {
+      "epoch": 0.8910300983411146,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007287215655110759,
+      "loss": 0.6627,
+      "step": 17940
+    },
+    {
+      "epoch": 0.8915267706367339,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0007286818317274263,
+      "loss": 0.6434,
+      "step": 17950
+    },
+    {
+      "epoch": 0.8920234429323532,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007286420979437767,
+      "loss": 0.636,
+      "step": 17960
+    },
+    {
+      "epoch": 0.8925201152279726,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007286023641601273,
+      "loss": 0.6488,
+      "step": 17970
+    },
+    {
+      "epoch": 0.8930167875235919,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0007285626303764776,
+      "loss": 0.6529,
+      "step": 17980
+    },
+    {
+      "epoch": 0.8935134598192113,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0007285228965928281,
+      "loss": 0.6544,
+      "step": 17990
+    },
+    {
+      "epoch": 0.8940101321148306,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007284831628091786,
+      "loss": 0.6757,
+      "step": 18000
+    },
+    {
+      "epoch": 0.8945068044104499,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0007284434290255289,
+      "loss": 0.6446,
+      "step": 18010
+    },
+    {
+      "epoch": 0.8950034767060694,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007284036952418794,
+      "loss": 0.6686,
+      "step": 18020
+    },
+    {
+      "epoch": 0.8955001490016887,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0007283639614582299,
+      "loss": 0.6257,
+      "step": 18030
+    },
+    {
+      "epoch": 0.8959968212973081,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007283242276745803,
+      "loss": 0.684,
+      "step": 18040
+    },
+    {
+      "epoch": 0.8964934935929274,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007282844938909308,
+      "loss": 0.6261,
+      "step": 18050
+    },
+    {
+      "epoch": 0.8969901658885467,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0007282447601072812,
+      "loss": 0.6511,
+      "step": 18060
+    },
+    {
+      "epoch": 0.8974868381841661,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0007282050263236317,
+      "loss": 0.669,
+      "step": 18070
+    },
+    {
+      "epoch": 0.8979835104797854,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007281652925399822,
+      "loss": 0.6212,
+      "step": 18080
+    },
+    {
+      "epoch": 0.8984801827754048,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0007281255587563326,
+      "loss": 0.6908,
+      "step": 18090
+    },
+    {
+      "epoch": 0.8989768550710241,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0007280858249726831,
+      "loss": 0.6437,
+      "step": 18100
+    },
+    {
+      "epoch": 0.8994735273666435,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0007280460911890335,
+      "loss": 0.6976,
+      "step": 18110
+    },
+    {
+      "epoch": 0.8999701996622629,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0007280063574053839,
+      "loss": 0.6431,
+      "step": 18120
+    },
+    {
+      "epoch": 0.9004668719578822,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0007279666236217345,
+      "loss": 0.6197,
+      "step": 18130
+    },
+    {
+      "epoch": 0.9009635442535016,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007279268898380849,
+      "loss": 0.6198,
+      "step": 18140
+    },
+    {
+      "epoch": 0.9014602165491209,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007278871560544353,
+      "loss": 0.6532,
+      "step": 18150
+    },
+    {
+      "epoch": 0.9019568888447402,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007278474222707858,
+      "loss": 0.6735,
+      "step": 18160
+    },
+    {
+      "epoch": 0.9024535611403596,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007278076884871361,
+      "loss": 0.6289,
+      "step": 18170
+    },
+    {
+      "epoch": 0.9029502334359789,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007277679547034867,
+      "loss": 0.6466,
+      "step": 18180
+    },
+    {
+      "epoch": 0.9034469057315982,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007277282209198372,
+      "loss": 0.6393,
+      "step": 18190
+    },
+    {
+      "epoch": 0.9039435780272176,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0007276884871361875,
+      "loss": 0.6556,
+      "step": 18200
+    },
+    {
+      "epoch": 0.904440250322837,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.000727648753352538,
+      "loss": 0.6643,
+      "step": 18210
+    },
+    {
+      "epoch": 0.9049369226184564,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0007276090195688885,
+      "loss": 0.6401,
+      "step": 18220
+    },
+    {
+      "epoch": 0.9054335949140757,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.000727569285785239,
+      "loss": 0.6655,
+      "step": 18230
+    },
+    {
+      "epoch": 0.905930267209695,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007275295520015894,
+      "loss": 0.6161,
+      "step": 18240
+    },
+    {
+      "epoch": 0.9064269395053144,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007274898182179398,
+      "loss": 0.6177,
+      "step": 18250
+    },
+    {
+      "epoch": 0.9069236118009337,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0007274500844342903,
+      "loss": 0.6307,
+      "step": 18260
+    },
+    {
+      "epoch": 0.9074202840965531,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0007274103506506407,
+      "loss": 0.6363,
+      "step": 18270
+    },
+    {
+      "epoch": 0.9079169563921724,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0007273706168669912,
+      "loss": 0.6561,
+      "step": 18280
+    },
+    {
+      "epoch": 0.9084136286877917,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007273308830833417,
+      "loss": 0.6347,
+      "step": 18290
+    },
+    {
+      "epoch": 0.9089103009834112,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007272911492996921,
+      "loss": 0.6362,
+      "step": 18300
+    },
+    {
+      "epoch": 0.9094069732790305,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007272514155160425,
+      "loss": 0.6456,
+      "step": 18310
+    },
+    {
+      "epoch": 0.9099036455746499,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.000727211681732393,
+      "loss": 0.6526,
+      "step": 18320
+    },
+    {
+      "epoch": 0.9104003178702692,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0007271719479487435,
+      "loss": 0.6516,
+      "step": 18330
+    },
+    {
+      "epoch": 0.9108969901658885,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007271322141650939,
+      "loss": 0.6862,
+      "step": 18340
+    },
+    {
+      "epoch": 0.9113936624615079,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007270924803814444,
+      "loss": 0.6343,
+      "step": 18350
+    },
+    {
+      "epoch": 0.9118903347571272,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0007270527465977948,
+      "loss": 0.6607,
+      "step": 18360
+    },
+    {
+      "epoch": 0.9123870070527466,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007270130128141452,
+      "loss": 0.6355,
+      "step": 18370
+    },
+    {
+      "epoch": 0.9128836793483659,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007269732790304958,
+      "loss": 0.6391,
+      "step": 18380
+    },
+    {
+      "epoch": 0.9133803516439853,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0007269335452468462,
+      "loss": 0.638,
+      "step": 18390
+    },
+    {
+      "epoch": 0.9138770239396047,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0007268938114631966,
+      "loss": 0.6353,
+      "step": 18400
+    },
+    {
+      "epoch": 0.914373696235224,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007268540776795471,
+      "loss": 0.6596,
+      "step": 18410
+    },
+    {
+      "epoch": 0.9148703685308434,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0007268143438958975,
+      "loss": 0.6434,
+      "step": 18420
+    },
+    {
+      "epoch": 0.9153670408264627,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.000726774610112248,
+      "loss": 0.6595,
+      "step": 18430
+    },
+    {
+      "epoch": 0.915863713122082,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007267348763285984,
+      "loss": 0.6257,
+      "step": 18440
+    },
+    {
+      "epoch": 0.9163603854177014,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007266951425449489,
+      "loss": 0.6293,
+      "step": 18450
+    },
+    {
+      "epoch": 0.9168570577133207,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007266554087612993,
+      "loss": 0.6326,
+      "step": 18460
+    },
+    {
+      "epoch": 0.91735373000894,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007266156749776497,
+      "loss": 0.6738,
+      "step": 18470
+    },
+    {
+      "epoch": 0.9178504023045595,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0007265759411940003,
+      "loss": 0.64,
+      "step": 18480
+    },
+    {
+      "epoch": 0.9183470746001788,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007265362074103507,
+      "loss": 0.6349,
+      "step": 18490
+    },
+    {
+      "epoch": 0.9188437468957982,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0007264964736267011,
+      "loss": 0.6403,
+      "step": 18500
+    },
+    {
+      "epoch": 0.9193404191914175,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0007264567398430516,
+      "loss": 0.666,
+      "step": 18510
+    },
+    {
+      "epoch": 0.9198370914870369,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.000726417006059402,
+      "loss": 0.642,
+      "step": 18520
+    },
+    {
+      "epoch": 0.9203337637826562,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0007263772722757525,
+      "loss": 0.6183,
+      "step": 18530
+    },
+    {
+      "epoch": 0.9208304360782755,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.000726337538492103,
+      "loss": 0.6488,
+      "step": 18540
+    },
+    {
+      "epoch": 0.9213271083738949,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0007262978047084534,
+      "loss": 0.6425,
+      "step": 18550
+    },
+    {
+      "epoch": 0.9218237806695142,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007262580709248038,
+      "loss": 0.6517,
+      "step": 18560
+    },
+    {
+      "epoch": 0.9223204529651337,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0007262183371411543,
+      "loss": 0.6648,
+      "step": 18570
+    },
+    {
+      "epoch": 0.922817125260753,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007261786033575048,
+      "loss": 0.6293,
+      "step": 18580
+    },
+    {
+      "epoch": 0.9233137975563723,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0007261388695738552,
+      "loss": 0.6352,
+      "step": 18590
+    },
+    {
+      "epoch": 0.9238104698519917,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007260991357902057,
+      "loss": 0.6556,
+      "step": 18600
+    },
+    {
+      "epoch": 0.924307142147611,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0007260594020065561,
+      "loss": 0.6334,
+      "step": 18610
+    },
+    {
+      "epoch": 0.9248038144432303,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0007260196682229065,
+      "loss": 0.6108,
+      "step": 18620
+    },
+    {
+      "epoch": 0.9253004867388497,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.000725979934439257,
+      "loss": 0.6333,
+      "step": 18630
+    },
+    {
+      "epoch": 0.925797159034469,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007259402006556075,
+      "loss": 0.6463,
+      "step": 18640
+    },
+    {
+      "epoch": 0.9262938313300884,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007259004668719579,
+      "loss": 0.6333,
+      "step": 18650
+    },
+    {
+      "epoch": 0.9267905036257078,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0007258607330883083,
+      "loss": 0.676,
+      "step": 18660
+    },
+    {
+      "epoch": 0.9272871759213271,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007258209993046588,
+      "loss": 0.6571,
+      "step": 18670
+    },
+    {
+      "epoch": 0.9277838482169465,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007257812655210094,
+      "loss": 0.6544,
+      "step": 18680
+    },
+    {
+      "epoch": 0.9282805205125658,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0007257415317373597,
+      "loss": 0.6293,
+      "step": 18690
+    },
+    {
+      "epoch": 0.9287771928081852,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0007257017979537102,
+      "loss": 0.6482,
+      "step": 18700
+    },
+    {
+      "epoch": 0.9292738651038045,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0007256620641700606,
+      "loss": 0.6875,
+      "step": 18710
+    },
+    {
+      "epoch": 0.9297705373994238,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000725622330386411,
+      "loss": 0.6797,
+      "step": 18720
+    },
+    {
+      "epoch": 0.9302672096950432,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0007255825966027616,
+      "loss": 0.6506,
+      "step": 18730
+    },
+    {
+      "epoch": 0.9307638819906625,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.000725542862819112,
+      "loss": 0.6089,
+      "step": 18740
+    },
+    {
+      "epoch": 0.931260554286282,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0007255031290354624,
+      "loss": 0.6626,
+      "step": 18750
+    },
+    {
+      "epoch": 0.9317572265819013,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007254633952518129,
+      "loss": 0.6508,
+      "step": 18760
+    },
+    {
+      "epoch": 0.9322538988775206,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0007254236614681633,
+      "loss": 0.6255,
+      "step": 18770
+    },
+    {
+      "epoch": 0.93275057117314,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0007253839276845138,
+      "loss": 0.6643,
+      "step": 18780
+    },
+    {
+      "epoch": 0.9332472434687593,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007253441939008643,
+      "loss": 0.6427,
+      "step": 18790
+    },
+    {
+      "epoch": 0.9337439157643787,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0007253044601172147,
+      "loss": 0.641,
+      "step": 18800
+    },
+    {
+      "epoch": 0.934240588059998,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0007252647263335652,
+      "loss": 0.6462,
+      "step": 18810
+    },
+    {
+      "epoch": 0.9347372603556173,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0007252249925499155,
+      "loss": 0.6797,
+      "step": 18820
+    },
+    {
+      "epoch": 0.9352339326512367,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.000725185258766266,
+      "loss": 0.6372,
+      "step": 18830
+    },
+    {
+      "epoch": 0.9357306049468561,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007251455249826166,
+      "loss": 0.6683,
+      "step": 18840
+    },
+    {
+      "epoch": 0.9362272772424755,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007251057911989669,
+      "loss": 0.6337,
+      "step": 18850
+    },
+    {
+      "epoch": 0.9367239495380948,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007250660574153174,
+      "loss": 0.6449,
+      "step": 18860
+    },
+    {
+      "epoch": 0.9372206218337141,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007250263236316679,
+      "loss": 0.6515,
+      "step": 18870
+    },
+    {
+      "epoch": 0.9377172941293335,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007249865898480182,
+      "loss": 0.6232,
+      "step": 18880
+    },
+    {
+      "epoch": 0.9382139664249528,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007249468560643688,
+      "loss": 0.6598,
+      "step": 18890
+    },
+    {
+      "epoch": 0.9387106387205721,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007249071222807192,
+      "loss": 0.6842,
+      "step": 18900
+    },
+    {
+      "epoch": 0.9392073110161915,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007248673884970696,
+      "loss": 0.6436,
+      "step": 18910
+    },
+    {
+      "epoch": 0.9397039833118108,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0007248276547134201,
+      "loss": 0.6466,
+      "step": 18920
+    },
+    {
+      "epoch": 0.9402006556074303,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0007247879209297706,
+      "loss": 0.6312,
+      "step": 18930
+    },
+    {
+      "epoch": 0.9406973279030496,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.000724748187146121,
+      "loss": 0.676,
+      "step": 18940
+    },
+    {
+      "epoch": 0.941194000198669,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007247084533624715,
+      "loss": 0.6399,
+      "step": 18950
+    },
+    {
+      "epoch": 0.9416906724942883,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007246687195788219,
+      "loss": 0.6462,
+      "step": 18960
+    },
+    {
+      "epoch": 0.9421873447899076,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0007246289857951724,
+      "loss": 0.6322,
+      "step": 18970
+    },
+    {
+      "epoch": 0.942684017085527,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0007245892520115229,
+      "loss": 0.6487,
+      "step": 18980
+    },
+    {
+      "epoch": 0.9431806893811463,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007245495182278733,
+      "loss": 0.6321,
+      "step": 18990
+    },
+    {
+      "epoch": 0.9436773616767656,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0007245097844442238,
+      "loss": 0.6716,
+      "step": 19000
+    },
+    {
+      "epoch": 0.944174033972385,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007244700506605742,
+      "loss": 0.6537,
+      "step": 19010
+    },
+    {
+      "epoch": 0.9446707062680044,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0007244303168769246,
+      "loss": 0.6177,
+      "step": 19020
+    },
+    {
+      "epoch": 0.9451673785636238,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0007243905830932752,
+      "loss": 0.6269,
+      "step": 19030
+    },
+    {
+      "epoch": 0.9456640508592431,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0007243508493096255,
+      "loss": 0.6405,
+      "step": 19040
+    },
+    {
+      "epoch": 0.9461607231548624,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.000724311115525976,
+      "loss": 0.6169,
+      "step": 19050
+    },
+    {
+      "epoch": 0.9466573954504818,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007242713817423265,
+      "loss": 0.6676,
+      "step": 19060
+    },
+    {
+      "epoch": 0.9471540677461011,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0007242316479586768,
+      "loss": 0.6352,
+      "step": 19070
+    },
+    {
+      "epoch": 0.9476507400417205,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007241919141750274,
+      "loss": 0.6393,
+      "step": 19080
+    },
+    {
+      "epoch": 0.9481474123373398,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0007241521803913778,
+      "loss": 0.6591,
+      "step": 19090
+    },
+    {
+      "epoch": 0.9486440846329591,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007241124466077282,
+      "loss": 0.6649,
+      "step": 19100
+    },
+    {
+      "epoch": 0.9491407569285786,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0007240727128240787,
+      "loss": 0.6774,
+      "step": 19110
+    },
+    {
+      "epoch": 0.9496374292241979,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0007240329790404291,
+      "loss": 0.6421,
+      "step": 19120
+    },
+    {
+      "epoch": 0.9501341015198173,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0007239932452567797,
+      "loss": 0.6432,
+      "step": 19130
+    },
+    {
+      "epoch": 0.9506307738154366,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0007239535114731301,
+      "loss": 0.6313,
+      "step": 19140
+    },
+    {
+      "epoch": 0.9511274461110559,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0007239137776894805,
+      "loss": 0.6938,
+      "step": 19150
+    },
+    {
+      "epoch": 0.9516241184066753,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.000723874043905831,
+      "loss": 0.6474,
+      "step": 19160
+    },
+    {
+      "epoch": 0.9521207907022946,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0007238343101221814,
+      "loss": 0.66,
+      "step": 19170
+    },
+    {
+      "epoch": 0.9526174629979139,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0007237945763385318,
+      "loss": 0.6707,
+      "step": 19180
+    },
+    {
+      "epoch": 0.9531141352935333,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0007237548425548824,
+      "loss": 0.6322,
+      "step": 19190
+    },
+    {
+      "epoch": 0.9536108075891527,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0007237151087712328,
+      "loss": 0.6423,
+      "step": 19200
+    },
+    {
+      "epoch": 0.9541074798847721,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0007236753749875832,
+      "loss": 0.6438,
+      "step": 19210
+    },
+    {
+      "epoch": 0.9546041521803914,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0007236356412039337,
+      "loss": 0.65,
+      "step": 19220
+    },
+    {
+      "epoch": 0.9551008244760107,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.000723595907420284,
+      "loss": 0.6024,
+      "step": 19230
+    },
+    {
+      "epoch": 0.9555974967716301,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007235561736366346,
+      "loss": 0.6478,
+      "step": 19240
+    },
+    {
+      "epoch": 0.9560941690672494,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0007235164398529851,
+      "loss": 0.6551,
+      "step": 19250
+    },
+    {
+      "epoch": 0.9565908413628688,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007234767060693355,
+      "loss": 0.6244,
+      "step": 19260
+    },
+    {
+      "epoch": 0.9570875136584881,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007234369722856859,
+      "loss": 0.6126,
+      "step": 19270
+    },
+    {
+      "epoch": 0.9575841859541074,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007233972385020365,
+      "loss": 0.6673,
+      "step": 19280
+    },
+    {
+      "epoch": 0.9580808582497269,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0007233575047183869,
+      "loss": 0.6338,
+      "step": 19290
+    },
+    {
+      "epoch": 0.9585775305453462,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0007233177709347373,
+      "loss": 0.6413,
+      "step": 19300
+    },
+    {
+      "epoch": 0.9590742028409656,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0007232780371510877,
+      "loss": 0.6325,
+      "step": 19310
+    },
+    {
+      "epoch": 0.9595708751365849,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0007232383033674382,
+      "loss": 0.6368,
+      "step": 19320
+    },
+    {
+      "epoch": 0.9600675474322042,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0007231985695837886,
+      "loss": 0.6868,
+      "step": 19330
+    },
+    {
+      "epoch": 0.9605642197278236,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0007231588358001391,
+      "loss": 0.6554,
+      "step": 19340
+    },
+    {
+      "epoch": 0.9610608920234429,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007231191020164896,
+      "loss": 0.6511,
+      "step": 19350
+    },
+    {
+      "epoch": 0.9615575643190623,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00072307936823284,
+      "loss": 0.601,
+      "step": 19360
+    },
+    {
+      "epoch": 0.9620542366146816,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0007230396344491904,
+      "loss": 0.6424,
+      "step": 19370
+    },
+    {
+      "epoch": 0.9625509089103009,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.000722999900665541,
+      "loss": 0.6213,
+      "step": 19380
+    },
+    {
+      "epoch": 0.9630475812059204,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0007229601668818914,
+      "loss": 0.6322,
+      "step": 19390
+    },
+    {
+      "epoch": 0.9635442535015397,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007229204330982418,
+      "loss": 0.639,
+      "step": 19400
+    },
+    {
+      "epoch": 0.964040925797159,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0007228806993145923,
+      "loss": 0.6287,
+      "step": 19410
+    },
+    {
+      "epoch": 0.9645375980927784,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0007228409655309427,
+      "loss": 0.6506,
+      "step": 19420
+    },
+    {
+      "epoch": 0.9650342703883977,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0007228012317472931,
+      "loss": 0.6512,
+      "step": 19430
+    },
+    {
+      "epoch": 0.9655309426840171,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007227614979636437,
+      "loss": 0.6371,
+      "step": 19440
+    },
+    {
+      "epoch": 0.9660276149796364,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0007227217641799941,
+      "loss": 0.6314,
+      "step": 19450
+    },
+    {
+      "epoch": 0.9665242872752557,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0007226820303963445,
+      "loss": 0.633,
+      "step": 19460
+    },
+    {
+      "epoch": 0.9670209595708751,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.000722642296612695,
+      "loss": 0.6049,
+      "step": 19470
+    },
+    {
+      "epoch": 0.9675176318664945,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0007226025628290454,
+      "loss": 0.6461,
+      "step": 19480
+    },
+    {
+      "epoch": 0.9680143041621139,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0007225628290453959,
+      "loss": 0.6392,
+      "step": 19490
+    },
+    {
+      "epoch": 0.9685109764577332,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007225230952617463,
+      "loss": 0.6055,
+      "step": 19500
+    },
+    {
+      "epoch": 0.9690076487533525,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007224833614780968,
+      "loss": 0.6556,
+      "step": 19510
+    },
+    {
+      "epoch": 0.9695043210489719,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007224436276944472,
+      "loss": 0.6339,
+      "step": 19520
+    },
+    {
+      "epoch": 0.9700009933445912,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007224038939107976,
+      "loss": 0.655,
+      "step": 19530
+    },
+    {
+      "epoch": 0.9704976656402106,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0007223641601271482,
+      "loss": 0.6511,
+      "step": 19540
+    },
+    {
+      "epoch": 0.9709943379358299,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007223244263434986,
+      "loss": 0.6243,
+      "step": 19550
+    },
+    {
+      "epoch": 0.9714910102314492,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.000722284692559849,
+      "loss": 0.6578,
+      "step": 19560
+    },
+    {
+      "epoch": 0.9719876825270687,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007222449587761995,
+      "loss": 0.6603,
+      "step": 19570
+    },
+    {
+      "epoch": 0.972484354822688,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0007222052249925499,
+      "loss": 0.6494,
+      "step": 19580
+    },
+    {
+      "epoch": 0.9729810271183074,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007221654912089004,
+      "loss": 0.6439,
+      "step": 19590
+    },
+    {
+      "epoch": 0.9734776994139267,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0007221257574252509,
+      "loss": 0.6378,
+      "step": 19600
+    },
+    {
+      "epoch": 0.973974371709546,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0007220860236416013,
+      "loss": 0.6525,
+      "step": 19610
+    },
+    {
+      "epoch": 0.9744710440051654,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0007220462898579517,
+      "loss": 0.6577,
+      "step": 19620
+    },
+    {
+      "epoch": 0.9749677163007847,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0007220065560743022,
+      "loss": 0.6399,
+      "step": 19630
+    },
+    {
+      "epoch": 0.975464388596404,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0007219668222906527,
+      "loss": 0.6335,
+      "step": 19640
+    },
+    {
+      "epoch": 0.9759610608920234,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007219270885070031,
+      "loss": 0.6269,
+      "step": 19650
+    },
+    {
+      "epoch": 0.9764577331876428,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007218873547233536,
+      "loss": 0.653,
+      "step": 19660
+    },
+    {
+      "epoch": 0.9769544054832622,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.000721847620939704,
+      "loss": 0.6416,
+      "step": 19670
+    },
+    {
+      "epoch": 0.9774510777788815,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0007218078871560544,
+      "loss": 0.6566,
+      "step": 19680
+    },
+    {
+      "epoch": 0.9779477500745009,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0007217681533724049,
+      "loss": 0.6215,
+      "step": 19690
+    },
+    {
+      "epoch": 0.9784444223701202,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0007217284195887554,
+      "loss": 0.6143,
+      "step": 19700
+    },
+    {
+      "epoch": 0.9789410946657395,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0007216886858051059,
+      "loss": 0.6546,
+      "step": 19710
+    },
+    {
+      "epoch": 0.9794377669613589,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007216489520214562,
+      "loss": 0.6309,
+      "step": 19720
+    },
+    {
+      "epoch": 0.9799344392569782,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0007216092182378067,
+      "loss": 0.6613,
+      "step": 19730
+    },
+    {
+      "epoch": 0.9804311115525975,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007215694844541573,
+      "loss": 0.634,
+      "step": 19740
+    },
+    {
+      "epoch": 0.980927783848217,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0007215297506705076,
+      "loss": 0.6508,
+      "step": 19750
+    },
+    {
+      "epoch": 0.9814244561438363,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0007214900168868581,
+      "loss": 0.6455,
+      "step": 19760
+    },
+    {
+      "epoch": 0.9819211284394557,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007214502831032085,
+      "loss": 0.6486,
+      "step": 19770
+    },
+    {
+      "epoch": 0.982417800735075,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0007214105493195589,
+      "loss": 0.6511,
+      "step": 19780
+    },
+    {
+      "epoch": 0.9829144730306943,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007213708155359095,
+      "loss": 0.6744,
+      "step": 19790
+    },
+    {
+      "epoch": 0.9834111453263137,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007213310817522599,
+      "loss": 0.6283,
+      "step": 19800
+    },
+    {
+      "epoch": 0.983907817621933,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0007212913479686103,
+      "loss": 0.666,
+      "step": 19810
+    },
+    {
+      "epoch": 0.9844044899175524,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0007212516141849608,
+      "loss": 0.6695,
+      "step": 19820
+    },
+    {
+      "epoch": 0.9849011622131717,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007212118804013112,
+      "loss": 0.6422,
+      "step": 19830
+    },
+    {
+      "epoch": 0.9853978345087911,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0007211721466176617,
+      "loss": 0.6549,
+      "step": 19840
+    },
+    {
+      "epoch": 0.9858945068044105,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0007211324128340122,
+      "loss": 0.6556,
+      "step": 19850
+    },
+    {
+      "epoch": 0.9863911791000298,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0007210926790503626,
+      "loss": 0.645,
+      "step": 19860
+    },
+    {
+      "epoch": 0.9868878513956492,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0007210529452667131,
+      "loss": 0.621,
+      "step": 19870
+    },
+    {
+      "epoch": 0.9873845236912685,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0007210132114830635,
+      "loss": 0.6638,
+      "step": 19880
+    },
+    {
+      "epoch": 0.9878811959868878,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.000720973477699414,
+      "loss": 0.6561,
+      "step": 19890
+    },
+    {
+      "epoch": 0.9883778682825072,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007209337439157645,
+      "loss": 0.6335,
+      "step": 19900
+    },
+    {
+      "epoch": 0.9888745405781265,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0007208940101321148,
+      "loss": 0.6598,
+      "step": 19910
+    },
+    {
+      "epoch": 0.9893712128737459,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0007208542763484653,
+      "loss": 0.6381,
+      "step": 19920
+    },
+    {
+      "epoch": 0.9898678851693653,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0007208145425648158,
+      "loss": 0.6276,
+      "step": 19930
+    },
+    {
+      "epoch": 0.9903645574649846,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0007207748087811662,
+      "loss": 0.6665,
+      "step": 19940
+    },
+    {
+      "epoch": 0.990861229760604,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007207350749975167,
+      "loss": 0.6354,
+      "step": 19950
+    },
+    {
+      "epoch": 0.9913579020562233,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007206953412138671,
+      "loss": 0.6357,
+      "step": 19960
+    },
+    {
+      "epoch": 0.9918545743518427,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007206556074302175,
+      "loss": 0.6177,
+      "step": 19970
+    },
+    {
+      "epoch": 0.992351246647462,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.000720615873646568,
+      "loss": 0.6346,
+      "step": 19980
+    },
+    {
+      "epoch": 0.9928479189430813,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0007205761398629185,
+      "loss": 0.6313,
+      "step": 19990
+    },
+    {
+      "epoch": 0.9933445912387007,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007205364060792689,
+      "loss": 0.6217,
+      "step": 20000
+    },
+    {
+      "epoch": 0.99384126353432,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0007204966722956194,
+      "loss": 0.6423,
+      "step": 20010
+    },
+    {
+      "epoch": 0.9943379358299395,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0007204569385119698,
+      "loss": 0.6366,
+      "step": 20020
+    },
+    {
+      "epoch": 0.9948346081255588,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0007204172047283203,
+      "loss": 0.6465,
+      "step": 20030
+    },
+    {
+      "epoch": 0.9953312804211781,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0007203774709446708,
+      "loss": 0.6485,
+      "step": 20040
+    },
+    {
+      "epoch": 0.9958279527167975,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0007203377371610212,
+      "loss": 0.6753,
+      "step": 20050
+    },
+    {
+      "epoch": 0.9963246250124168,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0007202980033773717,
+      "loss": 0.6464,
+      "step": 20060
+    },
+    {
+      "epoch": 0.9968212973080361,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007202582695937221,
+      "loss": 0.6352,
+      "step": 20070
+    },
+    {
+      "epoch": 0.9973179696036555,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0007202185358100725,
+      "loss": 0.6392,
+      "step": 20080
+    },
+    {
+      "epoch": 0.9978146418992748,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0007201788020264231,
+      "loss": 0.6418,
+      "step": 20090
+    },
+    {
+      "epoch": 0.9983113141948942,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0007201390682427734,
+      "loss": 0.6289,
+      "step": 20100
+    },
+    {
+      "epoch": 0.9988079864905136,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0007200993344591239,
+      "loss": 0.6458,
+      "step": 20110
+    },
+    {
+      "epoch": 0.999304658786133,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007200596006754744,
+      "loss": 0.6511,
+      "step": 20120
+    },
+    {
+      "epoch": 0.9998013310817523,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007200198668918247,
+      "loss": 0.6132,
+      "step": 20130
+    },
+    {
+      "epoch": 1.0002980033773716,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0007199801331081753,
+      "loss": 0.6431,
+      "step": 20140
+    },
+    {
+      "epoch": 1.0007946756729909,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007199403993245258,
+      "loss": 0.6414,
+      "step": 20150
+    },
+    {
+      "epoch": 1.0012913479686103,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0007199006655408762,
+      "loss": 0.6431,
+      "step": 20160
+    },
+    {
+      "epoch": 1.0017880202642298,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0007198609317572266,
+      "loss": 0.6413,
+      "step": 20170
+    },
+    {
+      "epoch": 1.002284692559849,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.000719821197973577,
+      "loss": 0.6061,
+      "step": 20180
+    },
+    {
+      "epoch": 1.0027813648554684,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0007197814641899276,
+      "loss": 0.6099,
+      "step": 20190
+    },
+    {
+      "epoch": 1.0032780371510877,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.000719741730406278,
+      "loss": 0.6247,
+      "step": 20200
+    },
+    {
+      "epoch": 1.003774709446707,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0007197019966226284,
+      "loss": 0.6223,
+      "step": 20210
+    },
+    {
+      "epoch": 1.0042713817423263,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007196622628389789,
+      "loss": 0.6178,
+      "step": 20220
+    },
+    {
+      "epoch": 1.0047680540379458,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0007196225290553293,
+      "loss": 0.6223,
+      "step": 20230
+    },
+    {
+      "epoch": 1.005264726333565,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007195827952716798,
+      "loss": 0.6333,
+      "step": 20240
+    },
+    {
+      "epoch": 1.0057613986291845,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007195430614880303,
+      "loss": 0.6397,
+      "step": 20250
+    },
+    {
+      "epoch": 1.006258070924804,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007195033277043807,
+      "loss": 0.6309,
+      "step": 20260
+    },
+    {
+      "epoch": 1.0067547432204231,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007194635939207311,
+      "loss": 0.6561,
+      "step": 20270
+    },
+    {
+      "epoch": 1.0072514155160426,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0007194238601370816,
+      "loss": 0.6394,
+      "step": 20280
+    },
+    {
+      "epoch": 1.0077480878116618,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0007193841263534319,
+      "loss": 0.6141,
+      "step": 20290
+    },
+    {
+      "epoch": 1.0082447601072813,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007193443925697825,
+      "loss": 0.6236,
+      "step": 20300
+    },
+    {
+      "epoch": 1.0087414324029005,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.000719304658786133,
+      "loss": 0.6215,
+      "step": 20310
+    },
+    {
+      "epoch": 1.00923810469852,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007192649250024834,
+      "loss": 0.5914,
+      "step": 20320
+    },
+    {
+      "epoch": 1.0097347769941392,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007192251912188338,
+      "loss": 0.6805,
+      "step": 20330
+    },
+    {
+      "epoch": 1.0102314492897586,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0007191854574351844,
+      "loss": 0.6498,
+      "step": 20340
+    },
+    {
+      "epoch": 1.010728121585378,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0007191457236515348,
+      "loss": 0.6419,
+      "step": 20350
+    },
+    {
+      "epoch": 1.0112247938809973,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0007191059898678852,
+      "loss": 0.614,
+      "step": 20360
+    },
+    {
+      "epoch": 1.0117214661766167,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0007190662560842356,
+      "loss": 0.6385,
+      "step": 20370
+    },
+    {
+      "epoch": 1.012218138472236,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0007190265223005861,
+      "loss": 0.6472,
+      "step": 20380
+    },
+    {
+      "epoch": 1.0127148107678554,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0007189867885169365,
+      "loss": 0.6339,
+      "step": 20390
+    },
+    {
+      "epoch": 1.0132114830634746,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.000718947054733287,
+      "loss": 0.629,
+      "step": 20400
+    },
+    {
+      "epoch": 1.013708155359094,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0007189073209496375,
+      "loss": 0.6393,
+      "step": 20410
+    },
+    {
+      "epoch": 1.0142048276547133,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0007188675871659879,
+      "loss": 0.615,
+      "step": 20420
+    },
+    {
+      "epoch": 1.0147014999503328,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007188278533823383,
+      "loss": 0.6203,
+      "step": 20430
+    },
+    {
+      "epoch": 1.0151981722459522,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0007187881195986889,
+      "loss": 0.6366,
+      "step": 20440
+    },
+    {
+      "epoch": 1.0156948445415714,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0007187483858150393,
+      "loss": 0.6522,
+      "step": 20450
+    },
+    {
+      "epoch": 1.016191516837191,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0007187086520313897,
+      "loss": 0.6664,
+      "step": 20460
+    },
+    {
+      "epoch": 1.0166881891328101,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007186689182477402,
+      "loss": 0.6364,
+      "step": 20470
+    },
+    {
+      "epoch": 1.0171848614284296,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007186291844640906,
+      "loss": 0.6269,
+      "step": 20480
+    },
+    {
+      "epoch": 1.0176815337240488,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.000718589450680441,
+      "loss": 0.6337,
+      "step": 20490
+    },
+    {
+      "epoch": 1.0181782060196682,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0007185497168967916,
+      "loss": 0.6397,
+      "step": 20500
+    },
+    {
+      "epoch": 1.0186748783152875,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.000718509983113142,
+      "loss": 0.6124,
+      "step": 20510
+    },
+    {
+      "epoch": 1.019171550610907,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007184702493294924,
+      "loss": 0.6223,
+      "step": 20520
+    },
+    {
+      "epoch": 1.0196682229065264,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0007184305155458429,
+      "loss": 0.6654,
+      "step": 20530
+    },
+    {
+      "epoch": 1.0201648952021456,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0007183907817621933,
+      "loss": 0.6354,
+      "step": 20540
+    },
+    {
+      "epoch": 1.020661567497765,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0007183510479785438,
+      "loss": 0.6546,
+      "step": 20550
+    },
+    {
+      "epoch": 1.0211582397933843,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0007183113141948942,
+      "loss": 0.6305,
+      "step": 20560
+    },
+    {
+      "epoch": 1.0216549120890037,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0007182715804112447,
+      "loss": 0.6463,
+      "step": 20570
+    },
+    {
+      "epoch": 1.022151584384623,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0007182318466275951,
+      "loss": 0.6449,
+      "step": 20580
+    },
+    {
+      "epoch": 1.0226482566802424,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0007181921128439455,
+      "loss": 0.6536,
+      "step": 20590
+    },
+    {
+      "epoch": 1.0231449289758616,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007181523790602961,
+      "loss": 0.6502,
+      "step": 20600
+    },
+    {
+      "epoch": 1.023641601271481,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0007181126452766466,
+      "loss": 0.6262,
+      "step": 20610
+    },
+    {
+      "epoch": 1.0241382735671005,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007180729114929969,
+      "loss": 0.6113,
+      "step": 20620
+    },
+    {
+      "epoch": 1.0246349458627197,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0007180331777093474,
+      "loss": 0.6546,
+      "step": 20630
+    },
+    {
+      "epoch": 1.0251316181583392,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0007179934439256978,
+      "loss": 0.6563,
+      "step": 20640
+    },
+    {
+      "epoch": 1.0256282904539584,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0007179537101420483,
+      "loss": 0.6274,
+      "step": 20650
+    },
+    {
+      "epoch": 1.0261249627495779,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0007179139763583988,
+      "loss": 0.636,
+      "step": 20660
+    },
+    {
+      "epoch": 1.026621635045197,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0007178742425747492,
+      "loss": 0.6469,
+      "step": 20670
+    },
+    {
+      "epoch": 1.0271183073408165,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007178345087910996,
+      "loss": 0.6293,
+      "step": 20680
+    },
+    {
+      "epoch": 1.0276149796364358,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007177947750074501,
+      "loss": 0.6118,
+      "step": 20690
+    },
+    {
+      "epoch": 1.0281116519320552,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007177550412238006,
+      "loss": 0.6164,
+      "step": 20700
+    },
+    {
+      "epoch": 1.0286083242276747,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.000717715307440151,
+      "loss": 0.6503,
+      "step": 20710
+    },
+    {
+      "epoch": 1.029104996523294,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007176755736565015,
+      "loss": 0.6428,
+      "step": 20720
+    },
+    {
+      "epoch": 1.0296016688189134,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0007176358398728519,
+      "loss": 0.6194,
+      "step": 20730
+    },
+    {
+      "epoch": 1.0300983411145326,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0007175961060892023,
+      "loss": 0.6299,
+      "step": 20740
+    },
+    {
+      "epoch": 1.030595013410152,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0007175563723055529,
+      "loss": 0.6277,
+      "step": 20750
+    },
+    {
+      "epoch": 1.0310916857057713,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0007175166385219033,
+      "loss": 0.6155,
+      "step": 20760
+    },
+    {
+      "epoch": 1.0315883580013907,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0007174769047382538,
+      "loss": 0.6441,
+      "step": 20770
+    },
+    {
+      "epoch": 1.03208503029701,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007174371709546041,
+      "loss": 0.6151,
+      "step": 20780
+    },
+    {
+      "epoch": 1.0325817025926294,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0007173974371709546,
+      "loss": 0.6207,
+      "step": 20790
+    },
+    {
+      "epoch": 1.0330783748882488,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007173577033873052,
+      "loss": 0.6286,
+      "step": 20800
+    },
+    {
+      "epoch": 1.033575047183868,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007173179696036555,
+      "loss": 0.661,
+      "step": 20810
+    },
+    {
+      "epoch": 1.0340717194794875,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.000717278235820006,
+      "loss": 0.6232,
+      "step": 20820
+    },
+    {
+      "epoch": 1.0345683917751067,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007172385020363564,
+      "loss": 0.6403,
+      "step": 20830
+    },
+    {
+      "epoch": 1.0350650640707262,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0007171987682527068,
+      "loss": 0.6231,
+      "step": 20840
+    },
+    {
+      "epoch": 1.0355617363663454,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007171590344690574,
+      "loss": 0.6646,
+      "step": 20850
+    },
+    {
+      "epoch": 1.0360584086619649,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007171193006854078,
+      "loss": 0.6139,
+      "step": 20860
+    },
+    {
+      "epoch": 1.036555080957584,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0007170795669017582,
+      "loss": 0.6497,
+      "step": 20870
+    },
+    {
+      "epoch": 1.0370517532532035,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0007170398331181087,
+      "loss": 0.644,
+      "step": 20880
+    },
+    {
+      "epoch": 1.037548425548823,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007170000993344591,
+      "loss": 0.6199,
+      "step": 20890
+    },
+    {
+      "epoch": 1.0380450978444422,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007169603655508097,
+      "loss": 0.6642,
+      "step": 20900
+    },
+    {
+      "epoch": 1.0385417701400617,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007169206317671601,
+      "loss": 0.6234,
+      "step": 20910
+    },
+    {
+      "epoch": 1.0390384424356809,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0007168808979835105,
+      "loss": 0.6432,
+      "step": 20920
+    },
+    {
+      "epoch": 1.0395351147313003,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.000716841164199861,
+      "loss": 0.6174,
+      "step": 20930
+    },
+    {
+      "epoch": 1.0400317870269196,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007168014304162114,
+      "loss": 0.6398,
+      "step": 20940
+    },
+    {
+      "epoch": 1.040528459322539,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007167616966325619,
+      "loss": 0.6372,
+      "step": 20950
+    },
+    {
+      "epoch": 1.0410251316181582,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007167219628489124,
+      "loss": 0.6382,
+      "step": 20960
+    },
+    {
+      "epoch": 1.0415218039137777,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0007166822290652627,
+      "loss": 0.6185,
+      "step": 20970
+    },
+    {
+      "epoch": 1.0420184762093971,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007166424952816132,
+      "loss": 0.6185,
+      "step": 20980
+    },
+    {
+      "epoch": 1.0425151485050164,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0007166027614979637,
+      "loss": 0.6616,
+      "step": 20990
+    },
+    {
+      "epoch": 1.0430118208006358,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007165630277143141,
+      "loss": 0.6378,
+      "step": 21000
+    },
+    {
+      "epoch": 1.043508493096255,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007165232939306646,
+      "loss": 0.6186,
+      "step": 21010
+    },
+    {
+      "epoch": 1.0440051653918745,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0007164835601470151,
+      "loss": 0.6414,
+      "step": 21020
+    },
+    {
+      "epoch": 1.0445018376874937,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0007164438263633654,
+      "loss": 0.6409,
+      "step": 21030
+    },
+    {
+      "epoch": 1.0449985099831132,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0007164040925797159,
+      "loss": 0.6121,
+      "step": 21040
+    },
+    {
+      "epoch": 1.0454951822787324,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0007163643587960664,
+      "loss": 0.614,
+      "step": 21050
+    },
+    {
+      "epoch": 1.0459918545743518,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0007163246250124169,
+      "loss": 0.6208,
+      "step": 21060
+    },
+    {
+      "epoch": 1.0464885268699713,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007162848912287673,
+      "loss": 0.6057,
+      "step": 21070
+    },
+    {
+      "epoch": 1.0469851991655905,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0007162451574451177,
+      "loss": 0.6317,
+      "step": 21080
+    },
+    {
+      "epoch": 1.04748187146121,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0007162054236614682,
+      "loss": 0.6259,
+      "step": 21090
+    },
+    {
+      "epoch": 1.0479785437568292,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0007161656898778187,
+      "loss": 0.6358,
+      "step": 21100
+    },
+    {
+      "epoch": 1.0484752160524486,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0007161259560941691,
+      "loss": 0.6219,
+      "step": 21110
+    },
+    {
+      "epoch": 1.0489718883480679,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007160862223105196,
+      "loss": 0.6467,
+      "step": 21120
+    },
+    {
+      "epoch": 1.0494685606436873,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00071604648852687,
+      "loss": 0.5926,
+      "step": 21130
+    },
+    {
+      "epoch": 1.0499652329393065,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0007160067547432204,
+      "loss": 0.6026,
+      "step": 21140
+    },
+    {
+      "epoch": 1.050461905234926,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.000715967020959571,
+      "loss": 0.6293,
+      "step": 21150
+    },
+    {
+      "epoch": 1.0509585775305454,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007159272871759213,
+      "loss": 0.6204,
+      "step": 21160
+    },
+    {
+      "epoch": 1.0514552498261647,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0007158875533922718,
+      "loss": 0.6413,
+      "step": 21170
+    },
+    {
+      "epoch": 1.0519519221217841,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0007158478196086223,
+      "loss": 0.6339,
+      "step": 21180
+    },
+    {
+      "epoch": 1.0524485944174033,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007158080858249726,
+      "loss": 0.6545,
+      "step": 21190
+    },
+    {
+      "epoch": 1.0529452667130228,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007157683520413232,
+      "loss": 0.6232,
+      "step": 21200
+    },
+    {
+      "epoch": 1.053441939008642,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0007157286182576737,
+      "loss": 0.6383,
+      "step": 21210
+    },
+    {
+      "epoch": 1.0539386113042615,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0007156888844740241,
+      "loss": 0.647,
+      "step": 21220
+    },
+    {
+      "epoch": 1.0544352835998807,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007156491506903745,
+      "loss": 0.6368,
+      "step": 21230
+    },
+    {
+      "epoch": 1.0549319558955002,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0007156094169067249,
+      "loss": 0.637,
+      "step": 21240
+    },
+    {
+      "epoch": 1.0554286281911196,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007155696831230755,
+      "loss": 0.6766,
+      "step": 21250
+    },
+    {
+      "epoch": 1.0559253004867388,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0007155299493394259,
+      "loss": 0.6468,
+      "step": 21260
+    },
+    {
+      "epoch": 1.0564219727823583,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0007154902155557763,
+      "loss": 0.6206,
+      "step": 21270
+    },
+    {
+      "epoch": 1.0569186450779775,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0007154504817721268,
+      "loss": 0.6368,
+      "step": 21280
+    },
+    {
+      "epoch": 1.057415317373597,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007154107479884772,
+      "loss": 0.6367,
+      "step": 21290
+    },
+    {
+      "epoch": 1.0579119896692162,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0007153710142048277,
+      "loss": 0.593,
+      "step": 21300
+    },
+    {
+      "epoch": 1.0584086619648356,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0007153312804211782,
+      "loss": 0.6003,
+      "step": 21310
+    },
+    {
+      "epoch": 1.0589053342604549,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0007152915466375286,
+      "loss": 0.6045,
+      "step": 21320
+    },
+    {
+      "epoch": 1.0594020065560743,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.000715251812853879,
+      "loss": 0.6367,
+      "step": 21330
+    },
+    {
+      "epoch": 1.0598986788516938,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007152120790702295,
+      "loss": 0.616,
+      "step": 21340
+    },
+    {
+      "epoch": 1.060395351147313,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00071517234528658,
+      "loss": 0.5995,
+      "step": 21350
+    },
+    {
+      "epoch": 1.0608920234429324,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0007151326115029304,
+      "loss": 0.6266,
+      "step": 21360
+    },
+    {
+      "epoch": 1.0613886957385517,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0007150928777192809,
+      "loss": 0.6305,
+      "step": 21370
+    },
+    {
+      "epoch": 1.061885368034171,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0007150531439356313,
+      "loss": 0.6097,
+      "step": 21380
+    },
+    {
+      "epoch": 1.0623820403297903,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0007150134101519817,
+      "loss": 0.6639,
+      "step": 21390
+    },
+    {
+      "epoch": 1.0628787126254098,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0007149736763683323,
+      "loss": 0.5871,
+      "step": 21400
+    },
+    {
+      "epoch": 1.063375384921029,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0007149339425846827,
+      "loss": 0.6388,
+      "step": 21410
+    },
+    {
+      "epoch": 1.0638720572166485,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0007148942088010331,
+      "loss": 0.6103,
+      "step": 21420
+    },
+    {
+      "epoch": 1.064368729512268,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0007148544750173835,
+      "loss": 0.6276,
+      "step": 21430
+    },
+    {
+      "epoch": 1.0648654018078871,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.000714814741233734,
+      "loss": 0.6362,
+      "step": 21440
+    },
+    {
+      "epoch": 1.0653620741035066,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007147750074500845,
+      "loss": 0.6413,
+      "step": 21450
+    },
+    {
+      "epoch": 1.0658587463991258,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007147352736664349,
+      "loss": 0.5809,
+      "step": 21460
+    },
+    {
+      "epoch": 1.0663554186947453,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0007146955398827854,
+      "loss": 0.6338,
+      "step": 21470
+    },
+    {
+      "epoch": 1.0668520909903645,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007146558060991358,
+      "loss": 0.6367,
+      "step": 21480
+    },
+    {
+      "epoch": 1.067348763285984,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007146160723154862,
+      "loss": 0.6383,
+      "step": 21490
+    },
+    {
+      "epoch": 1.0678454355816032,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0007145763385318368,
+      "loss": 0.6074,
+      "step": 21500
+    },
+    {
+      "epoch": 1.0683421078772226,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007145366047481872,
+      "loss": 0.606,
+      "step": 21510
+    },
+    {
+      "epoch": 1.068838780172842,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0007144968709645376,
+      "loss": 0.641,
+      "step": 21520
+    },
+    {
+      "epoch": 1.0693354524684613,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007144571371808881,
+      "loss": 0.6228,
+      "step": 21530
+    },
+    {
+      "epoch": 1.0698321247640807,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0007144174033972385,
+      "loss": 0.6205,
+      "step": 21540
+    },
+    {
+      "epoch": 1.0703287970597,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.000714377669613589,
+      "loss": 0.6452,
+      "step": 21550
+    },
+    {
+      "epoch": 1.0708254693553194,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007143379358299395,
+      "loss": 0.638,
+      "step": 21560
+    },
+    {
+      "epoch": 1.0713221416509386,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0007142982020462899,
+      "loss": 0.6696,
+      "step": 21570
+    },
+    {
+      "epoch": 1.071818813946558,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0007142584682626403,
+      "loss": 0.6129,
+      "step": 21580
+    },
+    {
+      "epoch": 1.0723154862421773,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0007142187344789908,
+      "loss": 0.6459,
+      "step": 21590
+    },
+    {
+      "epoch": 1.0728121585377968,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0007141790006953413,
+      "loss": 0.6551,
+      "step": 21600
+    },
+    {
+      "epoch": 1.0733088308334162,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0007141392669116917,
+      "loss": 0.6482,
+      "step": 21610
+    },
+    {
+      "epoch": 1.0738055031290354,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007140995331280422,
+      "loss": 0.5979,
+      "step": 21620
+    },
+    {
+      "epoch": 1.074302175424655,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007140597993443926,
+      "loss": 0.6278,
+      "step": 21630
+    },
+    {
+      "epoch": 1.0747988477202741,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.000714020065560743,
+      "loss": 0.627,
+      "step": 21640
+    },
+    {
+      "epoch": 1.0752955200158936,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0007139803317770934,
+      "loss": 0.6193,
+      "step": 21650
+    },
+    {
+      "epoch": 1.0757921923115128,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000713940597993444,
+      "loss": 0.6542,
+      "step": 21660
+    },
+    {
+      "epoch": 1.0762888646071322,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0007139008642097945,
+      "loss": 0.6136,
+      "step": 21670
+    },
+    {
+      "epoch": 1.0767855369027515,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0007138611304261448,
+      "loss": 0.6423,
+      "step": 21680
+    },
+    {
+      "epoch": 1.077282209198371,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0007138213966424953,
+      "loss": 0.6457,
+      "step": 21690
+    },
+    {
+      "epoch": 1.0777788814939901,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007137816628588457,
+      "loss": 0.6406,
+      "step": 21700
+    },
+    {
+      "epoch": 1.0782755537896096,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007137419290751962,
+      "loss": 0.6145,
+      "step": 21710
+    },
+    {
+      "epoch": 1.078772226085229,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0007137021952915467,
+      "loss": 0.6083,
+      "step": 21720
+    },
+    {
+      "epoch": 1.0792688983808483,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0007136624615078971,
+      "loss": 0.6084,
+      "step": 21730
+    },
+    {
+      "epoch": 1.0797655706764677,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007136227277242475,
+      "loss": 0.6413,
+      "step": 21740
+    },
+    {
+      "epoch": 1.080262242972087,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.000713582993940598,
+      "loss": 0.6146,
+      "step": 21750
+    },
+    {
+      "epoch": 1.0807589152677064,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007135432601569485,
+      "loss": 0.6419,
+      "step": 21760
+    },
+    {
+      "epoch": 1.0812555875633256,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0007135035263732989,
+      "loss": 0.6509,
+      "step": 21770
+    },
+    {
+      "epoch": 1.081752259858945,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007134637925896494,
+      "loss": 0.6484,
+      "step": 21780
+    },
+    {
+      "epoch": 1.0822489321545645,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0007134240588059998,
+      "loss": 0.6472,
+      "step": 21790
+    },
+    {
+      "epoch": 1.0827456044501838,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0007133843250223504,
+      "loss": 0.6434,
+      "step": 21800
+    },
+    {
+      "epoch": 1.0832422767458032,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0007133445912387008,
+      "loss": 0.6067,
+      "step": 21810
+    },
+    {
+      "epoch": 1.0837389490414224,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0007133048574550512,
+      "loss": 0.6255,
+      "step": 21820
+    },
+    {
+      "epoch": 1.0842356213370419,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0007132651236714017,
+      "loss": 0.6374,
+      "step": 21830
+    },
+    {
+      "epoch": 1.084732293632661,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.000713225389887752,
+      "loss": 0.6203,
+      "step": 21840
+    },
+    {
+      "epoch": 1.0852289659282806,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007131856561041025,
+      "loss": 0.6495,
+      "step": 21850
+    },
+    {
+      "epoch": 1.0857256382238998,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007131459223204531,
+      "loss": 0.6453,
+      "step": 21860
+    },
+    {
+      "epoch": 1.0862223105195192,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0007131061885368034,
+      "loss": 0.5869,
+      "step": 21870
+    },
+    {
+      "epoch": 1.0867189828151385,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0007130664547531539,
+      "loss": 0.6145,
+      "step": 21880
+    },
+    {
+      "epoch": 1.087215655110758,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0007130267209695044,
+      "loss": 0.6281,
+      "step": 21890
+    },
+    {
+      "epoch": 1.0877123274063774,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0007129869871858547,
+      "loss": 0.6336,
+      "step": 21900
+    },
+    {
+      "epoch": 1.0882089997019966,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0007129472534022053,
+      "loss": 0.6367,
+      "step": 21910
+    },
+    {
+      "epoch": 1.088705671997616,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007129075196185557,
+      "loss": 0.6467,
+      "step": 21920
+    },
+    {
+      "epoch": 1.0892023442932353,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007128677858349061,
+      "loss": 0.633,
+      "step": 21930
+    },
+    {
+      "epoch": 1.0896990165888547,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0007128280520512566,
+      "loss": 0.6505,
+      "step": 21940
+    },
+    {
+      "epoch": 1.090195688884474,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.000712788318267607,
+      "loss": 0.6666,
+      "step": 21950
+    },
+    {
+      "epoch": 1.0906923611800934,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0007127485844839576,
+      "loss": 0.6298,
+      "step": 21960
+    },
+    {
+      "epoch": 1.0911890334757128,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.000712708850700308,
+      "loss": 0.6172,
+      "step": 21970
+    },
+    {
+      "epoch": 1.091685705771332,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0007126691169166584,
+      "loss": 0.6471,
+      "step": 21980
+    },
+    {
+      "epoch": 1.0921823780669515,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007126293831330089,
+      "loss": 0.649,
+      "step": 21990
+    },
+    {
+      "epoch": 1.0926790503625707,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007125896493493593,
+      "loss": 0.5827,
+      "step": 22000
+    },
+    {
+      "epoch": 1.0931757226581902,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0007125499155657098,
+      "loss": 0.6342,
+      "step": 22010
+    },
+    {
+      "epoch": 1.0936723949538094,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0007125101817820603,
+      "loss": 0.6023,
+      "step": 22020
+    },
+    {
+      "epoch": 1.0941690672494289,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0007124704479984106,
+      "loss": 0.6409,
+      "step": 22030
+    },
+    {
+      "epoch": 1.094665739545048,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007124307142147611,
+      "loss": 0.6335,
+      "step": 22040
+    },
+    {
+      "epoch": 1.0951624118406675,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0007123909804311117,
+      "loss": 0.6375,
+      "step": 22050
+    },
+    {
+      "epoch": 1.0956590841362868,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.000712351246647462,
+      "loss": 0.6516,
+      "step": 22060
+    },
+    {
+      "epoch": 1.0961557564319062,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007123115128638125,
+      "loss": 0.6722,
+      "step": 22070
+    },
+    {
+      "epoch": 1.0966524287275257,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.000712271779080163,
+      "loss": 0.6483,
+      "step": 22080
+    },
+    {
+      "epoch": 1.097149101023145,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0007122320452965133,
+      "loss": 0.6566,
+      "step": 22090
+    },
+    {
+      "epoch": 1.0976457733187643,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0007121923115128638,
+      "loss": 0.6415,
+      "step": 22100
+    },
+    {
+      "epoch": 1.0981424456143836,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0007121525777292143,
+      "loss": 0.6417,
+      "step": 22110
+    },
+    {
+      "epoch": 1.098639117910003,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0007121128439455648,
+      "loss": 0.6333,
+      "step": 22120
+    },
+    {
+      "epoch": 1.0991357902056222,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0007120731101619152,
+      "loss": 0.6129,
+      "step": 22130
+    },
+    {
+      "epoch": 1.0996324625012417,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0007120333763782656,
+      "loss": 0.5927,
+      "step": 22140
+    },
+    {
+      "epoch": 1.1001291347968611,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0007119936425946161,
+      "loss": 0.627,
+      "step": 22150
+    },
+    {
+      "epoch": 1.1006258070924804,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0007119539088109666,
+      "loss": 0.5961,
+      "step": 22160
+    },
+    {
+      "epoch": 1.1011224793880998,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.000711914175027317,
+      "loss": 0.6226,
+      "step": 22170
+    },
+    {
+      "epoch": 1.101619151683719,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007118744412436675,
+      "loss": 0.6391,
+      "step": 22180
+    },
+    {
+      "epoch": 1.1021158239793385,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007118347074600179,
+      "loss": 0.6212,
+      "step": 22190
+    },
+    {
+      "epoch": 1.1026124962749577,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0007117949736763683,
+      "loss": 0.6268,
+      "step": 22200
+    },
+    {
+      "epoch": 1.1031091685705772,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0007117552398927189,
+      "loss": 0.6589,
+      "step": 22210
+    },
+    {
+      "epoch": 1.1036058408661964,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0007117155061090692,
+      "loss": 0.6267,
+      "step": 22220
+    },
+    {
+      "epoch": 1.1041025131618158,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007116757723254197,
+      "loss": 0.613,
+      "step": 22230
+    },
+    {
+      "epoch": 1.104599185457435,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007116360385417702,
+      "loss": 0.6083,
+      "step": 22240
+    },
+    {
+      "epoch": 1.1050958577530545,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007115963047581206,
+      "loss": 0.613,
+      "step": 22250
+    },
+    {
+      "epoch": 1.105592530048674,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0007115565709744711,
+      "loss": 0.5983,
+      "step": 22260
+    },
+    {
+      "epoch": 1.1060892023442932,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0007115168371908216,
+      "loss": 0.6171,
+      "step": 22270
+    },
+    {
+      "epoch": 1.1065858746399126,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.000711477103407172,
+      "loss": 0.6404,
+      "step": 22280
+    },
+    {
+      "epoch": 1.1070825469355319,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007114373696235224,
+      "loss": 0.6187,
+      "step": 22290
+    },
+    {
+      "epoch": 1.1075792192311513,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0007113976358398728,
+      "loss": 0.6012,
+      "step": 22300
+    },
+    {
+      "epoch": 1.1080758915267706,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0007113579020562234,
+      "loss": 0.6123,
+      "step": 22310
+    },
+    {
+      "epoch": 1.10857256382239,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0007113181682725738,
+      "loss": 0.6281,
+      "step": 22320
+    },
+    {
+      "epoch": 1.1090692361180094,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0007112784344889242,
+      "loss": 0.6013,
+      "step": 22330
+    },
+    {
+      "epoch": 1.1095659084136287,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0007112387007052747,
+      "loss": 0.6682,
+      "step": 22340
+    },
+    {
+      "epoch": 1.1100625807092481,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0007111989669216251,
+      "loss": 0.6218,
+      "step": 22350
+    },
+    {
+      "epoch": 1.1105592530048674,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0007111592331379756,
+      "loss": 0.6293,
+      "step": 22360
+    },
+    {
+      "epoch": 1.1110559253004868,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0007111194993543261,
+      "loss": 0.6096,
+      "step": 22370
+    },
+    {
+      "epoch": 1.111552597596106,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0007110797655706765,
+      "loss": 0.628,
+      "step": 22380
+    },
+    {
+      "epoch": 1.1120492698917255,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007110400317870269,
+      "loss": 0.6113,
+      "step": 22390
+    },
+    {
+      "epoch": 1.1125459421873447,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0007110002980033774,
+      "loss": 0.5945,
+      "step": 22400
+    },
+    {
+      "epoch": 1.1130426144829642,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0007109605642197279,
+      "loss": 0.6667,
+      "step": 22410
+    },
+    {
+      "epoch": 1.1135392867785834,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0007109208304360783,
+      "loss": 0.6661,
+      "step": 22420
+    },
+    {
+      "epoch": 1.1140359590742028,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0007108810966524288,
+      "loss": 0.6505,
+      "step": 22430
+    },
+    {
+      "epoch": 1.1145326313698223,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0007108413628687792,
+      "loss": 0.6354,
+      "step": 22440
+    },
+    {
+      "epoch": 1.1150293036654415,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0007108016290851296,
+      "loss": 0.6554,
+      "step": 22450
+    },
+    {
+      "epoch": 1.115525975961061,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0007107618953014802,
+      "loss": 0.6206,
+      "step": 22460
+    },
+    {
+      "epoch": 1.1160226482566802,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007107221615178306,
+      "loss": 0.6328,
+      "step": 22470
+    },
+    {
+      "epoch": 1.1165193205522996,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.000710682427734181,
+      "loss": 0.5999,
+      "step": 22480
+    },
+    {
+      "epoch": 1.1170159928479189,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007106426939505314,
+      "loss": 0.6318,
+      "step": 22490
+    },
+    {
+      "epoch": 1.1175126651435383,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007106029601668819,
+      "loss": 0.6513,
+      "step": 22500
+    },
+    {
+      "epoch": 1.1180093374391578,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007105632263832324,
+      "loss": 0.6258,
+      "step": 22510
+    },
+    {
+      "epoch": 1.118506009734777,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007105234925995828,
+      "loss": 0.6303,
+      "step": 22520
+    },
+    {
+      "epoch": 1.1190026820303964,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0007104837588159333,
+      "loss": 0.709,
+      "step": 22530
+    },
+    {
+      "epoch": 1.1194993543260157,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007104440250322838,
+      "loss": 0.6219,
+      "step": 22540
+    },
+    {
+      "epoch": 1.119996026621635,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007104042912486341,
+      "loss": 0.6181,
+      "step": 22550
+    },
+    {
+      "epoch": 1.1204926989172543,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0007103645574649847,
+      "loss": 0.621,
+      "step": 22560
+    },
+    {
+      "epoch": 1.1209893712128738,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0007103248236813351,
+      "loss": 0.619,
+      "step": 22570
+    },
+    {
+      "epoch": 1.121486043508493,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0007102850898976855,
+      "loss": 0.619,
+      "step": 22580
+    },
+    {
+      "epoch": 1.1219827158041125,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000710245356114036,
+      "loss": 0.6185,
+      "step": 22590
+    },
+    {
+      "epoch": 1.1224793880997317,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0007102056223303864,
+      "loss": 0.6314,
+      "step": 22600
+    },
+    {
+      "epoch": 1.1229760603953511,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0007101658885467369,
+      "loss": 0.6068,
+      "step": 22610
+    },
+    {
+      "epoch": 1.1234727326909706,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0007101261547630874,
+      "loss": 0.6185,
+      "step": 22620
+    },
+    {
+      "epoch": 1.1239694049865898,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0007100864209794378,
+      "loss": 0.6095,
+      "step": 22630
+    },
+    {
+      "epoch": 1.1244660772822093,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007100466871957882,
+      "loss": 0.6314,
+      "step": 22640
+    },
+    {
+      "epoch": 1.1249627495778285,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007100069534121387,
+      "loss": 0.6239,
+      "step": 22650
+    },
+    {
+      "epoch": 1.125459421873448,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0007099672196284892,
+      "loss": 0.6411,
+      "step": 22660
+    },
+    {
+      "epoch": 1.1259560941690672,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007099274858448396,
+      "loss": 0.6537,
+      "step": 22670
+    },
+    {
+      "epoch": 1.1264527664646866,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0007098877520611901,
+      "loss": 0.6395,
+      "step": 22680
+    },
+    {
+      "epoch": 1.126949438760306,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0007098480182775405,
+      "loss": 0.6465,
+      "step": 22690
+    },
+    {
+      "epoch": 1.1274461110559253,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.000709808284493891,
+      "loss": 0.5975,
+      "step": 22700
+    },
+    {
+      "epoch": 1.1279427833515447,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0007097685507102414,
+      "loss": 0.6149,
+      "step": 22710
+    },
+    {
+      "epoch": 1.128439455647164,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0007097288169265919,
+      "loss": 0.625,
+      "step": 22720
+    },
+    {
+      "epoch": 1.1289361279427834,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007096890831429424,
+      "loss": 0.6185,
+      "step": 22730
+    },
+    {
+      "epoch": 1.1294328002384026,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0007096493493592927,
+      "loss": 0.6329,
+      "step": 22740
+    },
+    {
+      "epoch": 1.129929472534022,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0007096096155756432,
+      "loss": 0.6179,
+      "step": 22750
+    },
+    {
+      "epoch": 1.1304261448296413,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0007095698817919937,
+      "loss": 0.6177,
+      "step": 22760
+    },
+    {
+      "epoch": 1.1309228171252608,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007095301480083441,
+      "loss": 0.6089,
+      "step": 22770
+    },
+    {
+      "epoch": 1.13141948942088,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007094904142246946,
+      "loss": 0.6127,
+      "step": 22780
+    },
+    {
+      "epoch": 1.1319161617164994,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.000709450680441045,
+      "loss": 0.6463,
+      "step": 22790
+    },
+    {
+      "epoch": 1.132412834012119,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007094109466573954,
+      "loss": 0.6197,
+      "step": 22800
+    },
+    {
+      "epoch": 1.1329095063077381,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.000709371212873746,
+      "loss": 0.6144,
+      "step": 22810
+    },
+    {
+      "epoch": 1.1334061786033576,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0007093314790900964,
+      "loss": 0.6395,
+      "step": 22820
+    },
+    {
+      "epoch": 1.1339028508989768,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0007092917453064468,
+      "loss": 0.6231,
+      "step": 22830
+    },
+    {
+      "epoch": 1.1343995231945962,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0007092520115227973,
+      "loss": 0.6377,
+      "step": 22840
+    },
+    {
+      "epoch": 1.1348961954902155,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0007092122777391477,
+      "loss": 0.6589,
+      "step": 22850
+    },
+    {
+      "epoch": 1.135392867785835,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007091725439554983,
+      "loss": 0.6392,
+      "step": 22860
+    },
+    {
+      "epoch": 1.1358895400814544,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007091328101718487,
+      "loss": 0.6901,
+      "step": 22870
+    },
+    {
+      "epoch": 1.1363862123770736,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0007090930763881991,
+      "loss": 0.6257,
+      "step": 22880
+    },
+    {
+      "epoch": 1.136882884672693,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0007090533426045496,
+      "loss": 0.609,
+      "step": 22890
+    },
+    {
+      "epoch": 1.1373795569683123,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0007090136088208999,
+      "loss": 0.6255,
+      "step": 22900
+    },
+    {
+      "epoch": 1.1378762292639317,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0007089738750372505,
+      "loss": 0.6209,
+      "step": 22910
+    },
+    {
+      "epoch": 1.138372901559551,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.000708934141253601,
+      "loss": 0.6024,
+      "step": 22920
+    },
+    {
+      "epoch": 1.1388695738551704,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0007088944074699513,
+      "loss": 0.6234,
+      "step": 22930
+    },
+    {
+      "epoch": 1.1393662461507896,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0007088546736863018,
+      "loss": 0.635,
+      "step": 22940
+    },
+    {
+      "epoch": 1.139862918446409,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007088149399026523,
+      "loss": 0.6387,
+      "step": 22950
+    },
+    {
+      "epoch": 1.1403595907420283,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0007087752061190026,
+      "loss": 0.6197,
+      "step": 22960
+    },
+    {
+      "epoch": 1.1408562630376478,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0007087354723353532,
+      "loss": 0.6047,
+      "step": 22970
+    },
+    {
+      "epoch": 1.1413529353332672,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0007086957385517036,
+      "loss": 0.6384,
+      "step": 22980
+    },
+    {
+      "epoch": 1.1418496076288864,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0007086560047680541,
+      "loss": 0.6246,
+      "step": 22990
+    },
+    {
+      "epoch": 1.1423462799245059,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007086162709844045,
+      "loss": 0.6167,
+      "step": 23000
+    },
+    {
+      "epoch": 1.142842952220125,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.000708576537200755,
+      "loss": 0.6267,
+      "step": 23010
+    },
+    {
+      "epoch": 1.1433396245157446,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0007085368034171055,
+      "loss": 0.628,
+      "step": 23020
+    },
+    {
+      "epoch": 1.1438362968113638,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0007084970696334559,
+      "loss": 0.6131,
+      "step": 23030
+    },
+    {
+      "epoch": 1.1443329691069832,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0007084573358498063,
+      "loss": 0.6425,
+      "step": 23040
+    },
+    {
+      "epoch": 1.1448296414026027,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0007084176020661568,
+      "loss": 0.6234,
+      "step": 23050
+    },
+    {
+      "epoch": 1.145326313698222,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007083778682825073,
+      "loss": 0.6404,
+      "step": 23060
+    },
+    {
+      "epoch": 1.1458229859938414,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0007083381344988577,
+      "loss": 0.617,
+      "step": 23070
+    },
+    {
+      "epoch": 1.1463196582894606,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007082984007152082,
+      "loss": 0.5922,
+      "step": 23080
+    },
+    {
+      "epoch": 1.14681633058508,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0007082586669315585,
+      "loss": 0.618,
+      "step": 23090
+    },
+    {
+      "epoch": 1.1473130028806993,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.000708218933147909,
+      "loss": 0.613,
+      "step": 23100
+    },
+    {
+      "epoch": 1.1478096751763187,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0007081791993642596,
+      "loss": 0.6369,
+      "step": 23110
+    },
+    {
+      "epoch": 1.148306347471938,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0007081394655806099,
+      "loss": 0.6305,
+      "step": 23120
+    },
+    {
+      "epoch": 1.1488030197675574,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0007080997317969604,
+      "loss": 0.6211,
+      "step": 23130
+    },
+    {
+      "epoch": 1.1492996920631766,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007080599980133109,
+      "loss": 0.6008,
+      "step": 23140
+    },
+    {
+      "epoch": 1.149796364358796,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0007080202642296613,
+      "loss": 0.6009,
+      "step": 23150
+    },
+    {
+      "epoch": 1.1502930366544155,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0007079805304460117,
+      "loss": 0.6652,
+      "step": 23160
+    },
+    {
+      "epoch": 1.1507897089500347,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007079407966623622,
+      "loss": 0.6498,
+      "step": 23170
+    },
+    {
+      "epoch": 1.1512863812456542,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0007079010628787127,
+      "loss": 0.6265,
+      "step": 23180
+    },
+    {
+      "epoch": 1.1517830535412734,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007078613290950631,
+      "loss": 0.6222,
+      "step": 23190
+    },
+    {
+      "epoch": 1.1522797258368929,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0007078215953114135,
+      "loss": 0.6105,
+      "step": 23200
+    },
+    {
+      "epoch": 1.152776398132512,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.000707781861527764,
+      "loss": 0.6407,
+      "step": 23210
+    },
+    {
+      "epoch": 1.1532730704281315,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007077421277441145,
+      "loss": 0.6244,
+      "step": 23220
+    },
+    {
+      "epoch": 1.153769742723751,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0007077023939604649,
+      "loss": 0.6351,
+      "step": 23230
+    },
+    {
+      "epoch": 1.1542664150193702,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0007076626601768154,
+      "loss": 0.6386,
+      "step": 23240
+    },
+    {
+      "epoch": 1.1547630873149897,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007076229263931658,
+      "loss": 0.639,
+      "step": 23250
+    },
+    {
+      "epoch": 1.155259759610609,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0007075831926095162,
+      "loss": 0.6513,
+      "step": 23260
+    },
+    {
+      "epoch": 1.1557564319062283,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0007075434588258668,
+      "loss": 0.6354,
+      "step": 23270
+    },
+    {
+      "epoch": 1.1562531042018476,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007075037250422172,
+      "loss": 0.6214,
+      "step": 23280
+    },
+    {
+      "epoch": 1.156749776497467,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007074639912585676,
+      "loss": 0.6353,
+      "step": 23290
+    },
+    {
+      "epoch": 1.1572464487930862,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0007074242574749181,
+      "loss": 0.6349,
+      "step": 23300
+    },
+    {
+      "epoch": 1.1577431210887057,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007073845236912685,
+      "loss": 0.6332,
+      "step": 23310
+    },
+    {
+      "epoch": 1.158239793384325,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.000707344789907619,
+      "loss": 0.6123,
+      "step": 23320
+    },
+    {
+      "epoch": 1.1587364656799444,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0007073050561239695,
+      "loss": 0.6024,
+      "step": 23330
+    },
+    {
+      "epoch": 1.1592331379755638,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007072653223403199,
+      "loss": 0.6049,
+      "step": 23340
+    },
+    {
+      "epoch": 1.159729810271183,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0007072255885566703,
+      "loss": 0.6323,
+      "step": 23350
+    },
+    {
+      "epoch": 1.1602264825668025,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007071858547730207,
+      "loss": 0.643,
+      "step": 23360
+    },
+    {
+      "epoch": 1.1607231548624217,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0007071461209893713,
+      "loss": 0.6096,
+      "step": 23370
+    },
+    {
+      "epoch": 1.1612198271580412,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007071063872057217,
+      "loss": 0.6407,
+      "step": 23380
+    },
+    {
+      "epoch": 1.1617164994536604,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007070666534220721,
+      "loss": 0.5905,
+      "step": 23390
+    },
+    {
+      "epoch": 1.1622131717492798,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007070269196384226,
+      "loss": 0.5927,
+      "step": 23400
+    },
+    {
+      "epoch": 1.1627098440448993,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.000706987185854773,
+      "loss": 0.6414,
+      "step": 23410
+    },
+    {
+      "epoch": 1.1632065163405185,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0007069474520711235,
+      "loss": 0.6149,
+      "step": 23420
+    },
+    {
+      "epoch": 1.1637031886361378,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.000706907718287474,
+      "loss": 0.6216,
+      "step": 23430
+    },
+    {
+      "epoch": 1.1641998609317572,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0007068679845038244,
+      "loss": 0.6228,
+      "step": 23440
+    },
+    {
+      "epoch": 1.1646965332273767,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0007068282507201748,
+      "loss": 0.6317,
+      "step": 23450
+    },
+    {
+      "epoch": 1.1651932055229959,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0007067885169365253,
+      "loss": 0.6172,
+      "step": 23460
+    },
+    {
+      "epoch": 1.1656898778186153,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0007067487831528758,
+      "loss": 0.6362,
+      "step": 23470
+    },
+    {
+      "epoch": 1.1661865501142346,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007067090493692262,
+      "loss": 0.6011,
+      "step": 23480
+    },
+    {
+      "epoch": 1.166683222409854,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0007066693155855767,
+      "loss": 0.6261,
+      "step": 23490
+    },
+    {
+      "epoch": 1.1671798947054732,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0007066295818019271,
+      "loss": 0.6458,
+      "step": 23500
+    },
+    {
+      "epoch": 1.1676765670010927,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0007065898480182775,
+      "loss": 0.6441,
+      "step": 23510
+    },
+    {
+      "epoch": 1.1681732392967121,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0007065501142346281,
+      "loss": 0.602,
+      "step": 23520
+    },
+    {
+      "epoch": 1.1686699115923314,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0007065103804509785,
+      "loss": 0.6454,
+      "step": 23530
+    },
+    {
+      "epoch": 1.1691665838879508,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007064706466673289,
+      "loss": 0.6184,
+      "step": 23540
+    },
+    {
+      "epoch": 1.16966325618357,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0007064309128836794,
+      "loss": 0.6447,
+      "step": 23550
+    },
+    {
+      "epoch": 1.1701599284791895,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0007063911791000298,
+      "loss": 0.6074,
+      "step": 23560
+    },
+    {
+      "epoch": 1.1706566007748087,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0007063514453163803,
+      "loss": 0.6108,
+      "step": 23570
+    },
+    {
+      "epoch": 1.1711532730704282,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0007063117115327307,
+      "loss": 0.6396,
+      "step": 23580
+    },
+    {
+      "epoch": 1.1716499453660474,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007062719777490812,
+      "loss": 0.6014,
+      "step": 23590
+    },
+    {
+      "epoch": 1.1721466176616668,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0007062322439654317,
+      "loss": 0.6143,
+      "step": 23600
+    },
+    {
+      "epoch": 1.172643289957286,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.000706192510181782,
+      "loss": 0.5961,
+      "step": 23610
+    },
+    {
+      "epoch": 1.1731399622529055,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0007061527763981326,
+      "loss": 0.6356,
+      "step": 23620
+    },
+    {
+      "epoch": 1.173636634548525,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.000706113042614483,
+      "loss": 0.6159,
+      "step": 23630
+    },
+    {
+      "epoch": 1.1741333068441442,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007060733088308334,
+      "loss": 0.6085,
+      "step": 23640
+    },
+    {
+      "epoch": 1.1746299791397636,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007060335750471839,
+      "loss": 0.5906,
+      "step": 23650
+    },
+    {
+      "epoch": 1.1751266514353829,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0007059938412635343,
+      "loss": 0.6355,
+      "step": 23660
+    },
+    {
+      "epoch": 1.1756233237310023,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0007059541074798848,
+      "loss": 0.6162,
+      "step": 23670
+    },
+    {
+      "epoch": 1.1761199960266215,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0007059143736962353,
+      "loss": 0.616,
+      "step": 23680
+    },
+    {
+      "epoch": 1.176616668322241,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0007058746399125857,
+      "loss": 0.5947,
+      "step": 23690
+    },
+    {
+      "epoch": 1.1771133406178604,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0007058349061289361,
+      "loss": 0.6463,
+      "step": 23700
+    },
+    {
+      "epoch": 1.1776100129134797,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0007057951723452866,
+      "loss": 0.6078,
+      "step": 23710
+    },
+    {
+      "epoch": 1.1781066852090991,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0007057554385616371,
+      "loss": 0.6306,
+      "step": 23720
+    },
+    {
+      "epoch": 1.1786033575047183,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0007057157047779875,
+      "loss": 0.6252,
+      "step": 23730
+    },
+    {
+      "epoch": 1.1791000298003378,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.000705675970994338,
+      "loss": 0.6268,
+      "step": 23740
+    },
+    {
+      "epoch": 1.179596702095957,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007056362372106884,
+      "loss": 0.6547,
+      "step": 23750
+    },
+    {
+      "epoch": 1.1800933743915765,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007055965034270389,
+      "loss": 0.5791,
+      "step": 23760
+    },
+    {
+      "epoch": 1.1805900466871957,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0007055567696433893,
+      "loss": 0.6184,
+      "step": 23770
+    },
+    {
+      "epoch": 1.1810867189828151,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0007055170358597398,
+      "loss": 0.6027,
+      "step": 23780
+    },
+    {
+      "epoch": 1.1815833912784344,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007054773020760903,
+      "loss": 0.6148,
+      "step": 23790
+    },
+    {
+      "epoch": 1.1820800635740538,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0007054375682924406,
+      "loss": 0.6185,
+      "step": 23800
+    },
+    {
+      "epoch": 1.1825767358696733,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007053978345087911,
+      "loss": 0.6371,
+      "step": 23810
+    },
+    {
+      "epoch": 1.1830734081652925,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0007053581007251417,
+      "loss": 0.6398,
+      "step": 23820
+    },
+    {
+      "epoch": 1.183570080460912,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.000705318366941492,
+      "loss": 0.6341,
+      "step": 23830
+    },
+    {
+      "epoch": 1.1840667527565312,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0007052786331578425,
+      "loss": 0.6218,
+      "step": 23840
+    },
+    {
+      "epoch": 1.1845634250521506,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007052388993741929,
+      "loss": 0.6058,
+      "step": 23850
+    },
+    {
+      "epoch": 1.1850600973477698,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0007051991655905433,
+      "loss": 0.6475,
+      "step": 23860
+    },
+    {
+      "epoch": 1.1855567696433893,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007051594318068939,
+      "loss": 0.6322,
+      "step": 23870
+    },
+    {
+      "epoch": 1.1860534419390087,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0007051196980232443,
+      "loss": 0.5993,
+      "step": 23880
+    },
+    {
+      "epoch": 1.186550114234628,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007050799642395948,
+      "loss": 0.6268,
+      "step": 23890
+    },
+    {
+      "epoch": 1.1870467865302474,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0007050402304559452,
+      "loss": 0.5902,
+      "step": 23900
+    },
+    {
+      "epoch": 1.1875434588258666,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0007050004966722956,
+      "loss": 0.6241,
+      "step": 23910
+    },
+    {
+      "epoch": 1.188040131121486,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0007049607628886462,
+      "loss": 0.6016,
+      "step": 23920
+    },
+    {
+      "epoch": 1.1885368034171053,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0007049210291049966,
+      "loss": 0.6004,
+      "step": 23930
+    },
+    {
+      "epoch": 1.1890334757127248,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.000704881295321347,
+      "loss": 0.6316,
+      "step": 23940
+    },
+    {
+      "epoch": 1.189530148008344,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007048415615376975,
+      "loss": 0.6467,
+      "step": 23950
+    },
+    {
+      "epoch": 1.1900268203039635,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007048018277540478,
+      "loss": 0.6248,
+      "step": 23960
+    },
+    {
+      "epoch": 1.1905234925995827,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007047620939703984,
+      "loss": 0.6385,
+      "step": 23970
+    },
+    {
+      "epoch": 1.1910201648952021,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0007047223601867489,
+      "loss": 0.6219,
+      "step": 23980
+    },
+    {
+      "epoch": 1.1915168371908216,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0007046826264030992,
+      "loss": 0.6193,
+      "step": 23990
+    },
+    {
+      "epoch": 1.1920135094864408,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0007046428926194497,
+      "loss": 0.6297,
+      "step": 24000
+    },
+    {
+      "epoch": 1.1925101817820603,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0007046031588358002,
+      "loss": 0.6158,
+      "step": 24010
+    },
+    {
+      "epoch": 1.1930068540776795,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0007045634250521506,
+      "loss": 0.6023,
+      "step": 24020
+    },
+    {
+      "epoch": 1.193503526373299,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0007045236912685011,
+      "loss": 0.6306,
+      "step": 24030
+    },
+    {
+      "epoch": 1.1940001986689182,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0007044839574848515,
+      "loss": 0.6125,
+      "step": 24040
+    },
+    {
+      "epoch": 1.1944968709645376,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.000704444223701202,
+      "loss": 0.5975,
+      "step": 24050
+    },
+    {
+      "epoch": 1.194993543260157,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007044044899175524,
+      "loss": 0.6436,
+      "step": 24060
+    },
+    {
+      "epoch": 1.1954902155557763,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007043647561339029,
+      "loss": 0.6371,
+      "step": 24070
+    },
+    {
+      "epoch": 1.1959868878513957,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007043250223502534,
+      "loss": 0.6063,
+      "step": 24080
+    },
+    {
+      "epoch": 1.196483560147015,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0007042852885666038,
+      "loss": 0.6436,
+      "step": 24090
+    },
+    {
+      "epoch": 1.1969802324426344,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0007042455547829542,
+      "loss": 0.6478,
+      "step": 24100
+    },
+    {
+      "epoch": 1.1974769047382536,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0007042058209993047,
+      "loss": 0.6237,
+      "step": 24110
+    },
+    {
+      "epoch": 1.197973577033873,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0007041660872156552,
+      "loss": 0.5966,
+      "step": 24120
+    },
+    {
+      "epoch": 1.1984702493294923,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0007041263534320056,
+      "loss": 0.6514,
+      "step": 24130
+    },
+    {
+      "epoch": 1.1989669216251118,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0007040866196483561,
+      "loss": 0.6267,
+      "step": 24140
+    },
+    {
+      "epoch": 1.199463593920731,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0007040468858647065,
+      "loss": 0.6382,
+      "step": 24150
+    },
+    {
+      "epoch": 1.1999602662163504,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007040071520810569,
+      "loss": 0.6262,
+      "step": 24160
+    },
+    {
+      "epoch": 1.2004569385119699,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0007039674182974075,
+      "loss": 0.6151,
+      "step": 24170
+    },
+    {
+      "epoch": 1.2009536108075891,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007039276845137578,
+      "loss": 0.6319,
+      "step": 24180
+    },
+    {
+      "epoch": 1.2014502831032086,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0007038879507301083,
+      "loss": 0.6521,
+      "step": 24190
+    },
+    {
+      "epoch": 1.2019469553988278,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0007038482169464588,
+      "loss": 0.5974,
+      "step": 24200
+    },
+    {
+      "epoch": 1.2024436276944472,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0007038084831628092,
+      "loss": 0.655,
+      "step": 24210
+    },
+    {
+      "epoch": 1.2029402999900665,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0007037687493791597,
+      "loss": 0.6258,
+      "step": 24220
+    },
+    {
+      "epoch": 1.203436972285686,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0007037290155955101,
+      "loss": 0.6048,
+      "step": 24230
+    },
+    {
+      "epoch": 1.2039336445813054,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0007036892818118606,
+      "loss": 0.6375,
+      "step": 24240
+    },
+    {
+      "epoch": 1.2044303168769246,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.000703649548028211,
+      "loss": 0.6155,
+      "step": 24250
+    },
+    {
+      "epoch": 1.204926989172544,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007036098142445614,
+      "loss": 0.595,
+      "step": 24260
+    },
+    {
+      "epoch": 1.2054236614681633,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.000703570080460912,
+      "loss": 0.605,
+      "step": 24270
+    },
+    {
+      "epoch": 1.2059203337637827,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0007035303466772624,
+      "loss": 0.6043,
+      "step": 24280
+    },
+    {
+      "epoch": 1.206417006059402,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0007034906128936128,
+      "loss": 0.6582,
+      "step": 24290
+    },
+    {
+      "epoch": 1.2069136783550214,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0007034508791099633,
+      "loss": 0.6373,
+      "step": 24300
+    },
+    {
+      "epoch": 1.2074103506506406,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0007034111453263137,
+      "loss": 0.6126,
+      "step": 24310
+    },
+    {
+      "epoch": 1.20790702294626,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007033714115426641,
+      "loss": 0.6255,
+      "step": 24320
+    },
+    {
+      "epoch": 1.2084036952418793,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0007033316777590147,
+      "loss": 0.6219,
+      "step": 24330
+    },
+    {
+      "epoch": 1.2089003675374987,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0007032919439753651,
+      "loss": 0.6258,
+      "step": 24340
+    },
+    {
+      "epoch": 1.2093970398331182,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0007032522101917155,
+      "loss": 0.6137,
+      "step": 24350
+    },
+    {
+      "epoch": 1.2098937121287374,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.000703212476408066,
+      "loss": 0.5958,
+      "step": 24360
+    },
+    {
+      "epoch": 1.2103903844243569,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007031727426244165,
+      "loss": 0.6078,
+      "step": 24370
+    },
+    {
+      "epoch": 1.210887056719976,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007031330088407669,
+      "loss": 0.6212,
+      "step": 24380
+    },
+    {
+      "epoch": 1.2113837290155955,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0007030932750571174,
+      "loss": 0.5888,
+      "step": 24390
+    },
+    {
+      "epoch": 1.2118804013112148,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0007030535412734678,
+      "loss": 0.611,
+      "step": 24400
+    },
+    {
+      "epoch": 1.2123770736068342,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0007030138074898182,
+      "loss": 0.5952,
+      "step": 24410
+    },
+    {
+      "epoch": 1.2128737459024537,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007029740737061688,
+      "loss": 0.6439,
+      "step": 24420
+    },
+    {
+      "epoch": 1.213370418198073,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0007029343399225192,
+      "loss": 0.6053,
+      "step": 24430
+    },
+    {
+      "epoch": 1.2138670904936923,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0007028946061388696,
+      "loss": 0.6181,
+      "step": 24440
+    },
+    {
+      "epoch": 1.2143637627893116,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00070285487235522,
+      "loss": 0.6271,
+      "step": 24450
+    },
+    {
+      "epoch": 1.214860435084931,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0007028151385715705,
+      "loss": 0.6332,
+      "step": 24460
+    },
+    {
+      "epoch": 1.2153571073805503,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.000702775404787921,
+      "loss": 0.6302,
+      "step": 24470
+    },
+    {
+      "epoch": 1.2158537796761697,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007027356710042714,
+      "loss": 0.5947,
+      "step": 24480
+    },
+    {
+      "epoch": 1.216350451971789,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0007026959372206219,
+      "loss": 0.645,
+      "step": 24490
+    },
+    {
+      "epoch": 1.2168471242674084,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007026562034369723,
+      "loss": 0.6371,
+      "step": 24500
+    },
+    {
+      "epoch": 1.2173437965630276,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0007026164696533227,
+      "loss": 0.6097,
+      "step": 24510
+    },
+    {
+      "epoch": 1.217840468858647,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0007025767358696733,
+      "loss": 0.6174,
+      "step": 24520
+    },
+    {
+      "epoch": 1.2183371411542665,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0007025370020860237,
+      "loss": 0.6005,
+      "step": 24530
+    },
+    {
+      "epoch": 1.2188338134498857,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0007024972683023741,
+      "loss": 0.59,
+      "step": 24540
+    },
+    {
+      "epoch": 1.2193304857455052,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0007024575345187246,
+      "loss": 0.6203,
+      "step": 24550
+    },
+    {
+      "epoch": 1.2198271580411244,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.000702417800735075,
+      "loss": 0.6041,
+      "step": 24560
+    },
+    {
+      "epoch": 1.2203238303367439,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0007023780669514254,
+      "loss": 0.6207,
+      "step": 24570
+    },
+    {
+      "epoch": 1.220820502632363,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.000702338333167776,
+      "loss": 0.6145,
+      "step": 24580
+    },
+    {
+      "epoch": 1.2213171749279825,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007022985993841264,
+      "loss": 0.6016,
+      "step": 24590
+    },
+    {
+      "epoch": 1.221813847223602,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0007022588656004768,
+      "loss": 0.615,
+      "step": 24600
+    },
+    {
+      "epoch": 1.2223105195192212,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007022191318168273,
+      "loss": 0.6351,
+      "step": 24610
+    },
+    {
+      "epoch": 1.2228071918148407,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0007021793980331777,
+      "loss": 0.6276,
+      "step": 24620
+    },
+    {
+      "epoch": 1.2233038641104599,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0007021396642495282,
+      "loss": 0.6095,
+      "step": 24630
+    },
+    {
+      "epoch": 1.2238005364060793,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0007020999304658786,
+      "loss": 0.6215,
+      "step": 24640
+    },
+    {
+      "epoch": 1.2242972087016986,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0007020601966822291,
+      "loss": 0.5976,
+      "step": 24650
+    },
+    {
+      "epoch": 1.224793880997318,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0007020204628985796,
+      "loss": 0.5812,
+      "step": 24660
+    },
+    {
+      "epoch": 1.2252905532929372,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0007019807291149299,
+      "loss": 0.6122,
+      "step": 24670
+    },
+    {
+      "epoch": 1.2257872255885567,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0007019409953312805,
+      "loss": 0.6131,
+      "step": 24680
+    },
+    {
+      "epoch": 1.226283897884176,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.000701901261547631,
+      "loss": 0.6229,
+      "step": 24690
+    },
+    {
+      "epoch": 1.2267805701797954,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0007018615277639813,
+      "loss": 0.594,
+      "step": 24700
+    },
+    {
+      "epoch": 1.2272772424754148,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0007018217939803318,
+      "loss": 0.5905,
+      "step": 24710
+    },
+    {
+      "epoch": 1.227773914771034,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0007017820601966822,
+      "loss": 0.6067,
+      "step": 24720
+    },
+    {
+      "epoch": 1.2282705870666535,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007017423264130327,
+      "loss": 0.6426,
+      "step": 24730
+    },
+    {
+      "epoch": 1.2287672593622727,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007017025926293832,
+      "loss": 0.6224,
+      "step": 24740
+    },
+    {
+      "epoch": 1.2292639316578922,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0007016628588457336,
+      "loss": 0.6083,
+      "step": 24750
+    },
+    {
+      "epoch": 1.2297606039535114,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.000701623125062084,
+      "loss": 0.6234,
+      "step": 24760
+    },
+    {
+      "epoch": 1.2302572762491308,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0007015833912784345,
+      "loss": 0.6315,
+      "step": 24770
+    },
+    {
+      "epoch": 1.2307539485447503,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.000701543657494785,
+      "loss": 0.6274,
+      "step": 24780
+    },
+    {
+      "epoch": 1.2312506208403695,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0007015039237111355,
+      "loss": 0.6201,
+      "step": 24790
+    },
+    {
+      "epoch": 1.231747293135989,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0007014641899274859,
+      "loss": 0.6321,
+      "step": 24800
+    },
+    {
+      "epoch": 1.2322439654316082,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0007014244561438363,
+      "loss": 0.6348,
+      "step": 24810
+    },
+    {
+      "epoch": 1.2327406377272276,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0007013847223601868,
+      "loss": 0.5963,
+      "step": 24820
+    },
+    {
+      "epoch": 1.2332373100228469,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0007013449885765372,
+      "loss": 0.6164,
+      "step": 24830
+    },
+    {
+      "epoch": 1.2337339823184663,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007013052547928877,
+      "loss": 0.6486,
+      "step": 24840
+    },
+    {
+      "epoch": 1.2342306546140855,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0007012655210092382,
+      "loss": 0.6338,
+      "step": 24850
+    },
+    {
+      "epoch": 1.234727326909705,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0007012257872255885,
+      "loss": 0.6248,
+      "step": 24860
+    },
+    {
+      "epoch": 1.2352239992053242,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.000701186053441939,
+      "loss": 0.622,
+      "step": 24870
+    },
+    {
+      "epoch": 1.2357206715009437,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0007011463196582896,
+      "loss": 0.5822,
+      "step": 24880
+    },
+    {
+      "epoch": 1.2362173437965631,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0007011065858746399,
+      "loss": 0.61,
+      "step": 24890
+    },
+    {
+      "epoch": 1.2367140160921823,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007010668520909904,
+      "loss": 0.6277,
+      "step": 24900
+    },
+    {
+      "epoch": 1.2372106883878018,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0007010271183073408,
+      "loss": 0.6312,
+      "step": 24910
+    },
+    {
+      "epoch": 1.237707360683421,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0007009873845236912,
+      "loss": 0.5847,
+      "step": 24920
+    },
+    {
+      "epoch": 1.2382040329790405,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0007009476507400418,
+      "loss": 0.6328,
+      "step": 24930
+    },
+    {
+      "epoch": 1.2387007052746597,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0007009079169563922,
+      "loss": 0.6421,
+      "step": 24940
+    },
+    {
+      "epoch": 1.2391973775702791,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0007008681831727427,
+      "loss": 0.6063,
+      "step": 24950
+    },
+    {
+      "epoch": 1.2396940498658986,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0007008284493890931,
+      "loss": 0.6355,
+      "step": 24960
+    },
+    {
+      "epoch": 1.2401907221615178,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0007007887156054435,
+      "loss": 0.6116,
+      "step": 24970
+    },
+    {
+      "epoch": 1.2406873944571373,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0007007489818217941,
+      "loss": 0.6389,
+      "step": 24980
+    },
+    {
+      "epoch": 1.2411840667527565,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0007007092480381445,
+      "loss": 0.641,
+      "step": 24990
+    },
+    {
+      "epoch": 1.241680739048376,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0007006695142544949,
+      "loss": 0.6491,
+      "step": 25000
+    },
+    {
+      "epoch": 1.2421774113439952,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007006297804708454,
+      "loss": 0.6464,
+      "step": 25010
+    },
+    {
+      "epoch": 1.2426740836396146,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0007005900466871958,
+      "loss": 0.6064,
+      "step": 25020
+    },
+    {
+      "epoch": 1.2431707559352339,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0007005503129035463,
+      "loss": 0.6449,
+      "step": 25030
+    },
+    {
+      "epoch": 1.2436674282308533,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0007005105791198968,
+      "loss": 0.6289,
+      "step": 25040
+    },
+    {
+      "epoch": 1.2441641005264725,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0007004708453362471,
+      "loss": 0.6168,
+      "step": 25050
+    },
+    {
+      "epoch": 1.244660772822092,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0007004311115525976,
+      "loss": 0.6012,
+      "step": 25060
+    },
+    {
+      "epoch": 1.2451574451177114,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0007003913777689481,
+      "loss": 0.6305,
+      "step": 25070
+    },
+    {
+      "epoch": 1.2456541174133307,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0007003516439852986,
+      "loss": 0.6238,
+      "step": 25080
+    },
+    {
+      "epoch": 1.24615078970895,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.000700311910201649,
+      "loss": 0.6106,
+      "step": 25090
+    },
+    {
+      "epoch": 1.2466474620045693,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0007002721764179994,
+      "loss": 0.6166,
+      "step": 25100
+    },
+    {
+      "epoch": 1.2471441343001888,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0007002324426343499,
+      "loss": 0.6587,
+      "step": 25110
+    },
+    {
+      "epoch": 1.247640806595808,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0007001927088507003,
+      "loss": 0.632,
+      "step": 25120
+    },
+    {
+      "epoch": 1.2481374788914275,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0007001529750670508,
+      "loss": 0.6218,
+      "step": 25130
+    },
+    {
+      "epoch": 1.248634151187047,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0007001132412834013,
+      "loss": 0.605,
+      "step": 25140
+    },
+    {
+      "epoch": 1.2491308234826661,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0007000735074997517,
+      "loss": 0.6058,
+      "step": 25150
+    },
+    {
+      "epoch": 1.2496274957782856,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0007000337737161021,
+      "loss": 0.6327,
+      "step": 25160
+    },
+    {
+      "epoch": 1.2501241680739048,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0006999940399324526,
+      "loss": 0.6189,
+      "step": 25170
+    },
+    {
+      "epoch": 1.2506208403695243,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0006999543061488031,
+      "loss": 0.6172,
+      "step": 25180
+    },
+    {
+      "epoch": 1.2511175126651435,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006999145723651535,
+      "loss": 0.6432,
+      "step": 25190
+    },
+    {
+      "epoch": 1.251614184960763,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.000699874838581504,
+      "loss": 0.6251,
+      "step": 25200
+    },
+    {
+      "epoch": 1.2521108572563822,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0006998351047978544,
+      "loss": 0.6038,
+      "step": 25210
+    },
+    {
+      "epoch": 1.2526075295520016,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006997953710142048,
+      "loss": 0.6242,
+      "step": 25220
+    },
+    {
+      "epoch": 1.2531042018476208,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006997556372305554,
+      "loss": 0.612,
+      "step": 25230
+    },
+    {
+      "epoch": 1.2536008741432403,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006997159034469058,
+      "loss": 0.6356,
+      "step": 25240
+    },
+    {
+      "epoch": 1.2540975464388597,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006996761696632562,
+      "loss": 0.5935,
+      "step": 25250
+    },
+    {
+      "epoch": 1.254594218734479,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0006996364358796067,
+      "loss": 0.5989,
+      "step": 25260
+    },
+    {
+      "epoch": 1.2550908910300984,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0006995967020959571,
+      "loss": 0.615,
+      "step": 25270
+    },
+    {
+      "epoch": 1.2555875633257176,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006995569683123076,
+      "loss": 0.634,
+      "step": 25280
+    },
+    {
+      "epoch": 1.256084235621337,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006995172345286581,
+      "loss": 0.6228,
+      "step": 25290
+    },
+    {
+      "epoch": 1.2565809079169563,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006994775007450085,
+      "loss": 0.6097,
+      "step": 25300
+    },
+    {
+      "epoch": 1.2570775802125758,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006994377669613589,
+      "loss": 0.6059,
+      "step": 25310
+    },
+    {
+      "epoch": 1.2575742525081952,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006993980331777093,
+      "loss": 0.6045,
+      "step": 25320
+    },
+    {
+      "epoch": 1.2580709248038144,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006993582993940599,
+      "loss": 0.6311,
+      "step": 25330
+    },
+    {
+      "epoch": 1.2585675970994337,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006993185656104103,
+      "loss": 0.5906,
+      "step": 25340
+    },
+    {
+      "epoch": 1.2590642693950531,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0006992788318267607,
+      "loss": 0.6082,
+      "step": 25350
+    },
+    {
+      "epoch": 1.2595609416906726,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006992390980431112,
+      "loss": 0.5942,
+      "step": 25360
+    },
+    {
+      "epoch": 1.2600576139862918,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0006991993642594616,
+      "loss": 0.606,
+      "step": 25370
+    },
+    {
+      "epoch": 1.2605542862819112,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.000699159630475812,
+      "loss": 0.6422,
+      "step": 25380
+    },
+    {
+      "epoch": 1.2610509585775305,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0006991198966921626,
+      "loss": 0.6,
+      "step": 25390
+    },
+    {
+      "epoch": 1.26154763087315,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.000699080162908513,
+      "loss": 0.6161,
+      "step": 25400
+    },
+    {
+      "epoch": 1.2620443031687691,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006990404291248634,
+      "loss": 0.5868,
+      "step": 25410
+    },
+    {
+      "epoch": 1.2625409754643886,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0006990006953412139,
+      "loss": 0.5998,
+      "step": 25420
+    },
+    {
+      "epoch": 1.263037647760008,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0006989609615575644,
+      "loss": 0.5932,
+      "step": 25430
+    },
+    {
+      "epoch": 1.2635343200556273,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006989212277739148,
+      "loss": 0.6253,
+      "step": 25440
+    },
+    {
+      "epoch": 1.2640309923512467,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006988814939902653,
+      "loss": 0.6573,
+      "step": 25450
+    },
+    {
+      "epoch": 1.264527664646866,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006988417602066157,
+      "loss": 0.6146,
+      "step": 25460
+    },
+    {
+      "epoch": 1.2650243369424854,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0006988020264229661,
+      "loss": 0.6223,
+      "step": 25470
+    },
+    {
+      "epoch": 1.2655210092381046,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006987622926393167,
+      "loss": 0.6207,
+      "step": 25480
+    },
+    {
+      "epoch": 1.266017681533724,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006987225588556671,
+      "loss": 0.6174,
+      "step": 25490
+    },
+    {
+      "epoch": 1.2665143538293435,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006986828250720175,
+      "loss": 0.6209,
+      "step": 25500
+    },
+    {
+      "epoch": 1.2670110261249627,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0006986430912883679,
+      "loss": 0.6172,
+      "step": 25510
+    },
+    {
+      "epoch": 1.267507698420582,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006986033575047184,
+      "loss": 0.61,
+      "step": 25520
+    },
+    {
+      "epoch": 1.2680043707162014,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.000698563623721069,
+      "loss": 0.6227,
+      "step": 25530
+    },
+    {
+      "epoch": 1.2685010430118209,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006985238899374193,
+      "loss": 0.6064,
+      "step": 25540
+    },
+    {
+      "epoch": 1.26899771530744,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006984841561537698,
+      "loss": 0.6189,
+      "step": 25550
+    },
+    {
+      "epoch": 1.2694943876030595,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006984444223701202,
+      "loss": 0.6138,
+      "step": 25560
+    },
+    {
+      "epoch": 1.2699910598986788,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0006984046885864706,
+      "loss": 0.6371,
+      "step": 25570
+    },
+    {
+      "epoch": 1.2704877321942982,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006983649548028212,
+      "loss": 0.6274,
+      "step": 25580
+    },
+    {
+      "epoch": 1.2709844044899175,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0006983252210191716,
+      "loss": 0.613,
+      "step": 25590
+    },
+    {
+      "epoch": 1.271481076785537,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000698285487235522,
+      "loss": 0.5777,
+      "step": 25600
+    },
+    {
+      "epoch": 1.2719777490811564,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006982457534518725,
+      "loss": 0.6199,
+      "step": 25610
+    },
+    {
+      "epoch": 1.2724744213767756,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0006982060196682229,
+      "loss": 0.6128,
+      "step": 25620
+    },
+    {
+      "epoch": 1.272971093672395,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006981662858845733,
+      "loss": 0.6039,
+      "step": 25630
+    },
+    {
+      "epoch": 1.2734677659680143,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0006981265521009239,
+      "loss": 0.5967,
+      "step": 25640
+    },
+    {
+      "epoch": 1.2739644382636337,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006980868183172743,
+      "loss": 0.6332,
+      "step": 25650
+    },
+    {
+      "epoch": 1.274461110559253,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0006980470845336247,
+      "loss": 0.6355,
+      "step": 25660
+    },
+    {
+      "epoch": 1.2749577828548724,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0006980073507499752,
+      "loss": 0.6018,
+      "step": 25670
+    },
+    {
+      "epoch": 1.2754544551504918,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0006979676169663257,
+      "loss": 0.618,
+      "step": 25680
+    },
+    {
+      "epoch": 1.275951127446111,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0006979278831826762,
+      "loss": 0.609,
+      "step": 25690
+    },
+    {
+      "epoch": 1.2764477997417303,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006978881493990265,
+      "loss": 0.631,
+      "step": 25700
+    },
+    {
+      "epoch": 1.2769444720373497,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.000697848415615377,
+      "loss": 0.6446,
+      "step": 25710
+    },
+    {
+      "epoch": 1.2774411443329692,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0006978086818317275,
+      "loss": 0.6415,
+      "step": 25720
+    },
+    {
+      "epoch": 1.2779378166285884,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0006977689480480778,
+      "loss": 0.6365,
+      "step": 25730
+    },
+    {
+      "epoch": 1.2784344889242079,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006977292142644284,
+      "loss": 0.669,
+      "step": 25740
+    },
+    {
+      "epoch": 1.278931161219827,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0006976894804807789,
+      "loss": 0.6098,
+      "step": 25750
+    },
+    {
+      "epoch": 1.2794278335154465,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006976497466971292,
+      "loss": 0.6255,
+      "step": 25760
+    },
+    {
+      "epoch": 1.2799245058110658,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0006976100129134797,
+      "loss": 0.6222,
+      "step": 25770
+    },
+    {
+      "epoch": 1.2804211781066852,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006975702791298301,
+      "loss": 0.6459,
+      "step": 25780
+    },
+    {
+      "epoch": 1.2809178504023047,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006975305453461806,
+      "loss": 0.6226,
+      "step": 25790
+    },
+    {
+      "epoch": 1.2814145226979239,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0006974908115625311,
+      "loss": 0.6342,
+      "step": 25800
+    },
+    {
+      "epoch": 1.2819111949935433,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0006974510777788815,
+      "loss": 0.6535,
+      "step": 25810
+    },
+    {
+      "epoch": 1.2824078672891626,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0006974113439952319,
+      "loss": 0.6288,
+      "step": 25820
+    },
+    {
+      "epoch": 1.282904539584782,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0006973716102115825,
+      "loss": 0.6131,
+      "step": 25830
+    },
+    {
+      "epoch": 1.2834012118804012,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006973318764279329,
+      "loss": 0.6023,
+      "step": 25840
+    },
+    {
+      "epoch": 1.2838978841760207,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006972921426442834,
+      "loss": 0.6053,
+      "step": 25850
+    },
+    {
+      "epoch": 1.2843945564716401,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0006972524088606338,
+      "loss": 0.6479,
+      "step": 25860
+    },
+    {
+      "epoch": 1.2848912287672594,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006972126750769842,
+      "loss": 0.6086,
+      "step": 25870
+    },
+    {
+      "epoch": 1.2853879010628786,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0006971729412933348,
+      "loss": 0.6075,
+      "step": 25880
+    },
+    {
+      "epoch": 1.285884573358498,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0006971332075096851,
+      "loss": 0.616,
+      "step": 25890
+    },
+    {
+      "epoch": 1.2863812456541175,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006970934737260356,
+      "loss": 0.6025,
+      "step": 25900
+    },
+    {
+      "epoch": 1.2868779179497367,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006970537399423861,
+      "loss": 0.5992,
+      "step": 25910
+    },
+    {
+      "epoch": 1.2873745902453562,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0006970140061587364,
+      "loss": 0.6205,
+      "step": 25920
+    },
+    {
+      "epoch": 1.2878712625409754,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.000696974272375087,
+      "loss": 0.6286,
+      "step": 25930
+    },
+    {
+      "epoch": 1.2883679348365948,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006969345385914375,
+      "loss": 0.648,
+      "step": 25940
+    },
+    {
+      "epoch": 1.288864607132214,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006968948048077878,
+      "loss": 0.6122,
+      "step": 25950
+    },
+    {
+      "epoch": 1.2893612794278335,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006968550710241383,
+      "loss": 0.6291,
+      "step": 25960
+    },
+    {
+      "epoch": 1.289857951723453,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006968153372404887,
+      "loss": 0.6159,
+      "step": 25970
+    },
+    {
+      "epoch": 1.2903546240190722,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006967756034568392,
+      "loss": 0.6452,
+      "step": 25980
+    },
+    {
+      "epoch": 1.2908512963146916,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0006967358696731897,
+      "loss": 0.6365,
+      "step": 25990
+    },
+    {
+      "epoch": 1.2913479686103109,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006966961358895401,
+      "loss": 0.6077,
+      "step": 26000
+    },
+    {
+      "epoch": 1.2918446409059303,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006966564021058906,
+      "loss": 0.6097,
+      "step": 26010
+    },
+    {
+      "epoch": 1.2923413132015495,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.000696616668322241,
+      "loss": 0.622,
+      "step": 26020
+    },
+    {
+      "epoch": 1.292837985497169,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006965769345385914,
+      "loss": 0.5867,
+      "step": 26030
+    },
+    {
+      "epoch": 1.2933346577927884,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.000696537200754942,
+      "loss": 0.6268,
+      "step": 26040
+    },
+    {
+      "epoch": 1.2938313300884077,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006964974669712924,
+      "loss": 0.6172,
+      "step": 26050
+    },
+    {
+      "epoch": 1.294328002384027,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006964577331876428,
+      "loss": 0.61,
+      "step": 26060
+    },
+    {
+      "epoch": 1.2948246746796463,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006964179994039933,
+      "loss": 0.6454,
+      "step": 26070
+    },
+    {
+      "epoch": 1.2953213469752658,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0006963782656203437,
+      "loss": 0.6162,
+      "step": 26080
+    },
+    {
+      "epoch": 1.295818019270885,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006963385318366942,
+      "loss": 0.6403,
+      "step": 26090
+    },
+    {
+      "epoch": 1.2963146915665045,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006962987980530447,
+      "loss": 0.6284,
+      "step": 26100
+    },
+    {
+      "epoch": 1.2968113638621237,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.000696259064269395,
+      "loss": 0.5865,
+      "step": 26110
+    },
+    {
+      "epoch": 1.2973080361577431,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0006962193304857455,
+      "loss": 0.5768,
+      "step": 26120
+    },
+    {
+      "epoch": 1.2978047084533624,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.000696179596702096,
+      "loss": 0.6388,
+      "step": 26130
+    },
+    {
+      "epoch": 1.2983013807489818,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006961398629184465,
+      "loss": 0.632,
+      "step": 26140
+    },
+    {
+      "epoch": 1.2987980530446013,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0006961001291347969,
+      "loss": 0.6346,
+      "step": 26150
+    },
+    {
+      "epoch": 1.2992947253402205,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0006960603953511473,
+      "loss": 0.6329,
+      "step": 26160
+    },
+    {
+      "epoch": 1.29979139763584,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0006960206615674978,
+      "loss": 0.6127,
+      "step": 26170
+    },
+    {
+      "epoch": 1.3002880699314592,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006959809277838482,
+      "loss": 0.6357,
+      "step": 26180
+    },
+    {
+      "epoch": 1.3007847422270786,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0006959411940001987,
+      "loss": 0.6332,
+      "step": 26190
+    },
+    {
+      "epoch": 1.3012814145226979,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0006959014602165492,
+      "loss": 0.612,
+      "step": 26200
+    },
+    {
+      "epoch": 1.3017780868183173,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0006958617264328996,
+      "loss": 0.6465,
+      "step": 26210
+    },
+    {
+      "epoch": 1.3022747591139368,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.00069582199264925,
+      "loss": 0.629,
+      "step": 26220
+    },
+    {
+      "epoch": 1.302771431409556,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0006957822588656005,
+      "loss": 0.5934,
+      "step": 26230
+    },
+    {
+      "epoch": 1.3032681037051752,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.000695742525081951,
+      "loss": 0.6104,
+      "step": 26240
+    },
+    {
+      "epoch": 1.3037647760007947,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006957027912983014,
+      "loss": 0.6124,
+      "step": 26250
+    },
+    {
+      "epoch": 1.304261448296414,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006956630575146519,
+      "loss": 0.6126,
+      "step": 26260
+    },
+    {
+      "epoch": 1.3047581205920333,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0006956233237310023,
+      "loss": 0.5908,
+      "step": 26270
+    },
+    {
+      "epoch": 1.3052547928876528,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006955835899473527,
+      "loss": 0.5986,
+      "step": 26280
+    },
+    {
+      "epoch": 1.305751465183272,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006955438561637033,
+      "loss": 0.6358,
+      "step": 26290
+    },
+    {
+      "epoch": 1.3062481374788915,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0006955041223800537,
+      "loss": 0.6091,
+      "step": 26300
+    },
+    {
+      "epoch": 1.3067448097745107,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0006954643885964041,
+      "loss": 0.6536,
+      "step": 26310
+    },
+    {
+      "epoch": 1.3072414820701301,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006954246548127546,
+      "loss": 0.6112,
+      "step": 26320
+    },
+    {
+      "epoch": 1.3077381543657496,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.000695384921029105,
+      "loss": 0.6137,
+      "step": 26330
+    },
+    {
+      "epoch": 1.3082348266613688,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006953451872454555,
+      "loss": 0.6373,
+      "step": 26340
+    },
+    {
+      "epoch": 1.3087314989569883,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.000695305453461806,
+      "loss": 0.6115,
+      "step": 26350
+    },
+    {
+      "epoch": 1.3092281712526075,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006952657196781564,
+      "loss": 0.6095,
+      "step": 26360
+    },
+    {
+      "epoch": 1.309724843548227,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0006952259858945068,
+      "loss": 0.594,
+      "step": 26370
+    },
+    {
+      "epoch": 1.3102215158438462,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0006951862521108572,
+      "loss": 0.6177,
+      "step": 26380
+    },
+    {
+      "epoch": 1.3107181881394656,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0006951465183272078,
+      "loss": 0.5909,
+      "step": 26390
+    },
+    {
+      "epoch": 1.311214860435085,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006951067845435582,
+      "loss": 0.5882,
+      "step": 26400
+    },
+    {
+      "epoch": 1.3117115327307043,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0006950670507599086,
+      "loss": 0.6251,
+      "step": 26410
+    },
+    {
+      "epoch": 1.3122082050263235,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006950273169762591,
+      "loss": 0.6158,
+      "step": 26420
+    },
+    {
+      "epoch": 1.312704877321943,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0006949875831926095,
+      "loss": 0.636,
+      "step": 26430
+    },
+    {
+      "epoch": 1.3132015496175624,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00069494784940896,
+      "loss": 0.6297,
+      "step": 26440
+    },
+    {
+      "epoch": 1.3136982219131816,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0006949081156253105,
+      "loss": 0.5981,
+      "step": 26450
+    },
+    {
+      "epoch": 1.314194894208801,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006948683818416609,
+      "loss": 0.6133,
+      "step": 26460
+    },
+    {
+      "epoch": 1.3146915665044203,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006948286480580113,
+      "loss": 0.6271,
+      "step": 26470
+    },
+    {
+      "epoch": 1.3151882388000398,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0006947889142743618,
+      "loss": 0.6233,
+      "step": 26480
+    },
+    {
+      "epoch": 1.315684911095659,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0006947491804907123,
+      "loss": 0.6188,
+      "step": 26490
+    },
+    {
+      "epoch": 1.3161815833912784,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006947094467070627,
+      "loss": 0.588,
+      "step": 26500
+    },
+    {
+      "epoch": 1.316678255686898,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006946697129234132,
+      "loss": 0.6201,
+      "step": 26510
+    },
+    {
+      "epoch": 1.3171749279825171,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0006946299791397636,
+      "loss": 0.6294,
+      "step": 26520
+    },
+    {
+      "epoch": 1.3176716002781366,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.000694590245356114,
+      "loss": 0.6203,
+      "step": 26530
+    },
+    {
+      "epoch": 1.3181682725737558,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006945505115724646,
+      "loss": 0.6211,
+      "step": 26540
+    },
+    {
+      "epoch": 1.3186649448693752,
+      "grad_norm": 0.125,
+      "learning_rate": 0.000694510777788815,
+      "loss": 0.6181,
+      "step": 26550
+    },
+    {
+      "epoch": 1.3191616171649945,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0006944710440051654,
+      "loss": 0.6553,
+      "step": 26560
+    },
+    {
+      "epoch": 1.319658289460614,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006944313102215158,
+      "loss": 0.5941,
+      "step": 26570
+    },
+    {
+      "epoch": 1.3201549617562334,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0006943915764378663,
+      "loss": 0.6245,
+      "step": 26580
+    },
+    {
+      "epoch": 1.3206516340518526,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0006943518426542169,
+      "loss": 0.6374,
+      "step": 26590
+    },
+    {
+      "epoch": 1.3211483063474718,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0006943121088705672,
+      "loss": 0.6443,
+      "step": 26600
+    },
+    {
+      "epoch": 1.3216449786430913,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0006942723750869177,
+      "loss": 0.6219,
+      "step": 26610
+    },
+    {
+      "epoch": 1.3221416509387107,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006942326413032682,
+      "loss": 0.6492,
+      "step": 26620
+    },
+    {
+      "epoch": 1.32263832323433,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0006941929075196185,
+      "loss": 0.6174,
+      "step": 26630
+    },
+    {
+      "epoch": 1.3231349955299494,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0006941531737359691,
+      "loss": 0.5641,
+      "step": 26640
+    },
+    {
+      "epoch": 1.3236316678255686,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006941134399523195,
+      "loss": 0.6142,
+      "step": 26650
+    },
+    {
+      "epoch": 1.324128340121188,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0006940737061686699,
+      "loss": 0.6077,
+      "step": 26660
+    },
+    {
+      "epoch": 1.3246250124168073,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006940339723850204,
+      "loss": 0.6305,
+      "step": 26670
+    },
+    {
+      "epoch": 1.3251216847124268,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006939942386013708,
+      "loss": 0.6343,
+      "step": 26680
+    },
+    {
+      "epoch": 1.3256183570080462,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006939545048177213,
+      "loss": 0.5751,
+      "step": 26690
+    },
+    {
+      "epoch": 1.3261150293036654,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006939147710340718,
+      "loss": 0.6058,
+      "step": 26700
+    },
+    {
+      "epoch": 1.3266117015992847,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0006938750372504222,
+      "loss": 0.6087,
+      "step": 26710
+    },
+    {
+      "epoch": 1.327108373894904,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0006938353034667726,
+      "loss": 0.6055,
+      "step": 26720
+    },
+    {
+      "epoch": 1.3276050461905236,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006937955696831231,
+      "loss": 0.6121,
+      "step": 26730
+    },
+    {
+      "epoch": 1.3281017184861428,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006937558358994736,
+      "loss": 0.5855,
+      "step": 26740
+    },
+    {
+      "epoch": 1.3285983907817622,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0006937161021158241,
+      "loss": 0.5932,
+      "step": 26750
+    },
+    {
+      "epoch": 1.3290950630773817,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006936763683321744,
+      "loss": 0.6314,
+      "step": 26760
+    },
+    {
+      "epoch": 1.329591735373001,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006936366345485249,
+      "loss": 0.6129,
+      "step": 26770
+    },
+    {
+      "epoch": 1.3300884076686201,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006935969007648754,
+      "loss": 0.5807,
+      "step": 26780
+    },
+    {
+      "epoch": 1.3305850799642396,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0006935571669812257,
+      "loss": 0.5593,
+      "step": 26790
+    },
+    {
+      "epoch": 1.331081752259859,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006935174331975763,
+      "loss": 0.6105,
+      "step": 26800
+    },
+    {
+      "epoch": 1.3315784245554783,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0006934776994139268,
+      "loss": 0.589,
+      "step": 26810
+    },
+    {
+      "epoch": 1.3320750968510977,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0006934379656302771,
+      "loss": 0.6002,
+      "step": 26820
+    },
+    {
+      "epoch": 1.332571769146717,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006933982318466276,
+      "loss": 0.5998,
+      "step": 26830
+    },
+    {
+      "epoch": 1.3330684414423364,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.000693358498062978,
+      "loss": 0.6258,
+      "step": 26840
+    },
+    {
+      "epoch": 1.3335651137379556,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006933187642793285,
+      "loss": 0.628,
+      "step": 26850
+    },
+    {
+      "epoch": 1.334061786033575,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.000693279030495679,
+      "loss": 0.6142,
+      "step": 26860
+    },
+    {
+      "epoch": 1.3345584583291945,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0006932392967120294,
+      "loss": 0.6212,
+      "step": 26870
+    },
+    {
+      "epoch": 1.3350551306248137,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006931995629283799,
+      "loss": 0.6142,
+      "step": 26880
+    },
+    {
+      "epoch": 1.335551802920433,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0006931598291447304,
+      "loss": 0.6429,
+      "step": 26890
+    },
+    {
+      "epoch": 1.3360484752160524,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006931200953610808,
+      "loss": 0.6278,
+      "step": 26900
+    },
+    {
+      "epoch": 1.3365451475116719,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0006930803615774313,
+      "loss": 0.6065,
+      "step": 26910
+    },
+    {
+      "epoch": 1.337041819807291,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0006930406277937817,
+      "loss": 0.62,
+      "step": 26920
+    },
+    {
+      "epoch": 1.3375384921029105,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006930008940101321,
+      "loss": 0.6124,
+      "step": 26930
+    },
+    {
+      "epoch": 1.33803516439853,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0006929611602264827,
+      "loss": 0.618,
+      "step": 26940
+    },
+    {
+      "epoch": 1.3385318366941492,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006929214264428331,
+      "loss": 0.6175,
+      "step": 26950
+    },
+    {
+      "epoch": 1.3390285089897684,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006928816926591835,
+      "loss": 0.6139,
+      "step": 26960
+    },
+    {
+      "epoch": 1.339525181285388,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.000692841958875534,
+      "loss": 0.6035,
+      "step": 26970
+    },
+    {
+      "epoch": 1.3400218535810073,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0006928022250918843,
+      "loss": 0.6273,
+      "step": 26980
+    },
+    {
+      "epoch": 1.3405185258766266,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006927624913082349,
+      "loss": 0.5981,
+      "step": 26990
+    },
+    {
+      "epoch": 1.341015198172246,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0006927227575245854,
+      "loss": 0.6299,
+      "step": 27000
+    },
+    {
+      "epoch": 1.3415118704678652,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0006926830237409357,
+      "loss": 0.6188,
+      "step": 27010
+    },
+    {
+      "epoch": 1.3420085427634847,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006926432899572862,
+      "loss": 0.6372,
+      "step": 27020
+    },
+    {
+      "epoch": 1.342505215059104,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0006926035561736366,
+      "loss": 0.6362,
+      "step": 27030
+    },
+    {
+      "epoch": 1.3430018873547234,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0006925638223899872,
+      "loss": 0.6245,
+      "step": 27040
+    },
+    {
+      "epoch": 1.3434985596503428,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006925240886063376,
+      "loss": 0.6176,
+      "step": 27050
+    },
+    {
+      "epoch": 1.343995231945962,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.000692484354822688,
+      "loss": 0.5695,
+      "step": 27060
+    },
+    {
+      "epoch": 1.3444919042415813,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006924446210390385,
+      "loss": 0.6038,
+      "step": 27070
+    },
+    {
+      "epoch": 1.3449885765372007,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006924048872553889,
+      "loss": 0.6044,
+      "step": 27080
+    },
+    {
+      "epoch": 1.3454852488328202,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0006923651534717393,
+      "loss": 0.606,
+      "step": 27090
+    },
+    {
+      "epoch": 1.3459819211284394,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0006923254196880899,
+      "loss": 0.5946,
+      "step": 27100
+    },
+    {
+      "epoch": 1.3464785934240588,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006922856859044403,
+      "loss": 0.6438,
+      "step": 27110
+    },
+    {
+      "epoch": 1.3469752657196783,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0006922459521207907,
+      "loss": 0.5927,
+      "step": 27120
+    },
+    {
+      "epoch": 1.3474719380152975,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0006922062183371412,
+      "loss": 0.5905,
+      "step": 27130
+    },
+    {
+      "epoch": 1.3479686103109167,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006921664845534917,
+      "loss": 0.6242,
+      "step": 27140
+    },
+    {
+      "epoch": 1.3484652826065362,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006921267507698421,
+      "loss": 0.6159,
+      "step": 27150
+    },
+    {
+      "epoch": 1.3489619549021556,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006920870169861926,
+      "loss": 0.6114,
+      "step": 27160
+    },
+    {
+      "epoch": 1.3494586271977749,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.000692047283202543,
+      "loss": 0.5995,
+      "step": 27170
+    },
+    {
+      "epoch": 1.3499552994933943,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006920075494188934,
+      "loss": 0.6224,
+      "step": 27180
+    },
+    {
+      "epoch": 1.3504519717890135,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.000691967815635244,
+      "loss": 0.6009,
+      "step": 27190
+    },
+    {
+      "epoch": 1.350948644084633,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0006919280818515944,
+      "loss": 0.5902,
+      "step": 27200
+    },
+    {
+      "epoch": 1.3514453163802522,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006918883480679448,
+      "loss": 0.606,
+      "step": 27210
+    },
+    {
+      "epoch": 1.3519419886758717,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006918486142842953,
+      "loss": 0.6061,
+      "step": 27220
+    },
+    {
+      "epoch": 1.3524386609714911,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0006918088805006457,
+      "loss": 0.6187,
+      "step": 27230
+    },
+    {
+      "epoch": 1.3529353332671104,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006917691467169961,
+      "loss": 0.6309,
+      "step": 27240
+    },
+    {
+      "epoch": 1.3534320055627296,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006917294129333466,
+      "loss": 0.625,
+      "step": 27250
+    },
+    {
+      "epoch": 1.353928677858349,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006916896791496971,
+      "loss": 0.6102,
+      "step": 27260
+    },
+    {
+      "epoch": 1.3544253501539685,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0006916499453660475,
+      "loss": 0.6332,
+      "step": 27270
+    },
+    {
+      "epoch": 1.3549220224495877,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006916102115823979,
+      "loss": 0.5904,
+      "step": 27280
+    },
+    {
+      "epoch": 1.3554186947452072,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006915704777987484,
+      "loss": 0.6378,
+      "step": 27290
+    },
+    {
+      "epoch": 1.3559153670408264,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006915307440150989,
+      "loss": 0.61,
+      "step": 27300
+    },
+    {
+      "epoch": 1.3564120393364458,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006914910102314493,
+      "loss": 0.6269,
+      "step": 27310
+    },
+    {
+      "epoch": 1.356908711632065,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006914512764477998,
+      "loss": 0.603,
+      "step": 27320
+    },
+    {
+      "epoch": 1.3574053839276845,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006914115426641502,
+      "loss": 0.619,
+      "step": 27330
+    },
+    {
+      "epoch": 1.357902056223304,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006913718088805006,
+      "loss": 0.6086,
+      "step": 27340
+    },
+    {
+      "epoch": 1.3583987285189232,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006913320750968512,
+      "loss": 0.608,
+      "step": 27350
+    },
+    {
+      "epoch": 1.3588954008145426,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0006912923413132016,
+      "loss": 0.595,
+      "step": 27360
+    },
+    {
+      "epoch": 1.3593920731101619,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.000691252607529552,
+      "loss": 0.6164,
+      "step": 27370
+    },
+    {
+      "epoch": 1.3598887454057813,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006912128737459025,
+      "loss": 0.6428,
+      "step": 27380
+    },
+    {
+      "epoch": 1.3603854177014005,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.000691173139962253,
+      "loss": 0.6131,
+      "step": 27390
+    },
+    {
+      "epoch": 1.36088208999702,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0006911334061786034,
+      "loss": 0.6363,
+      "step": 27400
+    },
+    {
+      "epoch": 1.3613787622926394,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0006910936723949539,
+      "loss": 0.5877,
+      "step": 27410
+    },
+    {
+      "epoch": 1.3618754345882587,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0006910539386113043,
+      "loss": 0.6,
+      "step": 27420
+    },
+    {
+      "epoch": 1.3623721068838779,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0006910142048276547,
+      "loss": 0.6287,
+      "step": 27430
+    },
+    {
+      "epoch": 1.3628687791794973,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006909744710440051,
+      "loss": 0.6438,
+      "step": 27440
+    },
+    {
+      "epoch": 1.3633654514751168,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0006909347372603557,
+      "loss": 0.6215,
+      "step": 27450
+    },
+    {
+      "epoch": 1.363862123770736,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006908950034767061,
+      "loss": 0.6013,
+      "step": 27460
+    },
+    {
+      "epoch": 1.3643587960663555,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006908552696930565,
+      "loss": 0.6294,
+      "step": 27470
+    },
+    {
+      "epoch": 1.3648554683619747,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.000690815535909407,
+      "loss": 0.6018,
+      "step": 27480
+    },
+    {
+      "epoch": 1.3653521406575941,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006907758021257576,
+      "loss": 0.5911,
+      "step": 27490
+    },
+    {
+      "epoch": 1.3658488129532134,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006907360683421079,
+      "loss": 0.6479,
+      "step": 27500
+    },
+    {
+      "epoch": 1.3663454852488328,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0006906963345584584,
+      "loss": 0.6117,
+      "step": 27510
+    },
+    {
+      "epoch": 1.3668421575444523,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006906566007748088,
+      "loss": 0.6127,
+      "step": 27520
+    },
+    {
+      "epoch": 1.3673388298400715,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006906168669911592,
+      "loss": 0.6233,
+      "step": 27530
+    },
+    {
+      "epoch": 1.367835502135691,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006905771332075097,
+      "loss": 0.6019,
+      "step": 27540
+    },
+    {
+      "epoch": 1.3683321744313102,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0006905373994238602,
+      "loss": 0.6169,
+      "step": 27550
+    },
+    {
+      "epoch": 1.3688288467269296,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0006904976656402106,
+      "loss": 0.5841,
+      "step": 27560
+    },
+    {
+      "epoch": 1.3693255190225488,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0006904579318565611,
+      "loss": 0.6046,
+      "step": 27570
+    },
+    {
+      "epoch": 1.3698221913181683,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006904181980729115,
+      "loss": 0.609,
+      "step": 27580
+    },
+    {
+      "epoch": 1.3703188636137877,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006903784642892619,
+      "loss": 0.6145,
+      "step": 27590
+    },
+    {
+      "epoch": 1.370815535909407,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0006903387305056125,
+      "loss": 0.6483,
+      "step": 27600
+    },
+    {
+      "epoch": 1.3713122082050262,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0006902989967219629,
+      "loss": 0.6199,
+      "step": 27610
+    },
+    {
+      "epoch": 1.3718088805006456,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0006902592629383134,
+      "loss": 0.6271,
+      "step": 27620
+    },
+    {
+      "epoch": 1.372305552796265,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006902195291546637,
+      "loss": 0.6004,
+      "step": 27630
+    },
+    {
+      "epoch": 1.3728022250918843,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0006901797953710142,
+      "loss": 0.6027,
+      "step": 27640
+    },
+    {
+      "epoch": 1.3732988973875038,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0006901400615873648,
+      "loss": 0.6139,
+      "step": 27650
+    },
+    {
+      "epoch": 1.373795569683123,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006901003278037151,
+      "loss": 0.6204,
+      "step": 27660
+    },
+    {
+      "epoch": 1.3742922419787424,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0006900605940200656,
+      "loss": 0.5861,
+      "step": 27670
+    },
+    {
+      "epoch": 1.3747889142743617,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006900208602364161,
+      "loss": 0.6278,
+      "step": 27680
+    },
+    {
+      "epoch": 1.3752855865699811,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0006899811264527664,
+      "loss": 0.6139,
+      "step": 27690
+    },
+    {
+      "epoch": 1.3757822588656006,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.000689941392669117,
+      "loss": 0.6202,
+      "step": 27700
+    },
+    {
+      "epoch": 1.3762789311612198,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0006899016588854674,
+      "loss": 0.5959,
+      "step": 27710
+    },
+    {
+      "epoch": 1.3767756034568392,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006898619251018178,
+      "loss": 0.6025,
+      "step": 27720
+    },
+    {
+      "epoch": 1.3772722757524585,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006898221913181683,
+      "loss": 0.5889,
+      "step": 27730
+    },
+    {
+      "epoch": 1.377768948048078,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0006897824575345187,
+      "loss": 0.6065,
+      "step": 27740
+    },
+    {
+      "epoch": 1.3782656203436972,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0006897427237508692,
+      "loss": 0.5845,
+      "step": 27750
+    },
+    {
+      "epoch": 1.3787622926393166,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006897029899672197,
+      "loss": 0.5899,
+      "step": 27760
+    },
+    {
+      "epoch": 1.379258964934936,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006896632561835701,
+      "loss": 0.6149,
+      "step": 27770
+    },
+    {
+      "epoch": 1.3797556372305553,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006896235223999206,
+      "loss": 0.6169,
+      "step": 27780
+    },
+    {
+      "epoch": 1.3802523095261745,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.000689583788616271,
+      "loss": 0.6069,
+      "step": 27790
+    },
+    {
+      "epoch": 1.380748981821794,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006895440548326215,
+      "loss": 0.6278,
+      "step": 27800
+    },
+    {
+      "epoch": 1.3812456541174134,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.000689504321048972,
+      "loss": 0.5892,
+      "step": 27810
+    },
+    {
+      "epoch": 1.3817423264130326,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0006894645872653224,
+      "loss": 0.6149,
+      "step": 27820
+    },
+    {
+      "epoch": 1.382238998708652,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0006894248534816728,
+      "loss": 0.6042,
+      "step": 27830
+    },
+    {
+      "epoch": 1.3827356710042713,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0006893851196980233,
+      "loss": 0.6224,
+      "step": 27840
+    },
+    {
+      "epoch": 1.3832323432998908,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006893453859143737,
+      "loss": 0.6311,
+      "step": 27850
+    },
+    {
+      "epoch": 1.38372901559551,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0006893056521307242,
+      "loss": 0.6114,
+      "step": 27860
+    },
+    {
+      "epoch": 1.3842256878911294,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0006892659183470747,
+      "loss": 0.6021,
+      "step": 27870
+    },
+    {
+      "epoch": 1.3847223601867489,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.000689226184563425,
+      "loss": 0.6095,
+      "step": 27880
+    },
+    {
+      "epoch": 1.385219032482368,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006891864507797755,
+      "loss": 0.5971,
+      "step": 27890
+    },
+    {
+      "epoch": 1.3857157047779876,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.000689146716996126,
+      "loss": 0.5634,
+      "step": 27900
+    },
+    {
+      "epoch": 1.3862123770736068,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0006891069832124764,
+      "loss": 0.6128,
+      "step": 27910
+    },
+    {
+      "epoch": 1.3867090493692262,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006890672494288269,
+      "loss": 0.6113,
+      "step": 27920
+    },
+    {
+      "epoch": 1.3872057216648455,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006890275156451773,
+      "loss": 0.5969,
+      "step": 27930
+    },
+    {
+      "epoch": 1.387702393960465,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0006889877818615278,
+      "loss": 0.6106,
+      "step": 27940
+    },
+    {
+      "epoch": 1.3881990662560844,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006889480480778783,
+      "loss": 0.6181,
+      "step": 27950
+    },
+    {
+      "epoch": 1.3886957385517036,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006889083142942287,
+      "loss": 0.6144,
+      "step": 27960
+    },
+    {
+      "epoch": 1.3891924108473228,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006888685805105792,
+      "loss": 0.5952,
+      "step": 27970
+    },
+    {
+      "epoch": 1.3896890831429423,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006888288467269296,
+      "loss": 0.6273,
+      "step": 27980
+    },
+    {
+      "epoch": 1.3901857554385617,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.00068878911294328,
+      "loss": 0.611,
+      "step": 27990
+    },
+    {
+      "epoch": 1.390682427734181,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006887493791596306,
+      "loss": 0.6432,
+      "step": 28000
+    },
+    {
+      "epoch": 1.3911791000298004,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.000688709645375981,
+      "loss": 0.5919,
+      "step": 28010
+    },
+    {
+      "epoch": 1.3916757723254196,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0006886699115923314,
+      "loss": 0.5956,
+      "step": 28020
+    },
+    {
+      "epoch": 1.392172444621039,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006886301778086819,
+      "loss": 0.6022,
+      "step": 28030
+    },
+    {
+      "epoch": 1.3926691169166583,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006885904440250322,
+      "loss": 0.6028,
+      "step": 28040
+    },
+    {
+      "epoch": 1.3931657892122777,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006885507102413828,
+      "loss": 0.6129,
+      "step": 28050
+    },
+    {
+      "epoch": 1.3936624615078972,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0006885109764577333,
+      "loss": 0.5869,
+      "step": 28060
+    },
+    {
+      "epoch": 1.3941591338035164,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006884712426740837,
+      "loss": 0.6021,
+      "step": 28070
+    },
+    {
+      "epoch": 1.3946558060991359,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0006884315088904341,
+      "loss": 0.6046,
+      "step": 28080
+    },
+    {
+      "epoch": 1.395152478394755,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006883917751067846,
+      "loss": 0.603,
+      "step": 28090
+    },
+    {
+      "epoch": 1.3956491506903745,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0006883520413231351,
+      "loss": 0.614,
+      "step": 28100
+    },
+    {
+      "epoch": 1.3961458229859938,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0006883123075394855,
+      "loss": 0.5915,
+      "step": 28110
+    },
+    {
+      "epoch": 1.3966424952816132,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0006882725737558359,
+      "loss": 0.6009,
+      "step": 28120
+    },
+    {
+      "epoch": 1.3971391675772327,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0006882328399721864,
+      "loss": 0.6098,
+      "step": 28130
+    },
+    {
+      "epoch": 1.397635839872852,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006881931061885368,
+      "loss": 0.6106,
+      "step": 28140
+    },
+    {
+      "epoch": 1.3981325121684711,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006881533724048873,
+      "loss": 0.6075,
+      "step": 28150
+    },
+    {
+      "epoch": 1.3986291844640906,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006881136386212378,
+      "loss": 0.6005,
+      "step": 28160
+    },
+    {
+      "epoch": 1.39912585675971,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0006880739048375882,
+      "loss": 0.5895,
+      "step": 28170
+    },
+    {
+      "epoch": 1.3996225290553292,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006880341710539386,
+      "loss": 0.6271,
+      "step": 28180
+    },
+    {
+      "epoch": 1.4001192013509487,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006879944372702891,
+      "loss": 0.6261,
+      "step": 28190
+    },
+    {
+      "epoch": 1.400615873646568,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0006879547034866396,
+      "loss": 0.6023,
+      "step": 28200
+    },
+    {
+      "epoch": 1.4011125459421874,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00068791496970299,
+      "loss": 0.64,
+      "step": 28210
+    },
+    {
+      "epoch": 1.4016092182378066,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0006878752359193405,
+      "loss": 0.5955,
+      "step": 28220
+    },
+    {
+      "epoch": 1.402105890533426,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006878355021356909,
+      "loss": 0.5777,
+      "step": 28230
+    },
+    {
+      "epoch": 1.4026025628290455,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0006877957683520413,
+      "loss": 0.6048,
+      "step": 28240
+    },
+    {
+      "epoch": 1.4030992351246647,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0006877560345683919,
+      "loss": 0.6006,
+      "step": 28250
+    },
+    {
+      "epoch": 1.4035959074202842,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0006877163007847423,
+      "loss": 0.5771,
+      "step": 28260
+    },
+    {
+      "epoch": 1.4040925797159034,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006876765670010927,
+      "loss": 0.6036,
+      "step": 28270
+    },
+    {
+      "epoch": 1.4045892520115228,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006876368332174432,
+      "loss": 0.5789,
+      "step": 28280
+    },
+    {
+      "epoch": 1.405085924307142,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006875970994337936,
+      "loss": 0.6106,
+      "step": 28290
+    },
+    {
+      "epoch": 1.4055825966027615,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.000687557365650144,
+      "loss": 0.6219,
+      "step": 28300
+    },
+    {
+      "epoch": 1.406079268898381,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0006875176318664945,
+      "loss": 0.6348,
+      "step": 28310
+    },
+    {
+      "epoch": 1.4065759411940002,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.000687477898082845,
+      "loss": 0.6038,
+      "step": 28320
+    },
+    {
+      "epoch": 1.4070726134896194,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006874381642991954,
+      "loss": 0.5846,
+      "step": 28330
+    },
+    {
+      "epoch": 1.4075692857852389,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0006873984305155458,
+      "loss": 0.6066,
+      "step": 28340
+    },
+    {
+      "epoch": 1.4080659580808583,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0006873586967318964,
+      "loss": 0.6239,
+      "step": 28350
+    },
+    {
+      "epoch": 1.4085626303764776,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006873189629482468,
+      "loss": 0.5947,
+      "step": 28360
+    },
+    {
+      "epoch": 1.409059302672097,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0006872792291645972,
+      "loss": 0.615,
+      "step": 28370
+    },
+    {
+      "epoch": 1.4095559749677162,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006872394953809477,
+      "loss": 0.5876,
+      "step": 28380
+    },
+    {
+      "epoch": 1.4100526472633357,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006871997615972981,
+      "loss": 0.6149,
+      "step": 28390
+    },
+    {
+      "epoch": 1.410549319558955,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0006871600278136485,
+      "loss": 0.5882,
+      "step": 28400
+    },
+    {
+      "epoch": 1.4110459918545744,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006871202940299991,
+      "loss": 0.6271,
+      "step": 28410
+    },
+    {
+      "epoch": 1.4115426641501938,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0006870805602463495,
+      "loss": 0.5835,
+      "step": 28420
+    },
+    {
+      "epoch": 1.412039336445813,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0006870408264626999,
+      "loss": 0.6312,
+      "step": 28430
+    },
+    {
+      "epoch": 1.4125360087414325,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006870010926790504,
+      "loss": 0.5719,
+      "step": 28440
+    },
+    {
+      "epoch": 1.4130326810370517,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0006869613588954009,
+      "loss": 0.5973,
+      "step": 28450
+    },
+    {
+      "epoch": 1.4135293533326712,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006869216251117513,
+      "loss": 0.6259,
+      "step": 28460
+    },
+    {
+      "epoch": 1.4140260256282904,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006868818913281018,
+      "loss": 0.6359,
+      "step": 28470
+    },
+    {
+      "epoch": 1.4145226979239098,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006868421575444522,
+      "loss": 0.5925,
+      "step": 28480
+    },
+    {
+      "epoch": 1.4150193702195293,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0006868024237608026,
+      "loss": 0.5965,
+      "step": 28490
+    },
+    {
+      "epoch": 1.4155160425151485,
+      "grad_norm": 0.125,
+      "learning_rate": 0.000686762689977153,
+      "loss": 0.6216,
+      "step": 28500
+    },
+    {
+      "epoch": 1.4160127148107677,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0006867229561935036,
+      "loss": 0.5953,
+      "step": 28510
+    },
+    {
+      "epoch": 1.4165093871063872,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006866832224098541,
+      "loss": 0.6156,
+      "step": 28520
+    },
+    {
+      "epoch": 1.4170060594020066,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006866434886262044,
+      "loss": 0.6186,
+      "step": 28530
+    },
+    {
+      "epoch": 1.4175027316976259,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0006866037548425549,
+      "loss": 0.6166,
+      "step": 28540
+    },
+    {
+      "epoch": 1.4179994039932453,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006865640210589055,
+      "loss": 0.5936,
+      "step": 28550
+    },
+    {
+      "epoch": 1.4184960762888645,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006865242872752558,
+      "loss": 0.6375,
+      "step": 28560
+    },
+    {
+      "epoch": 1.418992748584484,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0006864845534916063,
+      "loss": 0.6165,
+      "step": 28570
+    },
+    {
+      "epoch": 1.4194894208801032,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0006864448197079567,
+      "loss": 0.5936,
+      "step": 28580
+    },
+    {
+      "epoch": 1.4199860931757227,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006864050859243071,
+      "loss": 0.5865,
+      "step": 28590
+    },
+    {
+      "epoch": 1.4204827654713421,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006863653521406576,
+      "loss": 0.5765,
+      "step": 28600
+    },
+    {
+      "epoch": 1.4209794377669613,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006863256183570081,
+      "loss": 0.5794,
+      "step": 28610
+    },
+    {
+      "epoch": 1.4214761100625808,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006862858845733585,
+      "loss": 0.6259,
+      "step": 28620
+    },
+    {
+      "epoch": 1.4219727823582,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000686246150789709,
+      "loss": 0.6048,
+      "step": 28630
+    },
+    {
+      "epoch": 1.4224694546538195,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0006862064170060594,
+      "loss": 0.606,
+      "step": 28640
+    },
+    {
+      "epoch": 1.4229661269494387,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006861666832224098,
+      "loss": 0.6099,
+      "step": 28650
+    },
+    {
+      "epoch": 1.4234627992450581,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0006861269494387604,
+      "loss": 0.6104,
+      "step": 28660
+    },
+    {
+      "epoch": 1.4239594715406776,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006860872156551108,
+      "loss": 0.6408,
+      "step": 28670
+    },
+    {
+      "epoch": 1.4244561438362968,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006860474818714613,
+      "loss": 0.6107,
+      "step": 28680
+    },
+    {
+      "epoch": 1.424952816131916,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006860077480878117,
+      "loss": 0.6043,
+      "step": 28690
+    },
+    {
+      "epoch": 1.4254494884275355,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006859680143041621,
+      "loss": 0.6196,
+      "step": 28700
+    },
+    {
+      "epoch": 1.425946160723155,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0006859282805205127,
+      "loss": 0.5761,
+      "step": 28710
+    },
+    {
+      "epoch": 1.4264428330187742,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.000685888546736863,
+      "loss": 0.61,
+      "step": 28720
+    },
+    {
+      "epoch": 1.4269395053143936,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0006858488129532135,
+      "loss": 0.5906,
+      "step": 28730
+    },
+    {
+      "epoch": 1.4274361776100128,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.000685809079169564,
+      "loss": 0.578,
+      "step": 28740
+    },
+    {
+      "epoch": 1.4279328499056323,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0006857693453859143,
+      "loss": 0.6389,
+      "step": 28750
+    },
+    {
+      "epoch": 1.4284295222012515,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006857296116022649,
+      "loss": 0.6311,
+      "step": 28760
+    },
+    {
+      "epoch": 1.428926194496871,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0006856898778186153,
+      "loss": 0.6115,
+      "step": 28770
+    },
+    {
+      "epoch": 1.4294228667924904,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006856501440349657,
+      "loss": 0.5909,
+      "step": 28780
+    },
+    {
+      "epoch": 1.4299195390881096,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0006856104102513162,
+      "loss": 0.5808,
+      "step": 28790
+    },
+    {
+      "epoch": 1.430416211383729,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0006855706764676666,
+      "loss": 0.6197,
+      "step": 28800
+    },
+    {
+      "epoch": 1.4309128836793483,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006855309426840171,
+      "loss": 0.5979,
+      "step": 28810
+    },
+    {
+      "epoch": 1.4314095559749678,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006854912089003676,
+      "loss": 0.6095,
+      "step": 28820
+    },
+    {
+      "epoch": 1.431906228270587,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.000685451475116718,
+      "loss": 0.6002,
+      "step": 28830
+    },
+    {
+      "epoch": 1.4324029005662064,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0006854117413330685,
+      "loss": 0.6342,
+      "step": 28840
+    },
+    {
+      "epoch": 1.432899572861826,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006853720075494189,
+      "loss": 0.6459,
+      "step": 28850
+    },
+    {
+      "epoch": 1.4333962451574451,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0006853322737657694,
+      "loss": 0.5952,
+      "step": 28860
+    },
+    {
+      "epoch": 1.4338929174530644,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006852925399821199,
+      "loss": 0.6145,
+      "step": 28870
+    },
+    {
+      "epoch": 1.4343895897486838,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0006852528061984703,
+      "loss": 0.6077,
+      "step": 28880
+    },
+    {
+      "epoch": 1.4348862620443033,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006852130724148207,
+      "loss": 0.6086,
+      "step": 28890
+    },
+    {
+      "epoch": 1.4353829343399225,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006851733386311712,
+      "loss": 0.6101,
+      "step": 28900
+    },
+    {
+      "epoch": 1.435879606635542,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0006851336048475216,
+      "loss": 0.6171,
+      "step": 28910
+    },
+    {
+      "epoch": 1.4363762789311612,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0006850938710638721,
+      "loss": 0.5873,
+      "step": 28920
+    },
+    {
+      "epoch": 1.4368729512267806,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0006850541372802226,
+      "loss": 0.638,
+      "step": 28930
+    },
+    {
+      "epoch": 1.4373696235223998,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006850144034965729,
+      "loss": 0.6042,
+      "step": 28940
+    },
+    {
+      "epoch": 1.4378662958180193,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006849746697129234,
+      "loss": 0.5868,
+      "step": 28950
+    },
+    {
+      "epoch": 1.4383629681136387,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0006849349359292739,
+      "loss": 0.6008,
+      "step": 28960
+    },
+    {
+      "epoch": 1.438859640409258,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006848952021456244,
+      "loss": 0.6159,
+      "step": 28970
+    },
+    {
+      "epoch": 1.4393563127048772,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0006848554683619748,
+      "loss": 0.5899,
+      "step": 28980
+    },
+    {
+      "epoch": 1.4398529850004966,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0006848157345783252,
+      "loss": 0.5951,
+      "step": 28990
+    },
+    {
+      "epoch": 1.440349657296116,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006847760007946757,
+      "loss": 0.6214,
+      "step": 29000
+    },
+    {
+      "epoch": 1.4408463295917353,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0006847362670110262,
+      "loss": 0.6028,
+      "step": 29010
+    },
+    {
+      "epoch": 1.4413430018873548,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006846965332273766,
+      "loss": 0.6098,
+      "step": 29020
+    },
+    {
+      "epoch": 1.4418396741829742,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0006846567994437271,
+      "loss": 0.6203,
+      "step": 29030
+    },
+    {
+      "epoch": 1.4423363464785934,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006846170656600775,
+      "loss": 0.6068,
+      "step": 29040
+    },
+    {
+      "epoch": 1.4428330187742127,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006845773318764279,
+      "loss": 0.6099,
+      "step": 29050
+    },
+    {
+      "epoch": 1.443329691069832,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0006845375980927785,
+      "loss": 0.5605,
+      "step": 29060
+    },
+    {
+      "epoch": 1.4438263633654516,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006844978643091289,
+      "loss": 0.5866,
+      "step": 29070
+    },
+    {
+      "epoch": 1.4443230356610708,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0006844581305254793,
+      "loss": 0.6057,
+      "step": 29080
+    },
+    {
+      "epoch": 1.4448197079566902,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006844183967418298,
+      "loss": 0.6059,
+      "step": 29090
+    },
+    {
+      "epoch": 1.4453163802523095,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006843786629581801,
+      "loss": 0.5785,
+      "step": 29100
+    },
+    {
+      "epoch": 1.445813052547929,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006843389291745307,
+      "loss": 0.5752,
+      "step": 29110
+    },
+    {
+      "epoch": 1.4463097248435481,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0006842991953908812,
+      "loss": 0.5905,
+      "step": 29120
+    },
+    {
+      "epoch": 1.4468063971391676,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0006842594616072316,
+      "loss": 0.5859,
+      "step": 29130
+    },
+    {
+      "epoch": 1.447303069434787,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.000684219727823582,
+      "loss": 0.6083,
+      "step": 29140
+    },
+    {
+      "epoch": 1.4477997417304063,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0006841799940399325,
+      "loss": 0.5989,
+      "step": 29150
+    },
+    {
+      "epoch": 1.4482964140260255,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.000684140260256283,
+      "loss": 0.5948,
+      "step": 29160
+    },
+    {
+      "epoch": 1.448793086321645,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006841005264726334,
+      "loss": 0.6228,
+      "step": 29170
+    },
+    {
+      "epoch": 1.4492897586172644,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006840607926889838,
+      "loss": 0.599,
+      "step": 29180
+    },
+    {
+      "epoch": 1.4497864309128836,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006840210589053343,
+      "loss": 0.6125,
+      "step": 29190
+    },
+    {
+      "epoch": 1.450283103208503,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0006839813251216847,
+      "loss": 0.6139,
+      "step": 29200
+    },
+    {
+      "epoch": 1.4507797755041225,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006839415913380352,
+      "loss": 0.6245,
+      "step": 29210
+    },
+    {
+      "epoch": 1.4512764477997417,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006839018575543857,
+      "loss": 0.619,
+      "step": 29220
+    },
+    {
+      "epoch": 1.451773120095361,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0006838621237707361,
+      "loss": 0.6243,
+      "step": 29230
+    },
+    {
+      "epoch": 1.4522697923909804,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006838223899870865,
+      "loss": 0.5803,
+      "step": 29240
+    },
+    {
+      "epoch": 1.4527664646865999,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.000683782656203437,
+      "loss": 0.6058,
+      "step": 29250
+    },
+    {
+      "epoch": 1.453263136982219,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006837429224197875,
+      "loss": 0.6069,
+      "step": 29260
+    },
+    {
+      "epoch": 1.4537598092778385,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006837031886361379,
+      "loss": 0.5811,
+      "step": 29270
+    },
+    {
+      "epoch": 1.4542564815734578,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0006836634548524884,
+      "loss": 0.5839,
+      "step": 29280
+    },
+    {
+      "epoch": 1.4547531538690772,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0006836237210688388,
+      "loss": 0.6351,
+      "step": 29290
+    },
+    {
+      "epoch": 1.4552498261646964,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006835839872851892,
+      "loss": 0.6138,
+      "step": 29300
+    },
+    {
+      "epoch": 1.455746498460316,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0006835442535015398,
+      "loss": 0.6059,
+      "step": 29310
+    },
+    {
+      "epoch": 1.4562431707559353,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006835045197178902,
+      "loss": 0.5771,
+      "step": 29320
+    },
+    {
+      "epoch": 1.4567398430515546,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0006834647859342406,
+      "loss": 0.5944,
+      "step": 29330
+    },
+    {
+      "epoch": 1.4572365153471738,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0006834250521505911,
+      "loss": 0.5916,
+      "step": 29340
+    },
+    {
+      "epoch": 1.4577331876427932,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0006833853183669415,
+      "loss": 0.6127,
+      "step": 29350
+    },
+    {
+      "epoch": 1.4582298599384127,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.000683345584583292,
+      "loss": 0.6273,
+      "step": 29360
+    },
+    {
+      "epoch": 1.458726532234032,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006833058507996424,
+      "loss": 0.6252,
+      "step": 29370
+    },
+    {
+      "epoch": 1.4592232045296514,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0006832661170159929,
+      "loss": 0.5981,
+      "step": 29380
+    },
+    {
+      "epoch": 1.4597198768252708,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006832263832323433,
+      "loss": 0.6148,
+      "step": 29390
+    },
+    {
+      "epoch": 1.46021654912089,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006831866494486937,
+      "loss": 0.5942,
+      "step": 29400
+    },
+    {
+      "epoch": 1.4607132214165093,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0006831469156650443,
+      "loss": 0.6089,
+      "step": 29410
+    },
+    {
+      "epoch": 1.4612098937121287,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006831071818813948,
+      "loss": 0.5826,
+      "step": 29420
+    },
+    {
+      "epoch": 1.4617065660077482,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0006830674480977451,
+      "loss": 0.6103,
+      "step": 29430
+    },
+    {
+      "epoch": 1.4622032383033674,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006830277143140956,
+      "loss": 0.6019,
+      "step": 29440
+    },
+    {
+      "epoch": 1.4626999105989869,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.000682987980530446,
+      "loss": 0.6102,
+      "step": 29450
+    },
+    {
+      "epoch": 1.463196582894606,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0006829482467467965,
+      "loss": 0.6198,
+      "step": 29460
+    },
+    {
+      "epoch": 1.4636932551902255,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.000682908512963147,
+      "loss": 0.5934,
+      "step": 29470
+    },
+    {
+      "epoch": 1.4641899274858448,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0006828687791794974,
+      "loss": 0.5846,
+      "step": 29480
+    },
+    {
+      "epoch": 1.4646865997814642,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006828290453958478,
+      "loss": 0.6057,
+      "step": 29490
+    },
+    {
+      "epoch": 1.4651832720770837,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006827893116121983,
+      "loss": 0.633,
+      "step": 29500
+    },
+    {
+      "epoch": 1.4656799443727029,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0006827495778285488,
+      "loss": 0.6107,
+      "step": 29510
+    },
+    {
+      "epoch": 1.466176616668322,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0006827098440448992,
+      "loss": 0.6107,
+      "step": 29520
+    },
+    {
+      "epoch": 1.4666732889639416,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0006826701102612497,
+      "loss": 0.5991,
+      "step": 29530
+    },
+    {
+      "epoch": 1.467169961259561,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006826303764776001,
+      "loss": 0.6115,
+      "step": 29540
+    },
+    {
+      "epoch": 1.4676666335551802,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0006825906426939505,
+      "loss": 0.6148,
+      "step": 29550
+    },
+    {
+      "epoch": 1.4681633058507997,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.000682550908910301,
+      "loss": 0.6007,
+      "step": 29560
+    },
+    {
+      "epoch": 1.468659978146419,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006825111751266515,
+      "loss": 0.6115,
+      "step": 29570
+    },
+    {
+      "epoch": 1.4691566504420384,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.000682471441343002,
+      "loss": 0.6059,
+      "step": 29580
+    },
+    {
+      "epoch": 1.4696533227376576,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006824317075593523,
+      "loss": 0.6125,
+      "step": 29590
+    },
+    {
+      "epoch": 1.470149995033277,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006823919737757028,
+      "loss": 0.5979,
+      "step": 29600
+    },
+    {
+      "epoch": 1.4706466673288965,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0006823522399920534,
+      "loss": 0.598,
+      "step": 29610
+    },
+    {
+      "epoch": 1.4711433396245157,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0006823125062084037,
+      "loss": 0.6289,
+      "step": 29620
+    },
+    {
+      "epoch": 1.4716400119201352,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006822727724247542,
+      "loss": 0.6102,
+      "step": 29630
+    },
+    {
+      "epoch": 1.4721366842157544,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0006822330386411046,
+      "loss": 0.5902,
+      "step": 29640
+    },
+    {
+      "epoch": 1.4726333565113738,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.000682193304857455,
+      "loss": 0.5947,
+      "step": 29650
+    },
+    {
+      "epoch": 1.473130028806993,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006821535710738056,
+      "loss": 0.5978,
+      "step": 29660
+    },
+    {
+      "epoch": 1.4736267011026125,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.000682113837290156,
+      "loss": 0.6403,
+      "step": 29670
+    },
+    {
+      "epoch": 1.474123373398232,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0006820741035065064,
+      "loss": 0.607,
+      "step": 29680
+    },
+    {
+      "epoch": 1.4746200456938512,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006820343697228569,
+      "loss": 0.6096,
+      "step": 29690
+    },
+    {
+      "epoch": 1.4751167179894704,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006819946359392073,
+      "loss": 0.6089,
+      "step": 29700
+    },
+    {
+      "epoch": 1.4756133902850899,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0006819549021555579,
+      "loss": 0.6185,
+      "step": 29710
+    },
+    {
+      "epoch": 1.4761100625807093,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0006819151683719083,
+      "loss": 0.5886,
+      "step": 29720
+    },
+    {
+      "epoch": 1.4766067348763285,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006818754345882587,
+      "loss": 0.6284,
+      "step": 29730
+    },
+    {
+      "epoch": 1.477103407171948,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006818357008046092,
+      "loss": 0.587,
+      "step": 29740
+    },
+    {
+      "epoch": 1.4776000794675672,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006817959670209596,
+      "loss": 0.5764,
+      "step": 29750
+    },
+    {
+      "epoch": 1.4780967517631867,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00068175623323731,
+      "loss": 0.6388,
+      "step": 29760
+    },
+    {
+      "epoch": 1.478593424058806,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006817164994536606,
+      "loss": 0.5818,
+      "step": 29770
+    },
+    {
+      "epoch": 1.4790900963544253,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006816767656700109,
+      "loss": 0.6276,
+      "step": 29780
+    },
+    {
+      "epoch": 1.4795867686500448,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0006816370318863614,
+      "loss": 0.5938,
+      "step": 29790
+    },
+    {
+      "epoch": 1.480083440945664,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0006815972981027119,
+      "loss": 0.5975,
+      "step": 29800
+    },
+    {
+      "epoch": 1.4805801132412835,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006815575643190622,
+      "loss": 0.6227,
+      "step": 29810
+    },
+    {
+      "epoch": 1.4810767855369027,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0006815178305354128,
+      "loss": 0.612,
+      "step": 29820
+    },
+    {
+      "epoch": 1.4815734578325221,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006814780967517632,
+      "loss": 0.5949,
+      "step": 29830
+    },
+    {
+      "epoch": 1.4820701301281414,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006814383629681136,
+      "loss": 0.5996,
+      "step": 29840
+    },
+    {
+      "epoch": 1.4825668024237608,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006813986291844641,
+      "loss": 0.5888,
+      "step": 29850
+    },
+    {
+      "epoch": 1.4830634747193803,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006813588954008145,
+      "loss": 0.6183,
+      "step": 29860
+    },
+    {
+      "epoch": 1.4835601470149995,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0006813191616171651,
+      "loss": 0.626,
+      "step": 29870
+    },
+    {
+      "epoch": 1.4840568193106187,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0006812794278335155,
+      "loss": 0.6085,
+      "step": 29880
+    },
+    {
+      "epoch": 1.4845534916062382,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0006812396940498659,
+      "loss": 0.5884,
+      "step": 29890
+    },
+    {
+      "epoch": 1.4850501639018576,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006811999602662164,
+      "loss": 0.6011,
+      "step": 29900
+    },
+    {
+      "epoch": 1.4855468361974768,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006811602264825668,
+      "loss": 0.6363,
+      "step": 29910
+    },
+    {
+      "epoch": 1.4860435084930963,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006811204926989173,
+      "loss": 0.6158,
+      "step": 29920
+    },
+    {
+      "epoch": 1.4865401807887155,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0006810807589152678,
+      "loss": 0.5923,
+      "step": 29930
+    },
+    {
+      "epoch": 1.487036853084335,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0006810410251316182,
+      "loss": 0.6281,
+      "step": 29940
+    },
+    {
+      "epoch": 1.4875335253799542,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0006810012913479686,
+      "loss": 0.5919,
+      "step": 29950
+    },
+    {
+      "epoch": 1.4880301976755737,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006809615575643192,
+      "loss": 0.6249,
+      "step": 29960
+    },
+    {
+      "epoch": 1.488526869971193,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006809218237806695,
+      "loss": 0.5997,
+      "step": 29970
+    },
+    {
+      "epoch": 1.4890235422668123,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00068088208999702,
+      "loss": 0.6269,
+      "step": 29980
+    },
+    {
+      "epoch": 1.4895202145624318,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006808423562133705,
+      "loss": 0.6076,
+      "step": 29990
+    },
+    {
+      "epoch": 1.490016886858051,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0006808026224297208,
+      "loss": 0.5876,
+      "step": 30000
+    },
+    {
+      "epoch": 1.4905135591536705,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006807628886460713,
+      "loss": 0.6008,
+      "step": 30010
+    },
+    {
+      "epoch": 1.4910102314492897,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0006807231548624219,
+      "loss": 0.5796,
+      "step": 30020
+    },
+    {
+      "epoch": 1.4915069037449091,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006806834210787723,
+      "loss": 0.5966,
+      "step": 30030
+    },
+    {
+      "epoch": 1.4920035760405286,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006806436872951227,
+      "loss": 0.6028,
+      "step": 30040
+    },
+    {
+      "epoch": 1.4925002483361478,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0006806039535114731,
+      "loss": 0.6235,
+      "step": 30050
+    },
+    {
+      "epoch": 1.492996920631767,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006805642197278236,
+      "loss": 0.6222,
+      "step": 30060
+    },
+    {
+      "epoch": 1.4934935929273865,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0006805244859441741,
+      "loss": 0.5923,
+      "step": 30070
+    },
+    {
+      "epoch": 1.493990265223006,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0006804847521605245,
+      "loss": 0.5916,
+      "step": 30080
+    },
+    {
+      "epoch": 1.4944869375186252,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.000680445018376875,
+      "loss": 0.6256,
+      "step": 30090
+    },
+    {
+      "epoch": 1.4949836098142446,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006804052845932254,
+      "loss": 0.608,
+      "step": 30100
+    },
+    {
+      "epoch": 1.4954802821098638,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0006803655508095758,
+      "loss": 0.5927,
+      "step": 30110
+    },
+    {
+      "epoch": 1.4959769544054833,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006803258170259264,
+      "loss": 0.6394,
+      "step": 30120
+    },
+    {
+      "epoch": 1.4964736267011025,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0006802860832422768,
+      "loss": 0.6129,
+      "step": 30130
+    },
+    {
+      "epoch": 1.496970298996722,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0006802463494586272,
+      "loss": 0.6006,
+      "step": 30140
+    },
+    {
+      "epoch": 1.4974669712923414,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0006802066156749777,
+      "loss": 0.6076,
+      "step": 30150
+    },
+    {
+      "epoch": 1.4979636435879606,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0006801668818913281,
+      "loss": 0.6152,
+      "step": 30160
+    },
+    {
+      "epoch": 1.49846031588358,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006801271481076786,
+      "loss": 0.6272,
+      "step": 30170
+    },
+    {
+      "epoch": 1.4989569881791993,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006800874143240291,
+      "loss": 0.6088,
+      "step": 30180
+    },
+    {
+      "epoch": 1.4994536604748188,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006800476805403795,
+      "loss": 0.6184,
+      "step": 30190
+    },
+    {
+      "epoch": 1.499950332770438,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006800079467567299,
+      "loss": 0.594,
+      "step": 30200
+    },
+    {
+      "epoch": 1.5004470050660574,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0006799682129730804,
+      "loss": 0.617,
+      "step": 30210
+    },
+    {
+      "epoch": 1.5009436773616769,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006799284791894309,
+      "loss": 0.6101,
+      "step": 30220
+    },
+    {
+      "epoch": 1.5014403496572961,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0006798887454057813,
+      "loss": 0.6013,
+      "step": 30230
+    },
+    {
+      "epoch": 1.5019370219529153,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0006798490116221317,
+      "loss": 0.6145,
+      "step": 30240
+    },
+    {
+      "epoch": 1.5024336942485348,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0006798092778384822,
+      "loss": 0.6136,
+      "step": 30250
+    },
+    {
+      "epoch": 1.5029303665441542,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006797695440548326,
+      "loss": 0.6326,
+      "step": 30260
+    },
+    {
+      "epoch": 1.5034270388397735,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0006797298102711831,
+      "loss": 0.6127,
+      "step": 30270
+    },
+    {
+      "epoch": 1.503923711135393,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0006796900764875336,
+      "loss": 0.5979,
+      "step": 30280
+    },
+    {
+      "epoch": 1.5044203834310124,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.000679650342703884,
+      "loss": 0.6082,
+      "step": 30290
+    },
+    {
+      "epoch": 1.5049170557266316,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0006796106089202344,
+      "loss": 0.6313,
+      "step": 30300
+    },
+    {
+      "epoch": 1.5054137280222508,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006795708751365849,
+      "loss": 0.5912,
+      "step": 30310
+    },
+    {
+      "epoch": 1.5059104003178703,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0006795311413529354,
+      "loss": 0.6078,
+      "step": 30320
+    },
+    {
+      "epoch": 1.5064070726134897,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006794914075692858,
+      "loss": 0.5964,
+      "step": 30330
+    },
+    {
+      "epoch": 1.506903744909109,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006794516737856363,
+      "loss": 0.5968,
+      "step": 30340
+    },
+    {
+      "epoch": 1.5074004172047282,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006794119400019867,
+      "loss": 0.6018,
+      "step": 30350
+    },
+    {
+      "epoch": 1.5078970895003476,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006793722062183371,
+      "loss": 0.5776,
+      "step": 30360
+    },
+    {
+      "epoch": 1.508393761795967,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006793324724346877,
+      "loss": 0.616,
+      "step": 30370
+    },
+    {
+      "epoch": 1.5088904340915863,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006792927386510381,
+      "loss": 0.6041,
+      "step": 30380
+    },
+    {
+      "epoch": 1.5093871063872057,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0006792530048673885,
+      "loss": 0.6189,
+      "step": 30390
+    },
+    {
+      "epoch": 1.5098837786828252,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.000679213271083739,
+      "loss": 0.6155,
+      "step": 30400
+    },
+    {
+      "epoch": 1.5103804509784444,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006791735373000894,
+      "loss": 0.596,
+      "step": 30410
+    },
+    {
+      "epoch": 1.5108771232740636,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006791338035164399,
+      "loss": 0.614,
+      "step": 30420
+    },
+    {
+      "epoch": 1.511373795569683,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006790940697327903,
+      "loss": 0.5913,
+      "step": 30430
+    },
+    {
+      "epoch": 1.5118704678653025,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006790543359491408,
+      "loss": 0.6084,
+      "step": 30440
+    },
+    {
+      "epoch": 1.5123671401609218,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006790146021654912,
+      "loss": 0.5904,
+      "step": 30450
+    },
+    {
+      "epoch": 1.5128638124565412,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0006789748683818416,
+      "loss": 0.5973,
+      "step": 30460
+    },
+    {
+      "epoch": 1.5133604847521607,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006789351345981922,
+      "loss": 0.6363,
+      "step": 30470
+    },
+    {
+      "epoch": 1.51385715704778,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0006788954008145427,
+      "loss": 0.617,
+      "step": 30480
+    },
+    {
+      "epoch": 1.5143538293433991,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.000678855667030893,
+      "loss": 0.6019,
+      "step": 30490
+    },
+    {
+      "epoch": 1.5148505016390186,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006788159332472435,
+      "loss": 0.5968,
+      "step": 30500
+    },
+    {
+      "epoch": 1.515347173934638,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0006787761994635939,
+      "loss": 0.6072,
+      "step": 30510
+    },
+    {
+      "epoch": 1.5158438462302573,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006787364656799444,
+      "loss": 0.6382,
+      "step": 30520
+    },
+    {
+      "epoch": 1.5163405185258765,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0006786967318962949,
+      "loss": 0.5835,
+      "step": 30530
+    },
+    {
+      "epoch": 1.516837190821496,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0006786569981126453,
+      "loss": 0.6065,
+      "step": 30540
+    },
+    {
+      "epoch": 1.5173338631171154,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006786172643289957,
+      "loss": 0.6027,
+      "step": 30550
+    },
+    {
+      "epoch": 1.5178305354127346,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0006785775305453462,
+      "loss": 0.5881,
+      "step": 30560
+    },
+    {
+      "epoch": 1.518327207708354,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0006785377967616967,
+      "loss": 0.5985,
+      "step": 30570
+    },
+    {
+      "epoch": 1.5188238800039735,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006784980629780471,
+      "loss": 0.5662,
+      "step": 30580
+    },
+    {
+      "epoch": 1.5193205522995927,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0006784583291943976,
+      "loss": 0.5797,
+      "step": 30590
+    },
+    {
+      "epoch": 1.519817224595212,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.000678418595410748,
+      "loss": 0.565,
+      "step": 30600
+    },
+    {
+      "epoch": 1.5203138968908314,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0006783788616270985,
+      "loss": 0.5841,
+      "step": 30610
+    },
+    {
+      "epoch": 1.5208105691864509,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.000678339127843449,
+      "loss": 0.6007,
+      "step": 30620
+    },
+    {
+      "epoch": 1.52130724148207,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0006782993940597994,
+      "loss": 0.5948,
+      "step": 30630
+    },
+    {
+      "epoch": 1.5218039137776895,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006782596602761499,
+      "loss": 0.5944,
+      "step": 30640
+    },
+    {
+      "epoch": 1.522300586073309,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0006782199264925002,
+      "loss": 0.6135,
+      "step": 30650
+    },
+    {
+      "epoch": 1.5227972583689282,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0006781801927088507,
+      "loss": 0.6057,
+      "step": 30660
+    },
+    {
+      "epoch": 1.5232939306645474,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006781404589252013,
+      "loss": 0.6222,
+      "step": 30670
+    },
+    {
+      "epoch": 1.5237906029601669,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006781007251415516,
+      "loss": 0.5781,
+      "step": 30680
+    },
+    {
+      "epoch": 1.5242872752557863,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0006780609913579021,
+      "loss": 0.6204,
+      "step": 30690
+    },
+    {
+      "epoch": 1.5247839475514056,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006780212575742525,
+      "loss": 0.6038,
+      "step": 30700
+    },
+    {
+      "epoch": 1.5252806198470248,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006779815237906029,
+      "loss": 0.6073,
+      "step": 30710
+    },
+    {
+      "epoch": 1.5257772921426442,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006779417900069535,
+      "loss": 0.6317,
+      "step": 30720
+    },
+    {
+      "epoch": 1.5262739644382637,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006779020562233039,
+      "loss": 0.6054,
+      "step": 30730
+    },
+    {
+      "epoch": 1.526770636733883,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006778623224396543,
+      "loss": 0.5999,
+      "step": 30740
+    },
+    {
+      "epoch": 1.5272673090295024,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0006778225886560048,
+      "loss": 0.6027,
+      "step": 30750
+    },
+    {
+      "epoch": 1.5277639813251218,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006777828548723552,
+      "loss": 0.6004,
+      "step": 30760
+    },
+    {
+      "epoch": 1.528260653620741,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006777431210887058,
+      "loss": 0.5881,
+      "step": 30770
+    },
+    {
+      "epoch": 1.5287573259163603,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006777033873050562,
+      "loss": 0.6004,
+      "step": 30780
+    },
+    {
+      "epoch": 1.5292539982119797,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0006776636535214066,
+      "loss": 0.6021,
+      "step": 30790
+    },
+    {
+      "epoch": 1.5297506705075992,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0006776239197377571,
+      "loss": 0.6088,
+      "step": 30800
+    },
+    {
+      "epoch": 1.5302473428032184,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0006775841859541075,
+      "loss": 0.5897,
+      "step": 30810
+    },
+    {
+      "epoch": 1.5307440150988378,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.000677544452170458,
+      "loss": 0.5908,
+      "step": 30820
+    },
+    {
+      "epoch": 1.5312406873944573,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006775047183868085,
+      "loss": 0.5859,
+      "step": 30830
+    },
+    {
+      "epoch": 1.5317373596900765,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006774649846031588,
+      "loss": 0.5808,
+      "step": 30840
+    },
+    {
+      "epoch": 1.5322340319856957,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0006774252508195093,
+      "loss": 0.6055,
+      "step": 30850
+    },
+    {
+      "epoch": 1.5327307042813152,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0006773855170358598,
+      "loss": 0.5918,
+      "step": 30860
+    },
+    {
+      "epoch": 1.5332273765769346,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006773457832522101,
+      "loss": 0.6138,
+      "step": 30870
+    },
+    {
+      "epoch": 1.5337240488725539,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0006773060494685607,
+      "loss": 0.6295,
+      "step": 30880
+    },
+    {
+      "epoch": 1.534220721168173,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006772663156849112,
+      "loss": 0.6049,
+      "step": 30890
+    },
+    {
+      "epoch": 1.5347173934637925,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0006772265819012615,
+      "loss": 0.5895,
+      "step": 30900
+    },
+    {
+      "epoch": 1.535214065759412,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.000677186848117612,
+      "loss": 0.6292,
+      "step": 30910
+    },
+    {
+      "epoch": 1.5357107380550312,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006771471143339625,
+      "loss": 0.6287,
+      "step": 30920
+    },
+    {
+      "epoch": 1.5362074103506507,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.000677107380550313,
+      "loss": 0.5945,
+      "step": 30930
+    },
+    {
+      "epoch": 1.5367040826462701,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006770676467666634,
+      "loss": 0.6321,
+      "step": 30940
+    },
+    {
+      "epoch": 1.5372007549418893,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0006770279129830138,
+      "loss": 0.6027,
+      "step": 30950
+    },
+    {
+      "epoch": 1.5376974272375086,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006769881791993643,
+      "loss": 0.6035,
+      "step": 30960
+    },
+    {
+      "epoch": 1.538194099533128,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006769484454157148,
+      "loss": 0.6127,
+      "step": 30970
+    },
+    {
+      "epoch": 1.5386907718287475,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006769087116320652,
+      "loss": 0.5827,
+      "step": 30980
+    },
+    {
+      "epoch": 1.5391874441243667,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006768689778484157,
+      "loss": 0.617,
+      "step": 30990
+    },
+    {
+      "epoch": 1.5396841164199861,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006768292440647661,
+      "loss": 0.6028,
+      "step": 31000
+    },
+    {
+      "epoch": 1.5401807887156056,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006767895102811165,
+      "loss": 0.5821,
+      "step": 31010
+    },
+    {
+      "epoch": 1.5406774610112248,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0006767497764974671,
+      "loss": 0.602,
+      "step": 31020
+    },
+    {
+      "epoch": 1.541174133306844,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006767100427138174,
+      "loss": 0.6195,
+      "step": 31030
+    },
+    {
+      "epoch": 1.5416708056024635,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006766703089301679,
+      "loss": 0.6345,
+      "step": 31040
+    },
+    {
+      "epoch": 1.542167477898083,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0006766305751465184,
+      "loss": 0.6286,
+      "step": 31050
+    },
+    {
+      "epoch": 1.5426641501937022,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0006765908413628688,
+      "loss": 0.607,
+      "step": 31060
+    },
+    {
+      "epoch": 1.5431608224893214,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006765511075792192,
+      "loss": 0.6033,
+      "step": 31070
+    },
+    {
+      "epoch": 1.5436574947849409,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006765113737955698,
+      "loss": 0.5816,
+      "step": 31080
+    },
+    {
+      "epoch": 1.5441541670805603,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006764716400119202,
+      "loss": 0.6319,
+      "step": 31090
+    },
+    {
+      "epoch": 1.5446508393761795,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0006764319062282706,
+      "loss": 0.5986,
+      "step": 31100
+    },
+    {
+      "epoch": 1.545147511671799,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.000676392172444621,
+      "loss": 0.6509,
+      "step": 31110
+    },
+    {
+      "epoch": 1.5456441839674184,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0006763524386609716,
+      "loss": 0.5765,
+      "step": 31120
+    },
+    {
+      "epoch": 1.5461408562630377,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.000676312704877322,
+      "loss": 0.5904,
+      "step": 31130
+    },
+    {
+      "epoch": 1.5466375285586569,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0006762729710936724,
+      "loss": 0.5956,
+      "step": 31140
+    },
+    {
+      "epoch": 1.5471342008542763,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0006762332373100229,
+      "loss": 0.5829,
+      "step": 31150
+    },
+    {
+      "epoch": 1.5476308731498958,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0006761935035263733,
+      "loss": 0.5784,
+      "step": 31160
+    },
+    {
+      "epoch": 1.548127545445515,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0006761537697427237,
+      "loss": 0.6014,
+      "step": 31170
+    },
+    {
+      "epoch": 1.5486242177411342,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0006761140359590743,
+      "loss": 0.6188,
+      "step": 31180
+    },
+    {
+      "epoch": 1.549120890036754,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006760743021754247,
+      "loss": 0.6035,
+      "step": 31190
+    },
+    {
+      "epoch": 1.5496175623323731,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006760345683917751,
+      "loss": 0.5998,
+      "step": 31200
+    },
+    {
+      "epoch": 1.5501142346279924,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006759948346081256,
+      "loss": 0.5868,
+      "step": 31210
+    },
+    {
+      "epoch": 1.5506109069236118,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.000675955100824476,
+      "loss": 0.6067,
+      "step": 31220
+    },
+    {
+      "epoch": 1.5511075792192313,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006759153670408265,
+      "loss": 0.6151,
+      "step": 31230
+    },
+    {
+      "epoch": 1.5516042515148505,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.000675875633257177,
+      "loss": 0.6484,
+      "step": 31240
+    },
+    {
+      "epoch": 1.5521009238104697,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0006758358994735274,
+      "loss": 0.6143,
+      "step": 31250
+    },
+    {
+      "epoch": 1.5525975961060892,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0006757961656898778,
+      "loss": 0.5907,
+      "step": 31260
+    },
+    {
+      "epoch": 1.5530942684017086,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0006757564319062284,
+      "loss": 0.6016,
+      "step": 31270
+    },
+    {
+      "epoch": 1.5535909406973278,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0006757166981225788,
+      "loss": 0.6084,
+      "step": 31280
+    },
+    {
+      "epoch": 1.5540876129929473,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006756769643389292,
+      "loss": 0.6154,
+      "step": 31290
+    },
+    {
+      "epoch": 1.5545842852885667,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006756372305552796,
+      "loss": 0.6008,
+      "step": 31300
+    },
+    {
+      "epoch": 1.555080957584186,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006755974967716301,
+      "loss": 0.5986,
+      "step": 31310
+    },
+    {
+      "epoch": 1.5555776298798052,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0006755577629879805,
+      "loss": 0.5987,
+      "step": 31320
+    },
+    {
+      "epoch": 1.5560743021754246,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.000675518029204331,
+      "loss": 0.5984,
+      "step": 31330
+    },
+    {
+      "epoch": 1.556570974471044,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0006754782954206815,
+      "loss": 0.5834,
+      "step": 31340
+    },
+    {
+      "epoch": 1.5570676467666633,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0006754385616370319,
+      "loss": 0.6261,
+      "step": 31350
+    },
+    {
+      "epoch": 1.5575643190622825,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006753988278533823,
+      "loss": 0.5748,
+      "step": 31360
+    },
+    {
+      "epoch": 1.5580609913579022,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006753590940697328,
+      "loss": 0.5874,
+      "step": 31370
+    },
+    {
+      "epoch": 1.5585576636535214,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006753193602860833,
+      "loss": 0.5873,
+      "step": 31380
+    },
+    {
+      "epoch": 1.5590543359491407,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0006752796265024337,
+      "loss": 0.6083,
+      "step": 31390
+    },
+    {
+      "epoch": 1.5595510082447601,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006752398927187842,
+      "loss": 0.5917,
+      "step": 31400
+    },
+    {
+      "epoch": 1.5600476805403796,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0006752001589351346,
+      "loss": 0.6035,
+      "step": 31410
+    },
+    {
+      "epoch": 1.5605443528359988,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.000675160425151485,
+      "loss": 0.5833,
+      "step": 31420
+    },
+    {
+      "epoch": 1.561041025131618,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006751206913678356,
+      "loss": 0.5765,
+      "step": 31430
+    },
+    {
+      "epoch": 1.5615376974272375,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.000675080957584186,
+      "loss": 0.5868,
+      "step": 31440
+    },
+    {
+      "epoch": 1.562034369722857,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0006750412238005364,
+      "loss": 0.6094,
+      "step": 31450
+    },
+    {
+      "epoch": 1.5625310420184761,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0006750014900168869,
+      "loss": 0.5841,
+      "step": 31460
+    },
+    {
+      "epoch": 1.5630277143140956,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006749617562332373,
+      "loss": 0.5807,
+      "step": 31470
+    },
+    {
+      "epoch": 1.563524386609715,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006749220224495878,
+      "loss": 0.5957,
+      "step": 31480
+    },
+    {
+      "epoch": 1.5640210589053343,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0006748822886659383,
+      "loss": 0.6276,
+      "step": 31490
+    },
+    {
+      "epoch": 1.5645177312009535,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0006748425548822887,
+      "loss": 0.5916,
+      "step": 31500
+    },
+    {
+      "epoch": 1.565014403496573,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0006748028210986392,
+      "loss": 0.5754,
+      "step": 31510
+    },
+    {
+      "epoch": 1.5655110757921924,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006747630873149895,
+      "loss": 0.6124,
+      "step": 31520
+    },
+    {
+      "epoch": 1.5660077480878116,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006747233535313401,
+      "loss": 0.579,
+      "step": 31530
+    },
+    {
+      "epoch": 1.5665044203834309,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0006746836197476906,
+      "loss": 0.6369,
+      "step": 31540
+    },
+    {
+      "epoch": 1.5670010926790505,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006746438859640409,
+      "loss": 0.6231,
+      "step": 31550
+    },
+    {
+      "epoch": 1.5674977649746697,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0006746041521803914,
+      "loss": 0.6074,
+      "step": 31560
+    },
+    {
+      "epoch": 1.567994437270289,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006745644183967418,
+      "loss": 0.6241,
+      "step": 31570
+    },
+    {
+      "epoch": 1.5684911095659084,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0006745246846130923,
+      "loss": 0.5816,
+      "step": 31580
+    },
+    {
+      "epoch": 1.5689877818615279,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0006744849508294428,
+      "loss": 0.6155,
+      "step": 31590
+    },
+    {
+      "epoch": 1.569484454157147,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0006744452170457932,
+      "loss": 0.6011,
+      "step": 31600
+    },
+    {
+      "epoch": 1.5699811264527663,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006744054832621436,
+      "loss": 0.6116,
+      "step": 31610
+    },
+    {
+      "epoch": 1.5704777987483858,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0006743657494784941,
+      "loss": 0.5968,
+      "step": 31620
+    },
+    {
+      "epoch": 1.5709744710440052,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0006743260156948446,
+      "loss": 0.5831,
+      "step": 31630
+    },
+    {
+      "epoch": 1.5714711433396245,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.000674286281911195,
+      "loss": 0.614,
+      "step": 31640
+    },
+    {
+      "epoch": 1.571967815635244,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0006742465481275455,
+      "loss": 0.5724,
+      "step": 31650
+    },
+    {
+      "epoch": 1.5724644879308634,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0006742068143438959,
+      "loss": 0.5828,
+      "step": 31660
+    },
+    {
+      "epoch": 1.5729611602264826,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006741670805602464,
+      "loss": 0.5952,
+      "step": 31670
+    },
+    {
+      "epoch": 1.5734578325221018,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0006741273467765969,
+      "loss": 0.6089,
+      "step": 31680
+    },
+    {
+      "epoch": 1.5739545048177213,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0006740876129929473,
+      "loss": 0.6162,
+      "step": 31690
+    },
+    {
+      "epoch": 1.5744511771133407,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006740478792092978,
+      "loss": 0.5958,
+      "step": 31700
+    },
+    {
+      "epoch": 1.57494784940896,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006740081454256481,
+      "loss": 0.5772,
+      "step": 31710
+    },
+    {
+      "epoch": 1.5754445217045792,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006739684116419986,
+      "loss": 0.594,
+      "step": 31720
+    },
+    {
+      "epoch": 1.5759411940001988,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0006739286778583492,
+      "loss": 0.6195,
+      "step": 31730
+    },
+    {
+      "epoch": 1.576437866295818,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006738889440746995,
+      "loss": 0.6125,
+      "step": 31740
+    },
+    {
+      "epoch": 1.5769345385914373,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00067384921029105,
+      "loss": 0.6229,
+      "step": 31750
+    },
+    {
+      "epoch": 1.5774312108870567,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0006738094765074005,
+      "loss": 0.5876,
+      "step": 31760
+    },
+    {
+      "epoch": 1.5779278831826762,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006737697427237508,
+      "loss": 0.5819,
+      "step": 31770
+    },
+    {
+      "epoch": 1.5784245554782954,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0006737300089401014,
+      "loss": 0.59,
+      "step": 31780
+    },
+    {
+      "epoch": 1.5789212277739146,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006736902751564518,
+      "loss": 0.6214,
+      "step": 31790
+    },
+    {
+      "epoch": 1.579417900069534,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0006736505413728022,
+      "loss": 0.6092,
+      "step": 31800
+    },
+    {
+      "epoch": 1.5799145723651535,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0006736108075891527,
+      "loss": 0.601,
+      "step": 31810
+    },
+    {
+      "epoch": 1.5804112446607728,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006735710738055031,
+      "loss": 0.6004,
+      "step": 31820
+    },
+    {
+      "epoch": 1.5809079169563922,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0006735313400218537,
+      "loss": 0.5944,
+      "step": 31830
+    },
+    {
+      "epoch": 1.5814045892520117,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006734916062382041,
+      "loss": 0.6069,
+      "step": 31840
+    },
+    {
+      "epoch": 1.5819012615476309,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006734518724545545,
+      "loss": 0.5988,
+      "step": 31850
+    },
+    {
+      "epoch": 1.5823979338432501,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.000673412138670905,
+      "loss": 0.5958,
+      "step": 31860
+    },
+    {
+      "epoch": 1.5828946061388696,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0006733724048872554,
+      "loss": 0.5858,
+      "step": 31870
+    },
+    {
+      "epoch": 1.583391278434489,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006733326711036059,
+      "loss": 0.6077,
+      "step": 31880
+    },
+    {
+      "epoch": 1.5838879507301082,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0006732929373199564,
+      "loss": 0.5997,
+      "step": 31890
+    },
+    {
+      "epoch": 1.5843846230257275,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0006732532035363067,
+      "loss": 0.6122,
+      "step": 31900
+    },
+    {
+      "epoch": 1.5848812953213471,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0006732134697526572,
+      "loss": 0.5781,
+      "step": 31910
+    },
+    {
+      "epoch": 1.5853779676169664,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006731737359690077,
+      "loss": 0.5809,
+      "step": 31920
+    },
+    {
+      "epoch": 1.5858746399125856,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.000673134002185358,
+      "loss": 0.6108,
+      "step": 31930
+    },
+    {
+      "epoch": 1.586371312208205,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006730942684017086,
+      "loss": 0.604,
+      "step": 31940
+    },
+    {
+      "epoch": 1.5868679845038245,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006730545346180591,
+      "loss": 0.6119,
+      "step": 31950
+    },
+    {
+      "epoch": 1.5873646567994437,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0006730148008344095,
+      "loss": 0.5925,
+      "step": 31960
+    },
+    {
+      "epoch": 1.587861329095063,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006729750670507599,
+      "loss": 0.588,
+      "step": 31970
+    },
+    {
+      "epoch": 1.5883580013906824,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0006729353332671104,
+      "loss": 0.6016,
+      "step": 31980
+    },
+    {
+      "epoch": 1.5888546736863018,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0006728955994834609,
+      "loss": 0.6076,
+      "step": 31990
+    },
+    {
+      "epoch": 1.589351345981921,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006728558656998113,
+      "loss": 0.6245,
+      "step": 32000
+    },
+    {
+      "epoch": 1.5898480182775405,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0006728161319161617,
+      "loss": 0.5935,
+      "step": 32010
+    },
+    {
+      "epoch": 1.59034469057316,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0006727763981325122,
+      "loss": 0.6305,
+      "step": 32020
+    },
+    {
+      "epoch": 1.5908413628687792,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0006727366643488627,
+      "loss": 0.5974,
+      "step": 32030
+    },
+    {
+      "epoch": 1.5913380351643984,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0006726969305652131,
+      "loss": 0.5709,
+      "step": 32040
+    },
+    {
+      "epoch": 1.5918347074600179,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006726571967815636,
+      "loss": 0.5996,
+      "step": 32050
+    },
+    {
+      "epoch": 1.5923313797556373,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.000672617462997914,
+      "loss": 0.6167,
+      "step": 32060
+    },
+    {
+      "epoch": 1.5928280520512565,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0006725777292142644,
+      "loss": 0.6354,
+      "step": 32070
+    },
+    {
+      "epoch": 1.5933247243468758,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.000672537995430615,
+      "loss": 0.5934,
+      "step": 32080
+    },
+    {
+      "epoch": 1.5938213966424954,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006724982616469653,
+      "loss": 0.5804,
+      "step": 32090
+    },
+    {
+      "epoch": 1.5943180689381147,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006724585278633158,
+      "loss": 0.5837,
+      "step": 32100
+    },
+    {
+      "epoch": 1.594814741233734,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006724187940796663,
+      "loss": 0.5765,
+      "step": 32110
+    },
+    {
+      "epoch": 1.5953114135293534,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006723790602960167,
+      "loss": 0.6098,
+      "step": 32120
+    },
+    {
+      "epoch": 1.5958080858249728,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006723393265123672,
+      "loss": 0.5863,
+      "step": 32130
+    },
+    {
+      "epoch": 1.596304758120592,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006722995927287177,
+      "loss": 0.5996,
+      "step": 32140
+    },
+    {
+      "epoch": 1.5968014304162113,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0006722598589450681,
+      "loss": 0.6006,
+      "step": 32150
+    },
+    {
+      "epoch": 1.5972981027118307,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006722201251614185,
+      "loss": 0.5707,
+      "step": 32160
+    },
+    {
+      "epoch": 1.5977947750074502,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006721803913777689,
+      "loss": 0.5769,
+      "step": 32170
+    },
+    {
+      "epoch": 1.5982914473030694,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006721406575941195,
+      "loss": 0.607,
+      "step": 32180
+    },
+    {
+      "epoch": 1.5987881195986888,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006721009238104699,
+      "loss": 0.5976,
+      "step": 32190
+    },
+    {
+      "epoch": 1.5992847918943083,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006720611900268203,
+      "loss": 0.6053,
+      "step": 32200
+    },
+    {
+      "epoch": 1.5997814641899275,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006720214562431708,
+      "loss": 0.6023,
+      "step": 32210
+    },
+    {
+      "epoch": 1.6002781364855467,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006719817224595212,
+      "loss": 0.6075,
+      "step": 32220
+    },
+    {
+      "epoch": 1.6007748087811662,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006719419886758716,
+      "loss": 0.5845,
+      "step": 32230
+    },
+    {
+      "epoch": 1.6012714810767856,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006719022548922222,
+      "loss": 0.6106,
+      "step": 32240
+    },
+    {
+      "epoch": 1.6017681533724049,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006718625211085726,
+      "loss": 0.6036,
+      "step": 32250
+    },
+    {
+      "epoch": 1.602264825668024,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.000671822787324923,
+      "loss": 0.6072,
+      "step": 32260
+    },
+    {
+      "epoch": 1.6027614979636435,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006717830535412735,
+      "loss": 0.5904,
+      "step": 32270
+    },
+    {
+      "epoch": 1.603258170259263,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.000671743319757624,
+      "loss": 0.5886,
+      "step": 32280
+    },
+    {
+      "epoch": 1.6037548425548822,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006717035859739744,
+      "loss": 0.6086,
+      "step": 32290
+    },
+    {
+      "epoch": 1.6042515148505017,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0006716638521903249,
+      "loss": 0.6034,
+      "step": 32300
+    },
+    {
+      "epoch": 1.604748187146121,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0006716241184066753,
+      "loss": 0.6138,
+      "step": 32310
+    },
+    {
+      "epoch": 1.6052448594417403,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0006715843846230257,
+      "loss": 0.614,
+      "step": 32320
+    },
+    {
+      "epoch": 1.6057415317373596,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0006715446508393763,
+      "loss": 0.5995,
+      "step": 32330
+    },
+    {
+      "epoch": 1.606238204032979,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006715049170557267,
+      "loss": 0.5998,
+      "step": 32340
+    },
+    {
+      "epoch": 1.6067348763285985,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006714651832720771,
+      "loss": 0.5739,
+      "step": 32350
+    },
+    {
+      "epoch": 1.6072315486242177,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0006714254494884275,
+      "loss": 0.5946,
+      "step": 32360
+    },
+    {
+      "epoch": 1.6077282209198371,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.000671385715704778,
+      "loss": 0.5991,
+      "step": 32370
+    },
+    {
+      "epoch": 1.6082248932154566,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006713459819211284,
+      "loss": 0.612,
+      "step": 32380
+    },
+    {
+      "epoch": 1.6087215655110758,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0006713062481374789,
+      "loss": 0.5953,
+      "step": 32390
+    },
+    {
+      "epoch": 1.609218237806695,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006712665143538294,
+      "loss": 0.5863,
+      "step": 32400
+    },
+    {
+      "epoch": 1.6097149101023145,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0006712267805701799,
+      "loss": 0.5835,
+      "step": 32410
+    },
+    {
+      "epoch": 1.610211582397934,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006711870467865302,
+      "loss": 0.6002,
+      "step": 32420
+    },
+    {
+      "epoch": 1.6107082546935532,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006711473130028808,
+      "loss": 0.5809,
+      "step": 32430
+    },
+    {
+      "epoch": 1.6112049269891724,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006711075792192312,
+      "loss": 0.6159,
+      "step": 32440
+    },
+    {
+      "epoch": 1.6117015992847918,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0006710678454355816,
+      "loss": 0.6001,
+      "step": 32450
+    },
+    {
+      "epoch": 1.6121982715804113,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006710281116519321,
+      "loss": 0.625,
+      "step": 32460
+    },
+    {
+      "epoch": 1.6126949438760305,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006709883778682825,
+      "loss": 0.6024,
+      "step": 32470
+    },
+    {
+      "epoch": 1.61319161617165,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.000670948644084633,
+      "loss": 0.5837,
+      "step": 32480
+    },
+    {
+      "epoch": 1.6136882884672694,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0006709089103009835,
+      "loss": 0.5987,
+      "step": 32490
+    },
+    {
+      "epoch": 1.6141849607628886,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0006708691765173339,
+      "loss": 0.5843,
+      "step": 32500
+    },
+    {
+      "epoch": 1.6146816330585079,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0006708294427336843,
+      "loss": 0.5891,
+      "step": 32510
+    },
+    {
+      "epoch": 1.6151783053541273,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0006707897089500348,
+      "loss": 0.593,
+      "step": 32520
+    },
+    {
+      "epoch": 1.6156749776497468,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006707499751663852,
+      "loss": 0.6336,
+      "step": 32530
+    },
+    {
+      "epoch": 1.616171649945366,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0006707102413827357,
+      "loss": 0.5974,
+      "step": 32540
+    },
+    {
+      "epoch": 1.6166683222409854,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006706705075990862,
+      "loss": 0.5764,
+      "step": 32550
+    },
+    {
+      "epoch": 1.617164994536605,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0006706307738154366,
+      "loss": 0.6318,
+      "step": 32560
+    },
+    {
+      "epoch": 1.6176616668322241,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0006705910400317871,
+      "loss": 0.5853,
+      "step": 32570
+    },
+    {
+      "epoch": 1.6181583391278433,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006705513062481374,
+      "loss": 0.5914,
+      "step": 32580
+    },
+    {
+      "epoch": 1.6186550114234628,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.000670511572464488,
+      "loss": 0.5874,
+      "step": 32590
+    },
+    {
+      "epoch": 1.6191516837190822,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006704718386808385,
+      "loss": 0.5953,
+      "step": 32600
+    },
+    {
+      "epoch": 1.6196483560147015,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006704321048971888,
+      "loss": 0.5655,
+      "step": 32610
+    },
+    {
+      "epoch": 1.6201450283103207,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0006703923711135393,
+      "loss": 0.6293,
+      "step": 32620
+    },
+    {
+      "epoch": 1.6206417006059401,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006703526373298897,
+      "loss": 0.5849,
+      "step": 32630
+    },
+    {
+      "epoch": 1.6211383729015596,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006703129035462402,
+      "loss": 0.5957,
+      "step": 32640
+    },
+    {
+      "epoch": 1.6216350451971788,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0006702731697625907,
+      "loss": 0.5986,
+      "step": 32650
+    },
+    {
+      "epoch": 1.6221317174927983,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006702334359789411,
+      "loss": 0.6059,
+      "step": 32660
+    },
+    {
+      "epoch": 1.6226283897884177,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006701937021952915,
+      "loss": 0.6147,
+      "step": 32670
+    },
+    {
+      "epoch": 1.623125062084037,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.000670153968411642,
+      "loss": 0.602,
+      "step": 32680
+    },
+    {
+      "epoch": 1.6236217343796562,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006701142346279925,
+      "loss": 0.5796,
+      "step": 32690
+    },
+    {
+      "epoch": 1.6241184066752756,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.000670074500844343,
+      "loss": 0.6129,
+      "step": 32700
+    },
+    {
+      "epoch": 1.624615078970895,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006700347670606934,
+      "loss": 0.5866,
+      "step": 32710
+    },
+    {
+      "epoch": 1.6251117512665143,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0006699950332770438,
+      "loss": 0.609,
+      "step": 32720
+    },
+    {
+      "epoch": 1.6256084235621338,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006699552994933944,
+      "loss": 0.639,
+      "step": 32730
+    },
+    {
+      "epoch": 1.6261050958577532,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0006699155657097448,
+      "loss": 0.6427,
+      "step": 32740
+    },
+    {
+      "epoch": 1.6266017681533724,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0006698758319260952,
+      "loss": 0.5993,
+      "step": 32750
+    },
+    {
+      "epoch": 1.6270984404489917,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006698360981424457,
+      "loss": 0.5942,
+      "step": 32760
+    },
+    {
+      "epoch": 1.627595112744611,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.000669796364358796,
+      "loss": 0.5955,
+      "step": 32770
+    },
+    {
+      "epoch": 1.6280917850402306,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0006697566305751465,
+      "loss": 0.6021,
+      "step": 32780
+    },
+    {
+      "epoch": 1.6285884573358498,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006697168967914971,
+      "loss": 0.608,
+      "step": 32790
+    },
+    {
+      "epoch": 1.629085129631469,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0006696771630078474,
+      "loss": 0.572,
+      "step": 32800
+    },
+    {
+      "epoch": 1.6295818019270885,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006696374292241979,
+      "loss": 0.6285,
+      "step": 32810
+    },
+    {
+      "epoch": 1.630078474222708,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006695976954405484,
+      "loss": 0.5893,
+      "step": 32820
+    },
+    {
+      "epoch": 1.6305751465183271,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0006695579616568987,
+      "loss": 0.5894,
+      "step": 32830
+    },
+    {
+      "epoch": 1.6310718188139466,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006695182278732493,
+      "loss": 0.6047,
+      "step": 32840
+    },
+    {
+      "epoch": 1.631568491109566,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0006694784940895997,
+      "loss": 0.6013,
+      "step": 32850
+    },
+    {
+      "epoch": 1.6320651634051853,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0006694387603059502,
+      "loss": 0.6161,
+      "step": 32860
+    },
+    {
+      "epoch": 1.6325618357008045,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0006693990265223006,
+      "loss": 0.583,
+      "step": 32870
+    },
+    {
+      "epoch": 1.633058507996424,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000669359292738651,
+      "loss": 0.6175,
+      "step": 32880
+    },
+    {
+      "epoch": 1.6335551802920434,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0006693195589550016,
+      "loss": 0.5876,
+      "step": 32890
+    },
+    {
+      "epoch": 1.6340518525876626,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.000669279825171352,
+      "loss": 0.5955,
+      "step": 32900
+    },
+    {
+      "epoch": 1.634548524883282,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006692400913877024,
+      "loss": 0.5851,
+      "step": 32910
+    },
+    {
+      "epoch": 1.6350451971789015,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0006692003576040529,
+      "loss": 0.6354,
+      "step": 32920
+    },
+    {
+      "epoch": 1.6355418694745207,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006691606238204033,
+      "loss": 0.5891,
+      "step": 32930
+    },
+    {
+      "epoch": 1.63603854177014,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006691208900367538,
+      "loss": 0.6053,
+      "step": 32940
+    },
+    {
+      "epoch": 1.6365352140657594,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006690811562531043,
+      "loss": 0.5908,
+      "step": 32950
+    },
+    {
+      "epoch": 1.6370318863613789,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0006690414224694546,
+      "loss": 0.5854,
+      "step": 32960
+    },
+    {
+      "epoch": 1.637528558656998,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0006690016886858051,
+      "loss": 0.5946,
+      "step": 32970
+    },
+    {
+      "epoch": 1.6380252309526173,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0006689619549021556,
+      "loss": 0.6006,
+      "step": 32980
+    },
+    {
+      "epoch": 1.6385219032482368,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.000668922221118506,
+      "loss": 0.5873,
+      "step": 32990
+    },
+    {
+      "epoch": 1.6390185755438562,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006688824873348565,
+      "loss": 0.6015,
+      "step": 33000
+    },
+    {
+      "epoch": 1.6395152478394754,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.000668842753551207,
+      "loss": 0.5932,
+      "step": 33010
+    },
+    {
+      "epoch": 1.640011920135095,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0006688030197675574,
+      "loss": 0.601,
+      "step": 33020
+    },
+    {
+      "epoch": 1.6405085924307143,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006687632859839078,
+      "loss": 0.5883,
+      "step": 33030
+    },
+    {
+      "epoch": 1.6410052647263336,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006687235522002583,
+      "loss": 0.619,
+      "step": 33040
+    },
+    {
+      "epoch": 1.6415019370219528,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0006686838184166088,
+      "loss": 0.5785,
+      "step": 33050
+    },
+    {
+      "epoch": 1.6419986093175722,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0006686440846329592,
+      "loss": 0.6193,
+      "step": 33060
+    },
+    {
+      "epoch": 1.6424952816131917,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006686043508493096,
+      "loss": 0.6014,
+      "step": 33070
+    },
+    {
+      "epoch": 1.642991953908811,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006685646170656601,
+      "loss": 0.6019,
+      "step": 33080
+    },
+    {
+      "epoch": 1.6434886262044304,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006685248832820106,
+      "loss": 0.5937,
+      "step": 33090
+    },
+    {
+      "epoch": 1.6439852985000498,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.000668485149498361,
+      "loss": 0.6061,
+      "step": 33100
+    },
+    {
+      "epoch": 1.644481970795669,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0006684454157147115,
+      "loss": 0.571,
+      "step": 33110
+    },
+    {
+      "epoch": 1.6449786430912883,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0006684056819310619,
+      "loss": 0.6038,
+      "step": 33120
+    },
+    {
+      "epoch": 1.6454753153869077,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0006683659481474123,
+      "loss": 0.5675,
+      "step": 33130
+    },
+    {
+      "epoch": 1.6459719876825272,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0006683262143637629,
+      "loss": 0.6265,
+      "step": 33140
+    },
+    {
+      "epoch": 1.6464686599781464,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0006682864805801133,
+      "loss": 0.6078,
+      "step": 33150
+    },
+    {
+      "epoch": 1.6469653322737656,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006682467467964637,
+      "loss": 0.5742,
+      "step": 33160
+    },
+    {
+      "epoch": 1.647462004569385,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006682070130128142,
+      "loss": 0.5663,
+      "step": 33170
+    },
+    {
+      "epoch": 1.6479586768650045,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0006681672792291646,
+      "loss": 0.5814,
+      "step": 33180
+    },
+    {
+      "epoch": 1.6484553491606238,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006681275454455151,
+      "loss": 0.6041,
+      "step": 33190
+    },
+    {
+      "epoch": 1.6489520214562432,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0006680878116618656,
+      "loss": 0.5971,
+      "step": 33200
+    },
+    {
+      "epoch": 1.6494486937518626,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.000668048077878216,
+      "loss": 0.5873,
+      "step": 33210
+    },
+    {
+      "epoch": 1.6499453660474819,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006680083440945664,
+      "loss": 0.5809,
+      "step": 33220
+    },
+    {
+      "epoch": 1.650442038343101,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006679686103109168,
+      "loss": 0.5429,
+      "step": 33230
+    },
+    {
+      "epoch": 1.6509387106387206,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0006679288765272674,
+      "loss": 0.5753,
+      "step": 33240
+    },
+    {
+      "epoch": 1.65143538293434,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006678891427436178,
+      "loss": 0.5938,
+      "step": 33250
+    },
+    {
+      "epoch": 1.6519320552299592,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006678494089599682,
+      "loss": 0.6029,
+      "step": 33260
+    },
+    {
+      "epoch": 1.6524287275255787,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0006678096751763187,
+      "loss": 0.5796,
+      "step": 33270
+    },
+    {
+      "epoch": 1.6529253998211981,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006677699413926691,
+      "loss": 0.5814,
+      "step": 33280
+    },
+    {
+      "epoch": 1.6534220721168174,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0006677302076090196,
+      "loss": 0.5828,
+      "step": 33290
+    },
+    {
+      "epoch": 1.6539187444124366,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006676904738253701,
+      "loss": 0.5606,
+      "step": 33300
+    },
+    {
+      "epoch": 1.654415416708056,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0006676507400417205,
+      "loss": 0.6069,
+      "step": 33310
+    },
+    {
+      "epoch": 1.6549120890036755,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0006676110062580709,
+      "loss": 0.5975,
+      "step": 33320
+    },
+    {
+      "epoch": 1.6554087612992947,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006675712724744214,
+      "loss": 0.587,
+      "step": 33330
+    },
+    {
+      "epoch": 1.655905433594914,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006675315386907719,
+      "loss": 0.576,
+      "step": 33340
+    },
+    {
+      "epoch": 1.6564021058905334,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0006674918049071223,
+      "loss": 0.6055,
+      "step": 33350
+    },
+    {
+      "epoch": 1.6568987781861528,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0006674520711234728,
+      "loss": 0.5697,
+      "step": 33360
+    },
+    {
+      "epoch": 1.657395450481772,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0006674123373398232,
+      "loss": 0.5907,
+      "step": 33370
+    },
+    {
+      "epoch": 1.6578921227773915,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006673726035561736,
+      "loss": 0.5925,
+      "step": 33380
+    },
+    {
+      "epoch": 1.658388795073011,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006673328697725242,
+      "loss": 0.6118,
+      "step": 33390
+    },
+    {
+      "epoch": 1.6588854673686302,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0006672931359888746,
+      "loss": 0.6064,
+      "step": 33400
+    },
+    {
+      "epoch": 1.6593821396642494,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000667253402205225,
+      "loss": 0.5998,
+      "step": 33410
+    },
+    {
+      "epoch": 1.6598788119598689,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0006672136684215755,
+      "loss": 0.5894,
+      "step": 33420
+    },
+    {
+      "epoch": 1.6603754842554883,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0006671739346379259,
+      "loss": 0.5991,
+      "step": 33430
+    },
+    {
+      "epoch": 1.6608721565511075,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0006671342008542764,
+      "loss": 0.5902,
+      "step": 33440
+    },
+    {
+      "epoch": 1.6613688288467268,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006670944670706268,
+      "loss": 0.6083,
+      "step": 33450
+    },
+    {
+      "epoch": 1.6618655011423464,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006670547332869773,
+      "loss": 0.5757,
+      "step": 33460
+    },
+    {
+      "epoch": 1.6623621734379657,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0006670149995033278,
+      "loss": 0.5965,
+      "step": 33470
+    },
+    {
+      "epoch": 1.662858845733585,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006669752657196781,
+      "loss": 0.6245,
+      "step": 33480
+    },
+    {
+      "epoch": 1.6633555180292043,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0006669355319360287,
+      "loss": 0.6389,
+      "step": 33490
+    },
+    {
+      "epoch": 1.6638521903248238,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006668957981523791,
+      "loss": 0.5866,
+      "step": 33500
+    },
+    {
+      "epoch": 1.664348862620443,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006668560643687295,
+      "loss": 0.5713,
+      "step": 33510
+    },
+    {
+      "epoch": 1.6648455349160622,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00066681633058508,
+      "loss": 0.6158,
+      "step": 33520
+    },
+    {
+      "epoch": 1.6653422072116817,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0006667765968014304,
+      "loss": 0.5874,
+      "step": 33530
+    },
+    {
+      "epoch": 1.6658388795073011,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0006667368630177808,
+      "loss": 0.5814,
+      "step": 33540
+    },
+    {
+      "epoch": 1.6663355518029204,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0006666971292341314,
+      "loss": 0.586,
+      "step": 33550
+    },
+    {
+      "epoch": 1.6668322240985398,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006666573954504818,
+      "loss": 0.5789,
+      "step": 33560
+    },
+    {
+      "epoch": 1.6673288963941593,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006666176616668322,
+      "loss": 0.5843,
+      "step": 33570
+    },
+    {
+      "epoch": 1.6678255686897785,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0006665779278831827,
+      "loss": 0.5828,
+      "step": 33580
+    },
+    {
+      "epoch": 1.6683222409853977,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006665381940995332,
+      "loss": 0.5808,
+      "step": 33590
+    },
+    {
+      "epoch": 1.6688189132810172,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0006664984603158837,
+      "loss": 0.5988,
+      "step": 33600
+    },
+    {
+      "epoch": 1.6693155855766366,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0006664587265322341,
+      "loss": 0.5865,
+      "step": 33610
+    },
+    {
+      "epoch": 1.6698122578722558,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0006664189927485845,
+      "loss": 0.5933,
+      "step": 33620
+    },
+    {
+      "epoch": 1.670308930167875,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000666379258964935,
+      "loss": 0.5899,
+      "step": 33630
+    },
+    {
+      "epoch": 1.6708056024634947,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0006663395251812853,
+      "loss": 0.5751,
+      "step": 33640
+    },
+    {
+      "epoch": 1.671302274759114,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006662997913976359,
+      "loss": 0.5719,
+      "step": 33650
+    },
+    {
+      "epoch": 1.6717989470547332,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006662600576139864,
+      "loss": 0.5757,
+      "step": 33660
+    },
+    {
+      "epoch": 1.6722956193503526,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006662203238303367,
+      "loss": 0.6287,
+      "step": 33670
+    },
+    {
+      "epoch": 1.672792291645972,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006661805900466872,
+      "loss": 0.6213,
+      "step": 33680
+    },
+    {
+      "epoch": 1.6732889639415913,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006661408562630378,
+      "loss": 0.6159,
+      "step": 33690
+    },
+    {
+      "epoch": 1.6737856362372106,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006661011224793881,
+      "loss": 0.5905,
+      "step": 33700
+    },
+    {
+      "epoch": 1.67428230853283,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006660613886957386,
+      "loss": 0.6113,
+      "step": 33710
+    },
+    {
+      "epoch": 1.6747789808284494,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.000666021654912089,
+      "loss": 0.5731,
+      "step": 33720
+    },
+    {
+      "epoch": 1.6752756531240687,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0006659819211284394,
+      "loss": 0.5678,
+      "step": 33730
+    },
+    {
+      "epoch": 1.6757723254196881,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00066594218734479,
+      "loss": 0.5949,
+      "step": 33740
+    },
+    {
+      "epoch": 1.6762689977153076,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006659024535611404,
+      "loss": 0.5886,
+      "step": 33750
+    },
+    {
+      "epoch": 1.6767656700109268,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006658627197774909,
+      "loss": 0.5974,
+      "step": 33760
+    },
+    {
+      "epoch": 1.677262342306546,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006658229859938413,
+      "loss": 0.58,
+      "step": 33770
+    },
+    {
+      "epoch": 1.6777590146021655,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006657832522101917,
+      "loss": 0.5542,
+      "step": 33780
+    },
+    {
+      "epoch": 1.678255686897785,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0006657435184265423,
+      "loss": 0.5807,
+      "step": 33790
+    },
+    {
+      "epoch": 1.6787523591934042,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006657037846428927,
+      "loss": 0.5959,
+      "step": 33800
+    },
+    {
+      "epoch": 1.6792490314890234,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006656640508592431,
+      "loss": 0.5967,
+      "step": 33810
+    },
+    {
+      "epoch": 1.679745703784643,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006656243170755936,
+      "loss": 0.6256,
+      "step": 33820
+    },
+    {
+      "epoch": 1.6802423760802623,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006655845832919439,
+      "loss": 0.5835,
+      "step": 33830
+    },
+    {
+      "epoch": 1.6807390483758815,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006655448495082944,
+      "loss": 0.6092,
+      "step": 33840
+    },
+    {
+      "epoch": 1.681235720671501,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.000665505115724645,
+      "loss": 0.6094,
+      "step": 33850
+    },
+    {
+      "epoch": 1.6817323929671204,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0006654653819409953,
+      "loss": 0.6105,
+      "step": 33860
+    },
+    {
+      "epoch": 1.6822290652627396,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0006654256481573458,
+      "loss": 0.632,
+      "step": 33870
+    },
+    {
+      "epoch": 1.6827257375583589,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006653859143736963,
+      "loss": 0.6004,
+      "step": 33880
+    },
+    {
+      "epoch": 1.6832224098539783,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006653461805900466,
+      "loss": 0.5981,
+      "step": 33890
+    },
+    {
+      "epoch": 1.6837190821495978,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0006653064468063972,
+      "loss": 0.5944,
+      "step": 33900
+    },
+    {
+      "epoch": 1.684215754445217,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006652667130227476,
+      "loss": 0.599,
+      "step": 33910
+    },
+    {
+      "epoch": 1.6847124267408364,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0006652269792390981,
+      "loss": 0.5958,
+      "step": 33920
+    },
+    {
+      "epoch": 1.6852090990364559,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006651872454554485,
+      "loss": 0.5523,
+      "step": 33930
+    },
+    {
+      "epoch": 1.685705771332075,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0006651475116717989,
+      "loss": 0.5817,
+      "step": 33940
+    },
+    {
+      "epoch": 1.6862024436276943,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006651077778881495,
+      "loss": 0.579,
+      "step": 33950
+    },
+    {
+      "epoch": 1.6866991159233138,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0006650680441044999,
+      "loss": 0.6026,
+      "step": 33960
+    },
+    {
+      "epoch": 1.6871957882189332,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0006650283103208503,
+      "loss": 0.5954,
+      "step": 33970
+    },
+    {
+      "epoch": 1.6876924605145525,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0006649885765372008,
+      "loss": 0.5736,
+      "step": 33980
+    },
+    {
+      "epoch": 1.6881891328101717,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0006649488427535512,
+      "loss": 0.5884,
+      "step": 33990
+    },
+    {
+      "epoch": 1.6886858051057914,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0006649091089699017,
+      "loss": 0.6015,
+      "step": 34000
+    },
+    {
+      "epoch": 1.6891824774014106,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006648693751862522,
+      "loss": 0.5854,
+      "step": 34010
+    },
+    {
+      "epoch": 1.6896791496970298,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0006648296414026026,
+      "loss": 0.6298,
+      "step": 34020
+    },
+    {
+      "epoch": 1.6901758219926493,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.000664789907618953,
+      "loss": 0.576,
+      "step": 34030
+    },
+    {
+      "epoch": 1.6906724942882687,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006647501738353036,
+      "loss": 0.5603,
+      "step": 34040
+    },
+    {
+      "epoch": 1.691169166583888,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.000664710440051654,
+      "loss": 0.601,
+      "step": 34050
+    },
+    {
+      "epoch": 1.6916658388795072,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0006646707062680044,
+      "loss": 0.6415,
+      "step": 34060
+    },
+    {
+      "epoch": 1.6921625111751266,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006646309724843549,
+      "loss": 0.6118,
+      "step": 34070
+    },
+    {
+      "epoch": 1.692659183470746,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006645912387007053,
+      "loss": 0.5767,
+      "step": 34080
+    },
+    {
+      "epoch": 1.6931558557663653,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006645515049170557,
+      "loss": 0.5706,
+      "step": 34090
+    },
+    {
+      "epoch": 1.6936525280619847,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0006645117711334062,
+      "loss": 0.6036,
+      "step": 34100
+    },
+    {
+      "epoch": 1.6941492003576042,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006644720373497567,
+      "loss": 0.6352,
+      "step": 34110
+    },
+    {
+      "epoch": 1.6946458726532234,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006644323035661071,
+      "loss": 0.5675,
+      "step": 34120
+    },
+    {
+      "epoch": 1.6951425449488426,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0006643925697824575,
+      "loss": 0.6123,
+      "step": 34130
+    },
+    {
+      "epoch": 1.695639217244462,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.000664352835998808,
+      "loss": 0.5847,
+      "step": 34140
+    },
+    {
+      "epoch": 1.6961358895400815,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006643131022151585,
+      "loss": 0.589,
+      "step": 34150
+    },
+    {
+      "epoch": 1.6966325618357008,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0006642733684315089,
+      "loss": 0.5961,
+      "step": 34160
+    },
+    {
+      "epoch": 1.69712923413132,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0006642336346478594,
+      "loss": 0.6074,
+      "step": 34170
+    },
+    {
+      "epoch": 1.6976259064269397,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0006641939008642098,
+      "loss": 0.5817,
+      "step": 34180
+    },
+    {
+      "epoch": 1.698122578722559,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006641541670805602,
+      "loss": 0.5915,
+      "step": 34190
+    },
+    {
+      "epoch": 1.6986192510181781,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0006641144332969108,
+      "loss": 0.5905,
+      "step": 34200
+    },
+    {
+      "epoch": 1.6991159233137976,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0006640746995132612,
+      "loss": 0.5991,
+      "step": 34210
+    },
+    {
+      "epoch": 1.699612595609417,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0006640349657296116,
+      "loss": 0.5691,
+      "step": 34220
+    },
+    {
+      "epoch": 1.7001092679050362,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006639952319459621,
+      "loss": 0.5869,
+      "step": 34230
+    },
+    {
+      "epoch": 1.7006059402006555,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006639554981623125,
+      "loss": 0.6391,
+      "step": 34240
+    },
+    {
+      "epoch": 1.701102612496275,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.000663915764378663,
+      "loss": 0.583,
+      "step": 34250
+    },
+    {
+      "epoch": 1.7015992847918944,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006638760305950135,
+      "loss": 0.6054,
+      "step": 34260
+    },
+    {
+      "epoch": 1.7020959570875136,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006638362968113639,
+      "loss": 0.5867,
+      "step": 34270
+    },
+    {
+      "epoch": 1.702592629383133,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006637965630277143,
+      "loss": 0.6036,
+      "step": 34280
+    },
+    {
+      "epoch": 1.7030893016787525,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0006637568292440648,
+      "loss": 0.6233,
+      "step": 34290
+    },
+    {
+      "epoch": 1.7035859739743717,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0006637170954604153,
+      "loss": 0.5898,
+      "step": 34300
+    },
+    {
+      "epoch": 1.704082646269991,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006636773616767657,
+      "loss": 0.6078,
+      "step": 34310
+    },
+    {
+      "epoch": 1.7045793185656104,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006636376278931161,
+      "loss": 0.6156,
+      "step": 34320
+    },
+    {
+      "epoch": 1.7050759908612299,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006635978941094666,
+      "loss": 0.5982,
+      "step": 34330
+    },
+    {
+      "epoch": 1.705572663156849,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006635581603258171,
+      "loss": 0.5923,
+      "step": 34340
+    },
+    {
+      "epoch": 1.7060693354524683,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0006635184265421675,
+      "loss": 0.5583,
+      "step": 34350
+    },
+    {
+      "epoch": 1.706566007748088,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.000663478692758518,
+      "loss": 0.5799,
+      "step": 34360
+    },
+    {
+      "epoch": 1.7070626800437072,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0006634389589748684,
+      "loss": 0.5927,
+      "step": 34370
+    },
+    {
+      "epoch": 1.7075593523393264,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0006633992251912188,
+      "loss": 0.5996,
+      "step": 34380
+    },
+    {
+      "epoch": 1.7080560246349459,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006633594914075693,
+      "loss": 0.6004,
+      "step": 34390
+    },
+    {
+      "epoch": 1.7085526969305653,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006633197576239198,
+      "loss": 0.611,
+      "step": 34400
+    },
+    {
+      "epoch": 1.7090493692261846,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006632800238402702,
+      "loss": 0.5754,
+      "step": 34410
+    },
+    {
+      "epoch": 1.7095460415218038,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006632402900566207,
+      "loss": 0.599,
+      "step": 34420
+    },
+    {
+      "epoch": 1.7100427138174232,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006632005562729711,
+      "loss": 0.5795,
+      "step": 34430
+    },
+    {
+      "epoch": 1.7105393861130427,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0006631608224893215,
+      "loss": 0.5899,
+      "step": 34440
+    },
+    {
+      "epoch": 1.711036058408662,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0006631210887056721,
+      "loss": 0.5833,
+      "step": 34450
+    },
+    {
+      "epoch": 1.7115327307042814,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006630813549220225,
+      "loss": 0.5888,
+      "step": 34460
+    },
+    {
+      "epoch": 1.7120294029999008,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0006630416211383729,
+      "loss": 0.5978,
+      "step": 34470
+    },
+    {
+      "epoch": 1.71252607529552,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006630018873547234,
+      "loss": 0.5993,
+      "step": 34480
+    },
+    {
+      "epoch": 1.7130227475911393,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0006629621535710738,
+      "loss": 0.5815,
+      "step": 34490
+    },
+    {
+      "epoch": 1.7135194198867587,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0006629224197874244,
+      "loss": 0.5956,
+      "step": 34500
+    },
+    {
+      "epoch": 1.7140160921823782,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006628826860037747,
+      "loss": 0.5933,
+      "step": 34510
+    },
+    {
+      "epoch": 1.7145127644779974,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006628429522201252,
+      "loss": 0.5888,
+      "step": 34520
+    },
+    {
+      "epoch": 1.7150094367736166,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0006628032184364757,
+      "loss": 0.5999,
+      "step": 34530
+    },
+    {
+      "epoch": 1.715506109069236,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.000662763484652826,
+      "loss": 0.6141,
+      "step": 34540
+    },
+    {
+      "epoch": 1.7160027813648555,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006627237508691766,
+      "loss": 0.5664,
+      "step": 34550
+    },
+    {
+      "epoch": 1.7164994536604747,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006626840170855271,
+      "loss": 0.5732,
+      "step": 34560
+    },
+    {
+      "epoch": 1.7169961259560942,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0006626442833018774,
+      "loss": 0.5897,
+      "step": 34570
+    },
+    {
+      "epoch": 1.7174927982517136,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006626045495182279,
+      "loss": 0.6314,
+      "step": 34580
+    },
+    {
+      "epoch": 1.7179894705473329,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006625648157345783,
+      "loss": 0.5981,
+      "step": 34590
+    },
+    {
+      "epoch": 1.718486142842952,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006625250819509288,
+      "loss": 0.6238,
+      "step": 34600
+    },
+    {
+      "epoch": 1.7189828151385715,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006624853481672793,
+      "loss": 0.5775,
+      "step": 34610
+    },
+    {
+      "epoch": 1.719479487434191,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006624456143836297,
+      "loss": 0.5957,
+      "step": 34620
+    },
+    {
+      "epoch": 1.7199761597298102,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0006624058805999801,
+      "loss": 0.6131,
+      "step": 34630
+    },
+    {
+      "epoch": 1.7204728320254297,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0006623661468163306,
+      "loss": 0.6022,
+      "step": 34640
+    },
+    {
+      "epoch": 1.7209695043210491,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006623264130326811,
+      "loss": 0.601,
+      "step": 34650
+    },
+    {
+      "epoch": 1.7214661766166683,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006622866792490316,
+      "loss": 0.594,
+      "step": 34660
+    },
+    {
+      "epoch": 1.7219628489122876,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.000662246945465382,
+      "loss": 0.5892,
+      "step": 34670
+    },
+    {
+      "epoch": 1.722459521207907,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006622072116817324,
+      "loss": 0.5817,
+      "step": 34680
+    },
+    {
+      "epoch": 1.7229561935035265,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0006621674778980829,
+      "loss": 0.6195,
+      "step": 34690
+    },
+    {
+      "epoch": 1.7234528657991457,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006621277441144333,
+      "loss": 0.5867,
+      "step": 34700
+    },
+    {
+      "epoch": 1.723949538094765,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006620880103307838,
+      "loss": 0.6086,
+      "step": 34710
+    },
+    {
+      "epoch": 1.7244462103903844,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006620482765471343,
+      "loss": 0.5925,
+      "step": 34720
+    },
+    {
+      "epoch": 1.7249428826860038,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0006620085427634846,
+      "loss": 0.6208,
+      "step": 34730
+    },
+    {
+      "epoch": 1.725439554981623,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0006619688089798351,
+      "loss": 0.6107,
+      "step": 34740
+    },
+    {
+      "epoch": 1.7259362272772425,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006619290751961857,
+      "loss": 0.5579,
+      "step": 34750
+    },
+    {
+      "epoch": 1.726432899572862,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.000661889341412536,
+      "loss": 0.5842,
+      "step": 34760
+    },
+    {
+      "epoch": 1.7269295718684812,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006618496076288865,
+      "loss": 0.5636,
+      "step": 34770
+    },
+    {
+      "epoch": 1.7274262441641004,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006618098738452369,
+      "loss": 0.6168,
+      "step": 34780
+    },
+    {
+      "epoch": 1.7279229164597198,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006617701400615874,
+      "loss": 0.5637,
+      "step": 34790
+    },
+    {
+      "epoch": 1.7284195887553393,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0006617304062779379,
+      "loss": 0.5769,
+      "step": 34800
+    },
+    {
+      "epoch": 1.7289162610509585,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0006616906724942883,
+      "loss": 0.6049,
+      "step": 34810
+    },
+    {
+      "epoch": 1.729412933346578,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006616509387106388,
+      "loss": 0.5878,
+      "step": 34820
+    },
+    {
+      "epoch": 1.7299096056421974,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006616112049269892,
+      "loss": 0.5919,
+      "step": 34830
+    },
+    {
+      "epoch": 1.7304062779378167,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0006615714711433396,
+      "loss": 0.5782,
+      "step": 34840
+    },
+    {
+      "epoch": 1.7309029502334359,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0006615317373596902,
+      "loss": 0.634,
+      "step": 34850
+    },
+    {
+      "epoch": 1.7313996225290553,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006614920035760406,
+      "loss": 0.5835,
+      "step": 34860
+    },
+    {
+      "epoch": 1.7318962948246748,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.000661452269792391,
+      "loss": 0.6219,
+      "step": 34870
+    },
+    {
+      "epoch": 1.732392967120294,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006614125360087415,
+      "loss": 0.5947,
+      "step": 34880
+    },
+    {
+      "epoch": 1.7328896394159132,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006613728022250919,
+      "loss": 0.59,
+      "step": 34890
+    },
+    {
+      "epoch": 1.7333863117115327,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0006613330684414424,
+      "loss": 0.6176,
+      "step": 34900
+    },
+    {
+      "epoch": 1.7338829840071521,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0006612933346577929,
+      "loss": 0.619,
+      "step": 34910
+    },
+    {
+      "epoch": 1.7343796563027714,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0006612536008741432,
+      "loss": 0.5682,
+      "step": 34920
+    },
+    {
+      "epoch": 1.7348763285983908,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006612138670904937,
+      "loss": 0.5862,
+      "step": 34930
+    },
+    {
+      "epoch": 1.7353730008940103,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006611741333068442,
+      "loss": 0.5935,
+      "step": 34940
+    },
+    {
+      "epoch": 1.7358696731896295,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0006611343995231947,
+      "loss": 0.5896,
+      "step": 34950
+    },
+    {
+      "epoch": 1.7363663454852487,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0006610946657395451,
+      "loss": 0.5667,
+      "step": 34960
+    },
+    {
+      "epoch": 1.7368630177808682,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0006610549319558955,
+      "loss": 0.6185,
+      "step": 34970
+    },
+    {
+      "epoch": 1.7373596900764876,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.000661015198172246,
+      "loss": 0.5739,
+      "step": 34980
+    },
+    {
+      "epoch": 1.7378563623721068,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006609754643885964,
+      "loss": 0.5968,
+      "step": 34990
+    },
+    {
+      "epoch": 1.7383530346677263,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006609357306049468,
+      "loss": 0.6268,
+      "step": 35000
+    },
+    {
+      "epoch": 1.7388497069633457,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0006608959968212974,
+      "loss": 0.6147,
+      "step": 35010
+    },
+    {
+      "epoch": 1.739346379258965,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006608562630376478,
+      "loss": 0.577,
+      "step": 35020
+    },
+    {
+      "epoch": 1.7398430515545842,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006608165292539982,
+      "loss": 0.601,
+      "step": 35030
+    },
+    {
+      "epoch": 1.7403397238502036,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0006607767954703487,
+      "loss": 0.6041,
+      "step": 35040
+    },
+    {
+      "epoch": 1.740836396145823,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0006607370616866992,
+      "loss": 0.581,
+      "step": 35050
+    },
+    {
+      "epoch": 1.7413330684414423,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006606973279030496,
+      "loss": 0.5901,
+      "step": 35060
+    },
+    {
+      "epoch": 1.7418297407370615,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0006606575941194001,
+      "loss": 0.6036,
+      "step": 35070
+    },
+    {
+      "epoch": 1.742326413032681,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006606178603357505,
+      "loss": 0.6168,
+      "step": 35080
+    },
+    {
+      "epoch": 1.7428230853283004,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0006605781265521009,
+      "loss": 0.6085,
+      "step": 35090
+    },
+    {
+      "epoch": 1.7433197576239197,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006605383927684515,
+      "loss": 0.5746,
+      "step": 35100
+    },
+    {
+      "epoch": 1.7438164299195391,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0006604986589848019,
+      "loss": 0.581,
+      "step": 35110
+    },
+    {
+      "epoch": 1.7443131022151586,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0006604589252011523,
+      "loss": 0.5806,
+      "step": 35120
+    },
+    {
+      "epoch": 1.7448097745107778,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0006604191914175028,
+      "loss": 0.5574,
+      "step": 35130
+    },
+    {
+      "epoch": 1.745306446806397,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0006603794576338532,
+      "loss": 0.6143,
+      "step": 35140
+    },
+    {
+      "epoch": 1.7458031191020165,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0006603397238502036,
+      "loss": 0.5985,
+      "step": 35150
+    },
+    {
+      "epoch": 1.746299791397636,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0006602999900665542,
+      "loss": 0.6048,
+      "step": 35160
+    },
+    {
+      "epoch": 1.7467964636932551,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006602602562829046,
+      "loss": 0.5965,
+      "step": 35170
+    },
+    {
+      "epoch": 1.7472931359888746,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.000660220522499255,
+      "loss": 0.5958,
+      "step": 35180
+    },
+    {
+      "epoch": 1.747789808284494,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006601807887156054,
+      "loss": 0.607,
+      "step": 35190
+    },
+    {
+      "epoch": 1.7482864805801133,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.000660141054931956,
+      "loss": 0.5923,
+      "step": 35200
+    },
+    {
+      "epoch": 1.7487831528757325,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0006601013211483064,
+      "loss": 0.6049,
+      "step": 35210
+    },
+    {
+      "epoch": 1.749279825171352,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0006600615873646568,
+      "loss": 0.6206,
+      "step": 35220
+    },
+    {
+      "epoch": 1.7497764974669714,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0006600218535810073,
+      "loss": 0.6048,
+      "step": 35230
+    },
+    {
+      "epoch": 1.7502731697625906,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006599821197973577,
+      "loss": 0.6114,
+      "step": 35240
+    },
+    {
+      "epoch": 1.7507698420582098,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0006599423860137081,
+      "loss": 0.5983,
+      "step": 35250
+    },
+    {
+      "epoch": 1.7512665143538293,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006599026522300587,
+      "loss": 0.6022,
+      "step": 35260
+    },
+    {
+      "epoch": 1.7517631866494487,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006598629184464091,
+      "loss": 0.616,
+      "step": 35270
+    },
+    {
+      "epoch": 1.752259858945068,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006598231846627595,
+      "loss": 0.5992,
+      "step": 35280
+    },
+    {
+      "epoch": 1.7527565312406874,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00065978345087911,
+      "loss": 0.5907,
+      "step": 35290
+    },
+    {
+      "epoch": 1.7532532035363069,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0006597437170954604,
+      "loss": 0.5911,
+      "step": 35300
+    },
+    {
+      "epoch": 1.753749875831926,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006597039833118109,
+      "loss": 0.5878,
+      "step": 35310
+    },
+    {
+      "epoch": 1.7542465481275453,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0006596642495281614,
+      "loss": 0.58,
+      "step": 35320
+    },
+    {
+      "epoch": 1.7547432204231648,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0006596245157445118,
+      "loss": 0.6123,
+      "step": 35330
+    },
+    {
+      "epoch": 1.7552398927187842,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0006595847819608622,
+      "loss": 0.5974,
+      "step": 35340
+    },
+    {
+      "epoch": 1.7557365650144034,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0006595450481772127,
+      "loss": 0.5911,
+      "step": 35350
+    },
+    {
+      "epoch": 1.756233237310023,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006595053143935632,
+      "loss": 0.5947,
+      "step": 35360
+    },
+    {
+      "epoch": 1.7567299096056423,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006594655806099136,
+      "loss": 0.5844,
+      "step": 35370
+    },
+    {
+      "epoch": 1.7572265819012616,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.000659425846826264,
+      "loss": 0.5914,
+      "step": 35380
+    },
+    {
+      "epoch": 1.7577232541968808,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006593861130426145,
+      "loss": 0.5856,
+      "step": 35390
+    },
+    {
+      "epoch": 1.7582199264925003,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.000659346379258965,
+      "loss": 0.6051,
+      "step": 35400
+    },
+    {
+      "epoch": 1.7587165987881197,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006593066454753154,
+      "loss": 0.6096,
+      "step": 35410
+    },
+    {
+      "epoch": 1.759213271083739,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0006592669116916659,
+      "loss": 0.5855,
+      "step": 35420
+    },
+    {
+      "epoch": 1.7597099433793582,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0006592271779080163,
+      "loss": 0.6032,
+      "step": 35430
+    },
+    {
+      "epoch": 1.7602066156749776,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006591874441243667,
+      "loss": 0.5703,
+      "step": 35440
+    },
+    {
+      "epoch": 1.760703287970597,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006591477103407172,
+      "loss": 0.596,
+      "step": 35450
+    },
+    {
+      "epoch": 1.7611999602662163,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006591079765570677,
+      "loss": 0.5854,
+      "step": 35460
+    },
+    {
+      "epoch": 1.7616966325618357,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0006590682427734181,
+      "loss": 0.5599,
+      "step": 35470
+    },
+    {
+      "epoch": 1.7621933048574552,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006590285089897686,
+      "loss": 0.6016,
+      "step": 35480
+    },
+    {
+      "epoch": 1.7626899771530744,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.000658988775206119,
+      "loss": 0.597,
+      "step": 35490
+    },
+    {
+      "epoch": 1.7631866494486936,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006589490414224694,
+      "loss": 0.6352,
+      "step": 35500
+    },
+    {
+      "epoch": 1.763683321744313,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.00065890930763882,
+      "loss": 0.5891,
+      "step": 35510
+    },
+    {
+      "epoch": 1.7641799940399325,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0006588695738551704,
+      "loss": 0.5808,
+      "step": 35520
+    },
+    {
+      "epoch": 1.7646766663355518,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006588298400715208,
+      "loss": 0.585,
+      "step": 35530
+    },
+    {
+      "epoch": 1.7651733386311712,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006587901062878713,
+      "loss": 0.5931,
+      "step": 35540
+    },
+    {
+      "epoch": 1.7656700109267907,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0006587503725042217,
+      "loss": 0.5712,
+      "step": 35550
+    },
+    {
+      "epoch": 1.7661666832224099,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0006587106387205723,
+      "loss": 0.5875,
+      "step": 35560
+    },
+    {
+      "epoch": 1.766663355518029,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0006586709049369226,
+      "loss": 0.5762,
+      "step": 35570
+    },
+    {
+      "epoch": 1.7671600278136486,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0006586311711532731,
+      "loss": 0.586,
+      "step": 35580
+    },
+    {
+      "epoch": 1.767656700109268,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0006585914373696236,
+      "loss": 0.6036,
+      "step": 35590
+    },
+    {
+      "epoch": 1.7681533724048872,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006585517035859739,
+      "loss": 0.592,
+      "step": 35600
+    },
+    {
+      "epoch": 1.7686500447005065,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006585119698023245,
+      "loss": 0.5644,
+      "step": 35610
+    },
+    {
+      "epoch": 1.769146716996126,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.000658472236018675,
+      "loss": 0.579,
+      "step": 35620
+    },
+    {
+      "epoch": 1.7696433892917454,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006584325022350253,
+      "loss": 0.5832,
+      "step": 35630
+    },
+    {
+      "epoch": 1.7701400615873646,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006583927684513758,
+      "loss": 0.5951,
+      "step": 35640
+    },
+    {
+      "epoch": 1.770636733882984,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0006583530346677262,
+      "loss": 0.5847,
+      "step": 35650
+    },
+    {
+      "epoch": 1.7711334061786035,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0006583133008840767,
+      "loss": 0.5947,
+      "step": 35660
+    },
+    {
+      "epoch": 1.7716300784742227,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0006582735671004272,
+      "loss": 0.5862,
+      "step": 35670
+    },
+    {
+      "epoch": 1.772126750769842,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0006582338333167776,
+      "loss": 0.6173,
+      "step": 35680
+    },
+    {
+      "epoch": 1.7726234230654614,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0006581940995331281,
+      "loss": 0.5851,
+      "step": 35690
+    },
+    {
+      "epoch": 1.7731200953610808,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006581543657494785,
+      "loss": 0.5829,
+      "step": 35700
+    },
+    {
+      "epoch": 1.7736167676567,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.000658114631965829,
+      "loss": 0.5852,
+      "step": 35710
+    },
+    {
+      "epoch": 1.7741134399523193,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0006580748981821795,
+      "loss": 0.5783,
+      "step": 35720
+    },
+    {
+      "epoch": 1.774610112247939,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006580351643985299,
+      "loss": 0.5801,
+      "step": 35730
+    },
+    {
+      "epoch": 1.7751067845435582,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0006579954306148803,
+      "loss": 0.5973,
+      "step": 35740
+    },
+    {
+      "epoch": 1.7756034568391774,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0006579556968312308,
+      "loss": 0.611,
+      "step": 35750
+    },
+    {
+      "epoch": 1.7761001291347969,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0006579159630475812,
+      "loss": 0.5763,
+      "step": 35760
+    },
+    {
+      "epoch": 1.7765968014304163,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006578762292639317,
+      "loss": 0.5711,
+      "step": 35770
+    },
+    {
+      "epoch": 1.7770934737260355,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006578364954802822,
+      "loss": 0.6111,
+      "step": 35780
+    },
+    {
+      "epoch": 1.7775901460216548,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006577967616966325,
+      "loss": 0.5693,
+      "step": 35790
+    },
+    {
+      "epoch": 1.7780868183172742,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.000657757027912983,
+      "loss": 0.5779,
+      "step": 35800
+    },
+    {
+      "epoch": 1.7785834906128937,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0006577172941293336,
+      "loss": 0.5906,
+      "step": 35810
+    },
+    {
+      "epoch": 1.779080162908513,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0006576775603456839,
+      "loss": 0.5878,
+      "step": 35820
+    },
+    {
+      "epoch": 1.7795768352041323,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006576378265620344,
+      "loss": 0.5832,
+      "step": 35830
+    },
+    {
+      "epoch": 1.7800735074997518,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006575980927783848,
+      "loss": 0.5882,
+      "step": 35840
+    },
+    {
+      "epoch": 1.780570179795371,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006575583589947353,
+      "loss": 0.6045,
+      "step": 35850
+    },
+    {
+      "epoch": 1.7810668520909902,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006575186252110858,
+      "loss": 0.5789,
+      "step": 35860
+    },
+    {
+      "epoch": 1.7815635243866097,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0006574788914274362,
+      "loss": 0.5812,
+      "step": 35870
+    },
+    {
+      "epoch": 1.7820601966822291,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0006574391576437867,
+      "loss": 0.5801,
+      "step": 35880
+    },
+    {
+      "epoch": 1.7825568689778484,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006573994238601371,
+      "loss": 0.5915,
+      "step": 35890
+    },
+    {
+      "epoch": 1.7830535412734676,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0006573596900764875,
+      "loss": 0.5687,
+      "step": 35900
+    },
+    {
+      "epoch": 1.7835502135690873,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0006573199562928381,
+      "loss": 0.5588,
+      "step": 35910
+    },
+    {
+      "epoch": 1.7840468858647065,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006572802225091885,
+      "loss": 0.579,
+      "step": 35920
+    },
+    {
+      "epoch": 1.7845435581603257,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0006572404887255389,
+      "loss": 0.6217,
+      "step": 35930
+    },
+    {
+      "epoch": 1.7850402304559452,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0006572007549418894,
+      "loss": 0.5919,
+      "step": 35940
+    },
+    {
+      "epoch": 1.7855369027515646,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006571610211582398,
+      "loss": 0.5992,
+      "step": 35950
+    },
+    {
+      "epoch": 1.7860335750471839,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006571212873745903,
+      "loss": 0.5961,
+      "step": 35960
+    },
+    {
+      "epoch": 1.786530247342803,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0006570815535909408,
+      "loss": 0.6127,
+      "step": 35970
+    },
+    {
+      "epoch": 1.7870269196384225,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0006570418198072911,
+      "loss": 0.5762,
+      "step": 35980
+    },
+    {
+      "epoch": 1.787523591934042,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006570020860236416,
+      "loss": 0.5988,
+      "step": 35990
+    },
+    {
+      "epoch": 1.7880202642296612,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0006569623522399921,
+      "loss": 0.581,
+      "step": 36000
+    },
+    {
+      "epoch": 1.7885169365252807,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006569226184563426,
+      "loss": 0.5811,
+      "step": 36010
+    },
+    {
+      "epoch": 1.7890136088209,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000656882884672693,
+      "loss": 0.5708,
+      "step": 36020
+    },
+    {
+      "epoch": 1.7895102811165193,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006568431508890434,
+      "loss": 0.5819,
+      "step": 36030
+    },
+    {
+      "epoch": 1.7900069534121386,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006568034171053939,
+      "loss": 0.5852,
+      "step": 36040
+    },
+    {
+      "epoch": 1.790503625707758,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0006567636833217443,
+      "loss": 0.6267,
+      "step": 36050
+    },
+    {
+      "epoch": 1.7910002980033775,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006567239495380948,
+      "loss": 0.5895,
+      "step": 36060
+    },
+    {
+      "epoch": 1.7914969702989967,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006566842157544453,
+      "loss": 0.599,
+      "step": 36070
+    },
+    {
+      "epoch": 1.791993642594616,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006566444819707957,
+      "loss": 0.5701,
+      "step": 36080
+    },
+    {
+      "epoch": 1.7924903148902356,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0006566047481871461,
+      "loss": 0.5955,
+      "step": 36090
+    },
+    {
+      "epoch": 1.7929869871858548,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006565650144034966,
+      "loss": 0.5917,
+      "step": 36100
+    },
+    {
+      "epoch": 1.793483659481474,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006565252806198471,
+      "loss": 0.5633,
+      "step": 36110
+    },
+    {
+      "epoch": 1.7939803317770935,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006564855468361975,
+      "loss": 0.5555,
+      "step": 36120
+    },
+    {
+      "epoch": 1.794477004072713,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.000656445813052548,
+      "loss": 0.6074,
+      "step": 36130
+    },
+    {
+      "epoch": 1.7949736763683322,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006564060792688984,
+      "loss": 0.6201,
+      "step": 36140
+    },
+    {
+      "epoch": 1.7954703486639514,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0006563663454852488,
+      "loss": 0.586,
+      "step": 36150
+    },
+    {
+      "epoch": 1.7959670209595708,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006563266117015994,
+      "loss": 0.5746,
+      "step": 36160
+    },
+    {
+      "epoch": 1.7964636932551903,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006562868779179498,
+      "loss": 0.5907,
+      "step": 36170
+    },
+    {
+      "epoch": 1.7969603655508095,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006562471441343002,
+      "loss": 0.5704,
+      "step": 36180
+    },
+    {
+      "epoch": 1.797457037846429,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006562074103506507,
+      "loss": 0.6184,
+      "step": 36190
+    },
+    {
+      "epoch": 1.7979537101420484,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006561676765670011,
+      "loss": 0.6149,
+      "step": 36200
+    },
+    {
+      "epoch": 1.7984503824376676,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0006561279427833516,
+      "loss": 0.5642,
+      "step": 36210
+    },
+    {
+      "epoch": 1.7989470547332869,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0006560882089997021,
+      "loss": 0.5813,
+      "step": 36220
+    },
+    {
+      "epoch": 1.7994437270289063,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006560484752160525,
+      "loss": 0.612,
+      "step": 36230
+    },
+    {
+      "epoch": 1.7999403993245258,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006560087414324029,
+      "loss": 0.5886,
+      "step": 36240
+    },
+    {
+      "epoch": 1.800437071620145,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006559690076487533,
+      "loss": 0.6051,
+      "step": 36250
+    },
+    {
+      "epoch": 1.8009337439157642,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0006559292738651039,
+      "loss": 0.5852,
+      "step": 36260
+    },
+    {
+      "epoch": 1.801430416211384,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0006558895400814543,
+      "loss": 0.5926,
+      "step": 36270
+    },
+    {
+      "epoch": 1.8019270885070031,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0006558498062978047,
+      "loss": 0.604,
+      "step": 36280
+    },
+    {
+      "epoch": 1.8024237608026223,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0006558100725141552,
+      "loss": 0.609,
+      "step": 36290
+    },
+    {
+      "epoch": 1.8029204330982418,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0006557703387305056,
+      "loss": 0.5675,
+      "step": 36300
+    },
+    {
+      "epoch": 1.8034171053938612,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.000655730604946856,
+      "loss": 0.5771,
+      "step": 36310
+    },
+    {
+      "epoch": 1.8039137776894805,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0006556908711632066,
+      "loss": 0.5779,
+      "step": 36320
+    },
+    {
+      "epoch": 1.8044104499850997,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.000655651137379557,
+      "loss": 0.6223,
+      "step": 36330
+    },
+    {
+      "epoch": 1.8049071222807191,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006556114035959074,
+      "loss": 0.5769,
+      "step": 36340
+    },
+    {
+      "epoch": 1.8054037945763386,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0006555716698122579,
+      "loss": 0.5798,
+      "step": 36350
+    },
+    {
+      "epoch": 1.8059004668719578,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0006555319360286084,
+      "loss": 0.5818,
+      "step": 36360
+    },
+    {
+      "epoch": 1.8063971391675773,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0006554922022449588,
+      "loss": 0.6004,
+      "step": 36370
+    },
+    {
+      "epoch": 1.8068938114631967,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006554524684613093,
+      "loss": 0.597,
+      "step": 36380
+    },
+    {
+      "epoch": 1.807390483758816,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006554127346776597,
+      "loss": 0.5771,
+      "step": 36390
+    },
+    {
+      "epoch": 1.8078871560544352,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0006553730008940101,
+      "loss": 0.5779,
+      "step": 36400
+    },
+    {
+      "epoch": 1.8083838283500546,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006553332671103607,
+      "loss": 0.596,
+      "step": 36410
+    },
+    {
+      "epoch": 1.808880500645674,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006552935333267111,
+      "loss": 0.6053,
+      "step": 36420
+    },
+    {
+      "epoch": 1.8093771729412933,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0006552537995430615,
+      "loss": 0.5661,
+      "step": 36430
+    },
+    {
+      "epoch": 1.8098738452369125,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0006552140657594119,
+      "loss": 0.5906,
+      "step": 36440
+    },
+    {
+      "epoch": 1.8103705175325322,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0006551743319757624,
+      "loss": 0.5939,
+      "step": 36450
+    },
+    {
+      "epoch": 1.8108671898281514,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.000655134598192113,
+      "loss": 0.5859,
+      "step": 36460
+    },
+    {
+      "epoch": 1.8113638621237707,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006550948644084633,
+      "loss": 0.5821,
+      "step": 36470
+    },
+    {
+      "epoch": 1.81186053441939,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0006550551306248138,
+      "loss": 0.5904,
+      "step": 36480
+    },
+    {
+      "epoch": 1.8123572067150095,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0006550153968411643,
+      "loss": 0.5741,
+      "step": 36490
+    },
+    {
+      "epoch": 1.8128538790106288,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006549756630575146,
+      "loss": 0.6147,
+      "step": 36500
+    },
+    {
+      "epoch": 1.813350551306248,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0006549359292738652,
+      "loss": 0.5682,
+      "step": 36510
+    },
+    {
+      "epoch": 1.8138472236018675,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006548961954902156,
+      "loss": 0.6279,
+      "step": 36520
+    },
+    {
+      "epoch": 1.814343895897487,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.000654856461706566,
+      "loss": 0.5933,
+      "step": 36530
+    },
+    {
+      "epoch": 1.8148405681931061,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006548167279229165,
+      "loss": 0.5712,
+      "step": 36540
+    },
+    {
+      "epoch": 1.8153372404887256,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006547769941392669,
+      "loss": 0.5993,
+      "step": 36550
+    },
+    {
+      "epoch": 1.815833912784345,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006547372603556173,
+      "loss": 0.6267,
+      "step": 36560
+    },
+    {
+      "epoch": 1.8163305850799643,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006546975265719679,
+      "loss": 0.5799,
+      "step": 36570
+    },
+    {
+      "epoch": 1.8168272573755835,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006546577927883183,
+      "loss": 0.6024,
+      "step": 36580
+    },
+    {
+      "epoch": 1.817323929671203,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006546180590046688,
+      "loss": 0.6024,
+      "step": 36590
+    },
+    {
+      "epoch": 1.8178206019668224,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006545783252210192,
+      "loss": 0.5906,
+      "step": 36600
+    },
+    {
+      "epoch": 1.8183172742624416,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006545385914373696,
+      "loss": 0.5762,
+      "step": 36610
+    },
+    {
+      "epoch": 1.8188139465580608,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0006544988576537202,
+      "loss": 0.5811,
+      "step": 36620
+    },
+    {
+      "epoch": 1.8193106188536805,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006544591238700705,
+      "loss": 0.5927,
+      "step": 36630
+    },
+    {
+      "epoch": 1.8198072911492997,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.000654419390086421,
+      "loss": 0.5813,
+      "step": 36640
+    },
+    {
+      "epoch": 1.820303963444919,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006543796563027715,
+      "loss": 0.5863,
+      "step": 36650
+    },
+    {
+      "epoch": 1.8208006357405384,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006543399225191218,
+      "loss": 0.5647,
+      "step": 36660
+    },
+    {
+      "epoch": 1.8212973080361579,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0006543001887354724,
+      "loss": 0.5953,
+      "step": 36670
+    },
+    {
+      "epoch": 1.821793980331777,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0006542604549518229,
+      "loss": 0.5822,
+      "step": 36680
+    },
+    {
+      "epoch": 1.8222906526273963,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006542207211681732,
+      "loss": 0.593,
+      "step": 36690
+    },
+    {
+      "epoch": 1.8227873249230158,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0006541809873845237,
+      "loss": 0.5887,
+      "step": 36700
+    },
+    {
+      "epoch": 1.8232839972186352,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006541412536008741,
+      "loss": 0.6251,
+      "step": 36710
+    },
+    {
+      "epoch": 1.8237806695142544,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0006541015198172246,
+      "loss": 0.5714,
+      "step": 36720
+    },
+    {
+      "epoch": 1.8242773418098739,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0006540617860335751,
+      "loss": 0.5708,
+      "step": 36730
+    },
+    {
+      "epoch": 1.8247740141054933,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0006540220522499255,
+      "loss": 0.5896,
+      "step": 36740
+    },
+    {
+      "epoch": 1.8252706864011126,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.000653982318466276,
+      "loss": 0.578,
+      "step": 36750
+    },
+    {
+      "epoch": 1.8257673586967318,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0006539425846826264,
+      "loss": 0.6018,
+      "step": 36760
+    },
+    {
+      "epoch": 1.8262640309923512,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0006539028508989769,
+      "loss": 0.6012,
+      "step": 36770
+    },
+    {
+      "epoch": 1.8267607032879707,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0006538631171153274,
+      "loss": 0.6244,
+      "step": 36780
+    },
+    {
+      "epoch": 1.82725737558359,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006538233833316778,
+      "loss": 0.5579,
+      "step": 36790
+    },
+    {
+      "epoch": 1.8277540478792091,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006537836495480282,
+      "loss": 0.6142,
+      "step": 36800
+    },
+    {
+      "epoch": 1.8282507201748288,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0006537439157643787,
+      "loss": 0.5613,
+      "step": 36810
+    },
+    {
+      "epoch": 1.828747392470448,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0006537041819807292,
+      "loss": 0.6049,
+      "step": 36820
+    },
+    {
+      "epoch": 1.8292440647660673,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006536644481970796,
+      "loss": 0.5672,
+      "step": 36830
+    },
+    {
+      "epoch": 1.8297407370616867,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0006536247144134301,
+      "loss": 0.5831,
+      "step": 36840
+    },
+    {
+      "epoch": 1.8302374093573062,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0006535849806297804,
+      "loss": 0.5675,
+      "step": 36850
+    },
+    {
+      "epoch": 1.8307340816529254,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006535452468461309,
+      "loss": 0.6041,
+      "step": 36860
+    },
+    {
+      "epoch": 1.8312307539485446,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006535055130624815,
+      "loss": 0.5579,
+      "step": 36870
+    },
+    {
+      "epoch": 1.831727426244164,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0006534657792788319,
+      "loss": 0.5693,
+      "step": 36880
+    },
+    {
+      "epoch": 1.8322240985397835,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0006534260454951823,
+      "loss": 0.5783,
+      "step": 36890
+    },
+    {
+      "epoch": 1.8327207708354027,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006533863117115327,
+      "loss": 0.5867,
+      "step": 36900
+    },
+    {
+      "epoch": 1.8332174431310222,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0006533465779278832,
+      "loss": 0.5905,
+      "step": 36910
+    },
+    {
+      "epoch": 1.8337141154266416,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006533068441442337,
+      "loss": 0.6061,
+      "step": 36920
+    },
+    {
+      "epoch": 1.8342107877222609,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006532671103605841,
+      "loss": 0.6096,
+      "step": 36930
+    },
+    {
+      "epoch": 1.83470746001788,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006532273765769346,
+      "loss": 0.576,
+      "step": 36940
+    },
+    {
+      "epoch": 1.8352041323134995,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.000653187642793285,
+      "loss": 0.6109,
+      "step": 36950
+    },
+    {
+      "epoch": 1.835700804609119,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0006531479090096354,
+      "loss": 0.568,
+      "step": 36960
+    },
+    {
+      "epoch": 1.8361974769047382,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.000653108175225986,
+      "loss": 0.5775,
+      "step": 36970
+    },
+    {
+      "epoch": 1.8366941492003575,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0006530684414423364,
+      "loss": 0.5969,
+      "step": 36980
+    },
+    {
+      "epoch": 1.837190821495977,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006530287076586868,
+      "loss": 0.582,
+      "step": 36990
+    },
+    {
+      "epoch": 1.8376874937915963,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0006529889738750373,
+      "loss": 0.5714,
+      "step": 37000
+    },
+    {
+      "epoch": 1.8381841660872156,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006529492400913877,
+      "loss": 0.609,
+      "step": 37010
+    },
+    {
+      "epoch": 1.838680838382835,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006529095063077382,
+      "loss": 0.6116,
+      "step": 37020
+    },
+    {
+      "epoch": 1.8391775106784545,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0006528697725240887,
+      "loss": 0.6088,
+      "step": 37030
+    },
+    {
+      "epoch": 1.8396741829740737,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006528300387404391,
+      "loss": 0.6042,
+      "step": 37040
+    },
+    {
+      "epoch": 1.840170855269693,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0006527903049567895,
+      "loss": 0.6014,
+      "step": 37050
+    },
+    {
+      "epoch": 1.8406675275653124,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00065275057117314,
+      "loss": 0.5956,
+      "step": 37060
+    },
+    {
+      "epoch": 1.8411641998609318,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0006527108373894905,
+      "loss": 0.5895,
+      "step": 37070
+    },
+    {
+      "epoch": 1.841660872156551,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0006526711036058409,
+      "loss": 0.6098,
+      "step": 37080
+    },
+    {
+      "epoch": 1.8421575444521705,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006526313698221914,
+      "loss": 0.585,
+      "step": 37090
+    },
+    {
+      "epoch": 1.84265421674779,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0006525916360385418,
+      "loss": 0.6035,
+      "step": 37100
+    },
+    {
+      "epoch": 1.8431508890434092,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0006525519022548922,
+      "loss": 0.5677,
+      "step": 37110
+    },
+    {
+      "epoch": 1.8436475613390284,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006525121684712427,
+      "loss": 0.5505,
+      "step": 37120
+    },
+    {
+      "epoch": 1.8441442336346479,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0006524724346875932,
+      "loss": 0.5833,
+      "step": 37130
+    },
+    {
+      "epoch": 1.8446409059302673,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0006524327009039436,
+      "loss": 0.5779,
+      "step": 37140
+    },
+    {
+      "epoch": 1.8451375782258865,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.000652392967120294,
+      "loss": 0.5994,
+      "step": 37150
+    },
+    {
+      "epoch": 1.8456342505215058,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006523532333366445,
+      "loss": 0.5936,
+      "step": 37160
+    },
+    {
+      "epoch": 1.8461309228171252,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.000652313499552995,
+      "loss": 0.5973,
+      "step": 37170
+    },
+    {
+      "epoch": 1.8466275951127447,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0006522737657693454,
+      "loss": 0.5972,
+      "step": 37180
+    },
+    {
+      "epoch": 1.8471242674083639,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0006522340319856959,
+      "loss": 0.6032,
+      "step": 37190
+    },
+    {
+      "epoch": 1.8476209397039833,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006521942982020463,
+      "loss": 0.607,
+      "step": 37200
+    },
+    {
+      "epoch": 1.8481176119996028,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006521545644183967,
+      "loss": 0.6153,
+      "step": 37210
+    },
+    {
+      "epoch": 1.848614284295222,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0006521148306347473,
+      "loss": 0.5763,
+      "step": 37220
+    },
+    {
+      "epoch": 1.8491109565908412,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006520750968510977,
+      "loss": 0.5736,
+      "step": 37230
+    },
+    {
+      "epoch": 1.8496076288864607,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0006520353630674481,
+      "loss": 0.6079,
+      "step": 37240
+    },
+    {
+      "epoch": 1.8501043011820801,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0006519956292837986,
+      "loss": 0.5812,
+      "step": 37250
+    },
+    {
+      "epoch": 1.8506009734776994,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.000651955895500149,
+      "loss": 0.6019,
+      "step": 37260
+    },
+    {
+      "epoch": 1.8510976457733188,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006519161617164995,
+      "loss": 0.6007,
+      "step": 37270
+    },
+    {
+      "epoch": 1.8515943180689383,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00065187642793285,
+      "loss": 0.5725,
+      "step": 37280
+    },
+    {
+      "epoch": 1.8520909903645575,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006518366941492004,
+      "loss": 0.5952,
+      "step": 37290
+    },
+    {
+      "epoch": 1.8525876626601767,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006517969603655508,
+      "loss": 0.5955,
+      "step": 37300
+    },
+    {
+      "epoch": 1.8530843349557962,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0006517572265819012,
+      "loss": 0.5889,
+      "step": 37310
+    },
+    {
+      "epoch": 1.8535810072514156,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0006517174927982518,
+      "loss": 0.5997,
+      "step": 37320
+    },
+    {
+      "epoch": 1.8540776795470348,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0006516777590146023,
+      "loss": 0.5642,
+      "step": 37330
+    },
+    {
+      "epoch": 1.854574351842654,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0006516380252309526,
+      "loss": 0.595,
+      "step": 37340
+    },
+    {
+      "epoch": 1.8550710241382735,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0006515982914473031,
+      "loss": 0.5922,
+      "step": 37350
+    },
+    {
+      "epoch": 1.855567696433893,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006515585576636536,
+      "loss": 0.5832,
+      "step": 37360
+    },
+    {
+      "epoch": 1.8560643687295122,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.000651518823880004,
+      "loss": 0.5978,
+      "step": 37370
+    },
+    {
+      "epoch": 1.8565610410251316,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0006514790900963545,
+      "loss": 0.573,
+      "step": 37380
+    },
+    {
+      "epoch": 1.857057713320751,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0006514393563127049,
+      "loss": 0.5722,
+      "step": 37390
+    },
+    {
+      "epoch": 1.8575543856163703,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0006513996225290553,
+      "loss": 0.6139,
+      "step": 37400
+    },
+    {
+      "epoch": 1.8580510579119895,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006513598887454058,
+      "loss": 0.5884,
+      "step": 37410
+    },
+    {
+      "epoch": 1.858547730207609,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0006513201549617563,
+      "loss": 0.6201,
+      "step": 37420
+    },
+    {
+      "epoch": 1.8590444025032284,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006512804211781067,
+      "loss": 0.5874,
+      "step": 37430
+    },
+    {
+      "epoch": 1.8595410747988477,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006512406873944572,
+      "loss": 0.6012,
+      "step": 37440
+    },
+    {
+      "epoch": 1.8600377470944671,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0006512009536108076,
+      "loss": 0.5957,
+      "step": 37450
+    },
+    {
+      "epoch": 1.8605344193900866,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.000651161219827158,
+      "loss": 0.5916,
+      "step": 37460
+    },
+    {
+      "epoch": 1.8610310916857058,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0006511214860435086,
+      "loss": 0.586,
+      "step": 37470
+    },
+    {
+      "epoch": 1.861527763981325,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.000651081752259859,
+      "loss": 0.569,
+      "step": 37480
+    },
+    {
+      "epoch": 1.8620244362769445,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0006510420184762095,
+      "loss": 0.5574,
+      "step": 37490
+    },
+    {
+      "epoch": 1.862521108572564,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006510022846925598,
+      "loss": 0.5899,
+      "step": 37500
+    },
+    {
+      "epoch": 1.8630177808681831,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0006509625509089103,
+      "loss": 0.5996,
+      "step": 37510
+    },
+    {
+      "epoch": 1.8635144531638024,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0006509228171252609,
+      "loss": 0.6009,
+      "step": 37520
+    },
+    {
+      "epoch": 1.8640111254594218,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0006508830833416112,
+      "loss": 0.5804,
+      "step": 37530
+    },
+    {
+      "epoch": 1.8645077977550413,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0006508433495579617,
+      "loss": 0.6056,
+      "step": 37540
+    },
+    {
+      "epoch": 1.8650044700506605,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0006508036157743122,
+      "loss": 0.5867,
+      "step": 37550
+    },
+    {
+      "epoch": 1.86550114234628,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0006507638819906625,
+      "loss": 0.6001,
+      "step": 37560
+    },
+    {
+      "epoch": 1.8659978146418994,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.000650724148207013,
+      "loss": 0.5841,
+      "step": 37570
+    },
+    {
+      "epoch": 1.8664944869375186,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006506844144233635,
+      "loss": 0.5766,
+      "step": 37580
+    },
+    {
+      "epoch": 1.8669911592331379,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0006506446806397139,
+      "loss": 0.5876,
+      "step": 37590
+    },
+    {
+      "epoch": 1.8674878315287573,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0006506049468560644,
+      "loss": 0.5907,
+      "step": 37600
+    },
+    {
+      "epoch": 1.8679845038243768,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006505652130724148,
+      "loss": 0.5754,
+      "step": 37610
+    },
+    {
+      "epoch": 1.868481176119996,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0006505254792887652,
+      "loss": 0.5736,
+      "step": 37620
+    },
+    {
+      "epoch": 1.8689778484156154,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006504857455051158,
+      "loss": 0.6151,
+      "step": 37630
+    },
+    {
+      "epoch": 1.8694745207112349,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0006504460117214662,
+      "loss": 0.5984,
+      "step": 37640
+    },
+    {
+      "epoch": 1.869971193006854,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006504062779378167,
+      "loss": 0.6013,
+      "step": 37650
+    },
+    {
+      "epoch": 1.8704678653024733,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0006503665441541671,
+      "loss": 0.6051,
+      "step": 37660
+    },
+    {
+      "epoch": 1.8709645375980928,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0006503268103705176,
+      "loss": 0.5964,
+      "step": 37670
+    },
+    {
+      "epoch": 1.8714612098937122,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0006502870765868681,
+      "loss": 0.5812,
+      "step": 37680
+    },
+    {
+      "epoch": 1.8719578821893315,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006502473428032185,
+      "loss": 0.5946,
+      "step": 37690
+    },
+    {
+      "epoch": 1.8724545544849507,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006502076090195689,
+      "loss": 0.5612,
+      "step": 37700
+    },
+    {
+      "epoch": 1.8729512267805701,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006501678752359194,
+      "loss": 0.585,
+      "step": 37710
+    },
+    {
+      "epoch": 1.8734478990761896,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0006501281414522697,
+      "loss": 0.5948,
+      "step": 37720
+    },
+    {
+      "epoch": 1.8739445713718088,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0006500884076686203,
+      "loss": 0.5773,
+      "step": 37730
+    },
+    {
+      "epoch": 1.8744412436674283,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006500486738849708,
+      "loss": 0.5941,
+      "step": 37740
+    },
+    {
+      "epoch": 1.8749379159630477,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006500089401013211,
+      "loss": 0.5943,
+      "step": 37750
+    },
+    {
+      "epoch": 1.875434588258667,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006499692063176716,
+      "loss": 0.5959,
+      "step": 37760
+    },
+    {
+      "epoch": 1.8759312605542862,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.000649929472534022,
+      "loss": 0.5773,
+      "step": 37770
+    },
+    {
+      "epoch": 1.8764279328499056,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006498897387503726,
+      "loss": 0.5837,
+      "step": 37780
+    },
+    {
+      "epoch": 1.876924605145525,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.000649850004966723,
+      "loss": 0.5888,
+      "step": 37790
+    },
+    {
+      "epoch": 1.8774212774411443,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006498102711830734,
+      "loss": 0.6306,
+      "step": 37800
+    },
+    {
+      "epoch": 1.8779179497367637,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0006497705373994239,
+      "loss": 0.5537,
+      "step": 37810
+    },
+    {
+      "epoch": 1.8784146220323832,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006497308036157743,
+      "loss": 0.5905,
+      "step": 37820
+    },
+    {
+      "epoch": 1.8789112943280024,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006496910698321248,
+      "loss": 0.5726,
+      "step": 37830
+    },
+    {
+      "epoch": 1.8794079666236216,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006496513360484753,
+      "loss": 0.6031,
+      "step": 37840
+    },
+    {
+      "epoch": 1.879904638919241,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0006496116022648257,
+      "loss": 0.5986,
+      "step": 37850
+    },
+    {
+      "epoch": 1.8804013112148605,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0006495718684811761,
+      "loss": 0.5791,
+      "step": 37860
+    },
+    {
+      "epoch": 1.8808979835104798,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006495321346975267,
+      "loss": 0.5831,
+      "step": 37870
+    },
+    {
+      "epoch": 1.881394655806099,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006494924009138771,
+      "loss": 0.5692,
+      "step": 37880
+    },
+    {
+      "epoch": 1.8818913281017184,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006494526671302275,
+      "loss": 0.58,
+      "step": 37890
+    },
+    {
+      "epoch": 1.882388000397338,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.000649412933346578,
+      "loss": 0.5919,
+      "step": 37900
+    },
+    {
+      "epoch": 1.8828846726929571,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0006493731995629283,
+      "loss": 0.6251,
+      "step": 37910
+    },
+    {
+      "epoch": 1.8833813449885766,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0006493334657792788,
+      "loss": 0.579,
+      "step": 37920
+    },
+    {
+      "epoch": 1.883878017284196,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0006492937319956294,
+      "loss": 0.6264,
+      "step": 37930
+    },
+    {
+      "epoch": 1.8843746895798152,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006492539982119798,
+      "loss": 0.5701,
+      "step": 37940
+    },
+    {
+      "epoch": 1.8848713618754345,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0006492142644283302,
+      "loss": 0.604,
+      "step": 37950
+    },
+    {
+      "epoch": 1.885368034171054,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006491745306446807,
+      "loss": 0.5774,
+      "step": 37960
+    },
+    {
+      "epoch": 1.8858647064666734,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006491347968610311,
+      "loss": 0.5748,
+      "step": 37970
+    },
+    {
+      "epoch": 1.8863613787622926,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0006490950630773816,
+      "loss": 0.6001,
+      "step": 37980
+    },
+    {
+      "epoch": 1.886858051057912,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.000649055329293732,
+      "loss": 0.6058,
+      "step": 37990
+    },
+    {
+      "epoch": 1.8873547233535315,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0006490155955100825,
+      "loss": 0.5758,
+      "step": 38000
+    },
+    {
+      "epoch": 1.8878513956491507,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0006489758617264329,
+      "loss": 0.5878,
+      "step": 38010
+    },
+    {
+      "epoch": 1.88834806794477,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0006489361279427833,
+      "loss": 0.5932,
+      "step": 38020
+    },
+    {
+      "epoch": 1.8888447402403894,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006488963941591339,
+      "loss": 0.5892,
+      "step": 38030
+    },
+    {
+      "epoch": 1.8893414125360088,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006488566603754843,
+      "loss": 0.5921,
+      "step": 38040
+    },
+    {
+      "epoch": 1.889838084831628,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006488169265918347,
+      "loss": 0.5968,
+      "step": 38050
+    },
+    {
+      "epoch": 1.8903347571272473,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0006487771928081852,
+      "loss": 0.5817,
+      "step": 38060
+    },
+    {
+      "epoch": 1.8908314294228667,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006487374590245356,
+      "loss": 0.6205,
+      "step": 38070
+    },
+    {
+      "epoch": 1.8913281017184862,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0006486977252408861,
+      "loss": 0.5994,
+      "step": 38080
+    },
+    {
+      "epoch": 1.8918247740141054,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006486579914572366,
+      "loss": 0.5979,
+      "step": 38090
+    },
+    {
+      "epoch": 1.8923214463097249,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.000648618257673587,
+      "loss": 0.6016,
+      "step": 38100
+    },
+    {
+      "epoch": 1.8928181186053443,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0006485785238899374,
+      "loss": 0.615,
+      "step": 38110
+    },
+    {
+      "epoch": 1.8933147909009636,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.000648538790106288,
+      "loss": 0.5877,
+      "step": 38120
+    },
+    {
+      "epoch": 1.8938114631965828,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006484990563226384,
+      "loss": 0.594,
+      "step": 38130
+    },
+    {
+      "epoch": 1.8943081354922022,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0006484593225389888,
+      "loss": 0.5847,
+      "step": 38140
+    },
+    {
+      "epoch": 1.8948048077878217,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0006484195887553393,
+      "loss": 0.5837,
+      "step": 38150
+    },
+    {
+      "epoch": 1.895301480083441,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0006483798549716897,
+      "loss": 0.569,
+      "step": 38160
+    },
+    {
+      "epoch": 1.8957981523790601,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0006483401211880401,
+      "loss": 0.5629,
+      "step": 38170
+    },
+    {
+      "epoch": 1.8962948246746798,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0006483003874043906,
+      "loss": 0.592,
+      "step": 38180
+    },
+    {
+      "epoch": 1.896791496970299,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006482606536207411,
+      "loss": 0.5981,
+      "step": 38190
+    },
+    {
+      "epoch": 1.8972881692659183,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0006482209198370915,
+      "loss": 0.5692,
+      "step": 38200
+    },
+    {
+      "epoch": 1.8977848415615377,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0006481811860534419,
+      "loss": 0.582,
+      "step": 38210
+    },
+    {
+      "epoch": 1.8982815138571572,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0006481414522697924,
+      "loss": 0.5871,
+      "step": 38220
+    },
+    {
+      "epoch": 1.8987781861527764,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.000648101718486143,
+      "loss": 0.5967,
+      "step": 38230
+    },
+    {
+      "epoch": 1.8992748584483956,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0006480619847024933,
+      "loss": 0.5846,
+      "step": 38240
+    },
+    {
+      "epoch": 1.899771530744015,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006480222509188438,
+      "loss": 0.5886,
+      "step": 38250
+    },
+    {
+      "epoch": 1.9002682030396345,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006479825171351942,
+      "loss": 0.593,
+      "step": 38260
+    },
+    {
+      "epoch": 1.9007648753352537,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0006479427833515446,
+      "loss": 0.5926,
+      "step": 38270
+    },
+    {
+      "epoch": 1.9012615476308732,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0006479030495678952,
+      "loss": 0.5937,
+      "step": 38280
+    },
+    {
+      "epoch": 1.9017582199264926,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006478633157842456,
+      "loss": 0.5817,
+      "step": 38290
+    },
+    {
+      "epoch": 1.9022548922221119,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.000647823582000596,
+      "loss": 0.5878,
+      "step": 38300
+    },
+    {
+      "epoch": 1.902751564517731,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006477838482169465,
+      "loss": 0.6043,
+      "step": 38310
+    },
+    {
+      "epoch": 1.9032482368133505,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006477441144332969,
+      "loss": 0.5627,
+      "step": 38320
+    },
+    {
+      "epoch": 1.90374490910897,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006477043806496474,
+      "loss": 0.578,
+      "step": 38330
+    },
+    {
+      "epoch": 1.9042415814045892,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0006476646468659979,
+      "loss": 0.5814,
+      "step": 38340
+    },
+    {
+      "epoch": 1.9047382537002084,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006476249130823483,
+      "loss": 0.6276,
+      "step": 38350
+    },
+    {
+      "epoch": 1.905234925995828,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006475851792986987,
+      "loss": 0.5758,
+      "step": 38360
+    },
+    {
+      "epoch": 1.9057315982914473,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0006475454455150491,
+      "loss": 0.5622,
+      "step": 38370
+    },
+    {
+      "epoch": 1.9062282705870666,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006475057117313997,
+      "loss": 0.5823,
+      "step": 38380
+    },
+    {
+      "epoch": 1.906724942882686,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0006474659779477502,
+      "loss": 0.5854,
+      "step": 38390
+    },
+    {
+      "epoch": 1.9072216151783055,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0006474262441641005,
+      "loss": 0.6088,
+      "step": 38400
+    },
+    {
+      "epoch": 1.9077182874739247,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.000647386510380451,
+      "loss": 0.5652,
+      "step": 38410
+    },
+    {
+      "epoch": 1.908214959769544,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006473467765968015,
+      "loss": 0.6183,
+      "step": 38420
+    },
+    {
+      "epoch": 1.9087116320651634,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006473070428131519,
+      "loss": 0.6121,
+      "step": 38430
+    },
+    {
+      "epoch": 1.9092083043607828,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006472673090295024,
+      "loss": 0.5846,
+      "step": 38440
+    },
+    {
+      "epoch": 1.909704976656402,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006472275752458528,
+      "loss": 0.5984,
+      "step": 38450
+    },
+    {
+      "epoch": 1.9102016489520215,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006471878414622032,
+      "loss": 0.5955,
+      "step": 38460
+    },
+    {
+      "epoch": 1.910698321247641,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006471481076785537,
+      "loss": 0.5847,
+      "step": 38470
+    },
+    {
+      "epoch": 1.9111949935432602,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006471083738949042,
+      "loss": 0.561,
+      "step": 38480
+    },
+    {
+      "epoch": 1.9116916658388794,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0006470686401112546,
+      "loss": 0.5744,
+      "step": 38490
+    },
+    {
+      "epoch": 1.9121883381344988,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006470289063276051,
+      "loss": 0.581,
+      "step": 38500
+    },
+    {
+      "epoch": 1.9126850104301183,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006469891725439555,
+      "loss": 0.5776,
+      "step": 38510
+    },
+    {
+      "epoch": 1.9131816827257375,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0006469494387603059,
+      "loss": 0.5793,
+      "step": 38520
+    },
+    {
+      "epoch": 1.9136783550213567,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0006469097049766565,
+      "loss": 0.5699,
+      "step": 38530
+    },
+    {
+      "epoch": 1.9141750273169764,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006468699711930069,
+      "loss": 0.584,
+      "step": 38540
+    },
+    {
+      "epoch": 1.9146716996125956,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0006468302374093574,
+      "loss": 0.5715,
+      "step": 38550
+    },
+    {
+      "epoch": 1.9151683719082149,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006467905036257078,
+      "loss": 0.5985,
+      "step": 38560
+    },
+    {
+      "epoch": 1.9156650442038343,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006467507698420582,
+      "loss": 0.5708,
+      "step": 38570
+    },
+    {
+      "epoch": 1.9161617164994538,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0006467110360584088,
+      "loss": 0.5726,
+      "step": 38580
+    },
+    {
+      "epoch": 1.916658388795073,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0006466713022747591,
+      "loss": 0.5902,
+      "step": 38590
+    },
+    {
+      "epoch": 1.9171550610906922,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0006466315684911096,
+      "loss": 0.5852,
+      "step": 38600
+    },
+    {
+      "epoch": 1.9176517333863117,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0006465918347074601,
+      "loss": 0.5814,
+      "step": 38610
+    },
+    {
+      "epoch": 1.9181484056819311,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0006465521009238104,
+      "loss": 0.5683,
+      "step": 38620
+    },
+    {
+      "epoch": 1.9186450779775504,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.000646512367140161,
+      "loss": 0.5598,
+      "step": 38630
+    },
+    {
+      "epoch": 1.9191417502731698,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006464726333565114,
+      "loss": 0.5908,
+      "step": 38640
+    },
+    {
+      "epoch": 1.9196384225687892,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0006464328995728618,
+      "loss": 0.5623,
+      "step": 38650
+    },
+    {
+      "epoch": 1.9201350948644085,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006463931657892123,
+      "loss": 0.5959,
+      "step": 38660
+    },
+    {
+      "epoch": 1.9206317671600277,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0006463534320055627,
+      "loss": 0.5941,
+      "step": 38670
+    },
+    {
+      "epoch": 1.9211284394556472,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006463136982219133,
+      "loss": 0.6107,
+      "step": 38680
+    },
+    {
+      "epoch": 1.9216251117512666,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006462739644382637,
+      "loss": 0.5719,
+      "step": 38690
+    },
+    {
+      "epoch": 1.9221217840468858,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006462342306546141,
+      "loss": 0.5812,
+      "step": 38700
+    },
+    {
+      "epoch": 1.922618456342505,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0006461944968709646,
+      "loss": 0.6058,
+      "step": 38710
+    },
+    {
+      "epoch": 1.9231151286381247,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.000646154763087315,
+      "loss": 0.5757,
+      "step": 38720
+    },
+    {
+      "epoch": 1.923611800933744,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0006461150293036655,
+      "loss": 0.5796,
+      "step": 38730
+    },
+    {
+      "epoch": 1.9241084732293632,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.000646075295520016,
+      "loss": 0.5969,
+      "step": 38740
+    },
+    {
+      "epoch": 1.9246051455249826,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0006460355617363664,
+      "loss": 0.6022,
+      "step": 38750
+    },
+    {
+      "epoch": 1.925101817820602,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006459958279527168,
+      "loss": 0.5901,
+      "step": 38760
+    },
+    {
+      "epoch": 1.9255984901162213,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006459560941690673,
+      "loss": 0.5949,
+      "step": 38770
+    },
+    {
+      "epoch": 1.9260951624118405,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006459163603854176,
+      "loss": 0.565,
+      "step": 38780
+    },
+    {
+      "epoch": 1.92659183470746,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0006458766266017682,
+      "loss": 0.5685,
+      "step": 38790
+    },
+    {
+      "epoch": 1.9270885070030794,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006458368928181187,
+      "loss": 0.5727,
+      "step": 38800
+    },
+    {
+      "epoch": 1.9275851792986987,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.000645797159034469,
+      "loss": 0.5729,
+      "step": 38810
+    },
+    {
+      "epoch": 1.928081851594318,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0006457574252508195,
+      "loss": 0.5759,
+      "step": 38820
+    },
+    {
+      "epoch": 1.9285785238899376,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00064571769146717,
+      "loss": 0.5774,
+      "step": 38830
+    },
+    {
+      "epoch": 1.9290751961855568,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006456779576835205,
+      "loss": 0.596,
+      "step": 38840
+    },
+    {
+      "epoch": 1.929571868481176,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0006456382238998709,
+      "loss": 0.5776,
+      "step": 38850
+    },
+    {
+      "epoch": 1.9300685407767955,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0006455984901162213,
+      "loss": 0.5806,
+      "step": 38860
+    },
+    {
+      "epoch": 1.930565213072415,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0006455587563325718,
+      "loss": 0.5667,
+      "step": 38870
+    },
+    {
+      "epoch": 1.9310618853680341,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0006455190225489223,
+      "loss": 0.5859,
+      "step": 38880
+    },
+    {
+      "epoch": 1.9315585576636534,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006454792887652727,
+      "loss": 0.5778,
+      "step": 38890
+    },
+    {
+      "epoch": 1.932055229959273,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006454395549816232,
+      "loss": 0.5954,
+      "step": 38900
+    },
+    {
+      "epoch": 1.9325519022548923,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006453998211979736,
+      "loss": 0.6011,
+      "step": 38910
+    },
+    {
+      "epoch": 1.9330485745505115,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.000645360087414324,
+      "loss": 0.5982,
+      "step": 38920
+    },
+    {
+      "epoch": 1.933545246846131,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006453203536306746,
+      "loss": 0.6026,
+      "step": 38930
+    },
+    {
+      "epoch": 1.9340419191417504,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.000645280619847025,
+      "loss": 0.5653,
+      "step": 38940
+    },
+    {
+      "epoch": 1.9345385914373696,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006452408860633754,
+      "loss": 0.5959,
+      "step": 38950
+    },
+    {
+      "epoch": 1.9350352637329888,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006452011522797259,
+      "loss": 0.6307,
+      "step": 38960
+    },
+    {
+      "epoch": 1.9355319360286083,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0006451614184960763,
+      "loss": 0.6209,
+      "step": 38970
+    },
+    {
+      "epoch": 1.9360286083242277,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006451216847124268,
+      "loss": 0.5748,
+      "step": 38980
+    },
+    {
+      "epoch": 1.936525280619847,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006450819509287773,
+      "loss": 0.5798,
+      "step": 38990
+    },
+    {
+      "epoch": 1.9370219529154664,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0006450422171451277,
+      "loss": 0.6054,
+      "step": 39000
+    },
+    {
+      "epoch": 1.9375186252110859,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006450024833614781,
+      "loss": 0.5824,
+      "step": 39010
+    },
+    {
+      "epoch": 1.938015297506705,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006449627495778286,
+      "loss": 0.6086,
+      "step": 39020
+    },
+    {
+      "epoch": 1.9385119698023243,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.000644923015794179,
+      "loss": 0.6038,
+      "step": 39030
+    },
+    {
+      "epoch": 1.9390086420979438,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0006448832820105295,
+      "loss": 0.5818,
+      "step": 39040
+    },
+    {
+      "epoch": 1.9395053143935632,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006448435482268799,
+      "loss": 0.5801,
+      "step": 39050
+    },
+    {
+      "epoch": 1.9400019866891824,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0006448038144432304,
+      "loss": 0.6027,
+      "step": 39060
+    },
+    {
+      "epoch": 1.9404986589848017,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0006447640806595808,
+      "loss": 0.5803,
+      "step": 39070
+    },
+    {
+      "epoch": 1.9409953312804213,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006447243468759312,
+      "loss": 0.584,
+      "step": 39080
+    },
+    {
+      "epoch": 1.9414920035760406,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0006446846130922818,
+      "loss": 0.5877,
+      "step": 39090
+    },
+    {
+      "epoch": 1.9419886758716598,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0006446448793086322,
+      "loss": 0.6003,
+      "step": 39100
+    },
+    {
+      "epoch": 1.9424853481672792,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0006446051455249826,
+      "loss": 0.5956,
+      "step": 39110
+    },
+    {
+      "epoch": 1.9429820204628987,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0006445654117413331,
+      "loss": 0.5991,
+      "step": 39120
+    },
+    {
+      "epoch": 1.943478692758518,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006445256779576835,
+      "loss": 0.5997,
+      "step": 39130
+    },
+    {
+      "epoch": 1.9439753650541371,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.000644485944174034,
+      "loss": 0.5702,
+      "step": 39140
+    },
+    {
+      "epoch": 1.9444720373497566,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006444462103903845,
+      "loss": 0.594,
+      "step": 39150
+    },
+    {
+      "epoch": 1.944968709645376,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006444064766067349,
+      "loss": 0.5686,
+      "step": 39160
+    },
+    {
+      "epoch": 1.9454653819409953,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0006443667428230853,
+      "loss": 0.6089,
+      "step": 39170
+    },
+    {
+      "epoch": 1.9459620542366147,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0006443270090394359,
+      "loss": 0.5801,
+      "step": 39180
+    },
+    {
+      "epoch": 1.9464587265322342,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006442872752557863,
+      "loss": 0.5689,
+      "step": 39190
+    },
+    {
+      "epoch": 1.9469553988278534,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006442475414721367,
+      "loss": 0.5747,
+      "step": 39200
+    },
+    {
+      "epoch": 1.9474520711234726,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006442078076884872,
+      "loss": 0.5758,
+      "step": 39210
+    },
+    {
+      "epoch": 1.947948743419092,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006441680739048376,
+      "loss": 0.5688,
+      "step": 39220
+    },
+    {
+      "epoch": 1.9484454157147115,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.000644128340121188,
+      "loss": 0.5876,
+      "step": 39230
+    },
+    {
+      "epoch": 1.9489420880103308,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006440886063375385,
+      "loss": 0.565,
+      "step": 39240
+    },
+    {
+      "epoch": 1.94943876030595,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.000644048872553889,
+      "loss": 0.6057,
+      "step": 39250
+    },
+    {
+      "epoch": 1.9499354326015694,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006440091387702394,
+      "loss": 0.5712,
+      "step": 39260
+    },
+    {
+      "epoch": 1.9504321048971889,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0006439694049865898,
+      "loss": 0.6069,
+      "step": 39270
+    },
+    {
+      "epoch": 1.950928777192808,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0006439296712029403,
+      "loss": 0.5796,
+      "step": 39280
+    },
+    {
+      "epoch": 1.9514254494884276,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0006438899374192909,
+      "loss": 0.5689,
+      "step": 39290
+    },
+    {
+      "epoch": 1.951922121784047,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006438502036356412,
+      "loss": 0.5608,
+      "step": 39300
+    },
+    {
+      "epoch": 1.9524187940796662,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006438104698519917,
+      "loss": 0.5871,
+      "step": 39310
+    },
+    {
+      "epoch": 1.9529154663752855,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0006437707360683421,
+      "loss": 0.6047,
+      "step": 39320
+    },
+    {
+      "epoch": 1.953412138670905,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006437310022846925,
+      "loss": 0.5833,
+      "step": 39330
+    },
+    {
+      "epoch": 1.9539088109665244,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006436912685010431,
+      "loss": 0.6137,
+      "step": 39340
+    },
+    {
+      "epoch": 1.9544054832621436,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0006436515347173935,
+      "loss": 0.5939,
+      "step": 39350
+    },
+    {
+      "epoch": 1.954902155557763,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006436118009337439,
+      "loss": 0.5941,
+      "step": 39360
+    },
+    {
+      "epoch": 1.9553988278533825,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0006435720671500944,
+      "loss": 0.5745,
+      "step": 39370
+    },
+    {
+      "epoch": 1.9558955001490017,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0006435323333664448,
+      "loss": 0.5842,
+      "step": 39380
+    },
+    {
+      "epoch": 1.956392172444621,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0006434925995827953,
+      "loss": 0.5685,
+      "step": 39390
+    },
+    {
+      "epoch": 1.9568888447402404,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0006434528657991458,
+      "loss": 0.5829,
+      "step": 39400
+    },
+    {
+      "epoch": 1.9573855170358598,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006434131320154962,
+      "loss": 0.6016,
+      "step": 39410
+    },
+    {
+      "epoch": 1.957882189331479,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0006433733982318467,
+      "loss": 0.559,
+      "step": 39420
+    },
+    {
+      "epoch": 1.9583788616270983,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.000643333664448197,
+      "loss": 0.5896,
+      "step": 39430
+    },
+    {
+      "epoch": 1.9588755339227177,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0006432939306645476,
+      "loss": 0.5995,
+      "step": 39440
+    },
+    {
+      "epoch": 1.9593722062183372,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0006432541968808981,
+      "loss": 0.5651,
+      "step": 39450
+    },
+    {
+      "epoch": 1.9598688785139564,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006432144630972484,
+      "loss": 0.5804,
+      "step": 39460
+    },
+    {
+      "epoch": 1.9603655508095759,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0006431747293135989,
+      "loss": 0.5789,
+      "step": 39470
+    },
+    {
+      "epoch": 1.9608622231051953,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0006431349955299495,
+      "loss": 0.6015,
+      "step": 39480
+    },
+    {
+      "epoch": 1.9613588954008145,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0006430952617462998,
+      "loss": 0.5799,
+      "step": 39490
+    },
+    {
+      "epoch": 1.9618555676964338,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006430555279626503,
+      "loss": 0.6069,
+      "step": 39500
+    },
+    {
+      "epoch": 1.9623522399920532,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0006430157941790007,
+      "loss": 0.589,
+      "step": 39510
+    },
+    {
+      "epoch": 1.9628489122876727,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0006429760603953511,
+      "loss": 0.6054,
+      "step": 39520
+    },
+    {
+      "epoch": 1.963345584583292,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006429363266117016,
+      "loss": 0.5633,
+      "step": 39530
+    },
+    {
+      "epoch": 1.9638422568789113,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0006428965928280521,
+      "loss": 0.5687,
+      "step": 39540
+    },
+    {
+      "epoch": 1.9643389291745308,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0006428568590444025,
+      "loss": 0.6061,
+      "step": 39550
+    },
+    {
+      "epoch": 1.96483560147015,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.000642817125260753,
+      "loss": 0.5849,
+      "step": 39560
+    },
+    {
+      "epoch": 1.9653322737657692,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006427773914771034,
+      "loss": 0.5754,
+      "step": 39570
+    },
+    {
+      "epoch": 1.9658289460613887,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.000642737657693454,
+      "loss": 0.5902,
+      "step": 39580
+    },
+    {
+      "epoch": 1.9663256183570081,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006426979239098044,
+      "loss": 0.5815,
+      "step": 39590
+    },
+    {
+      "epoch": 1.9668222906526274,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006426581901261548,
+      "loss": 0.584,
+      "step": 39600
+    },
+    {
+      "epoch": 1.9673189629482466,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006426184563425053,
+      "loss": 0.5894,
+      "step": 39610
+    },
+    {
+      "epoch": 1.967815635243866,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006425787225588557,
+      "loss": 0.5796,
+      "step": 39620
+    },
+    {
+      "epoch": 1.9683123075394855,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0006425389887752061,
+      "loss": 0.6132,
+      "step": 39630
+    },
+    {
+      "epoch": 1.9688089798351047,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006424992549915567,
+      "loss": 0.5728,
+      "step": 39640
+    },
+    {
+      "epoch": 1.9693056521307242,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.000642459521207907,
+      "loss": 0.5985,
+      "step": 39650
+    },
+    {
+      "epoch": 1.9698023244263436,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006424197874242575,
+      "loss": 0.5922,
+      "step": 39660
+    },
+    {
+      "epoch": 1.9702989967219628,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.000642380053640608,
+      "loss": 0.6181,
+      "step": 39670
+    },
+    {
+      "epoch": 1.970795669017582,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0006423403198569583,
+      "loss": 0.5975,
+      "step": 39680
+    },
+    {
+      "epoch": 1.9712923413132015,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0006423005860733089,
+      "loss": 0.6012,
+      "step": 39690
+    },
+    {
+      "epoch": 1.971789013608821,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006422608522896593,
+      "loss": 0.567,
+      "step": 39700
+    },
+    {
+      "epoch": 1.9722856859044402,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006422211185060097,
+      "loss": 0.5704,
+      "step": 39710
+    },
+    {
+      "epoch": 1.9727823582000596,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0006421813847223602,
+      "loss": 0.6102,
+      "step": 39720
+    },
+    {
+      "epoch": 1.973279030495679,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0006421416509387106,
+      "loss": 0.6061,
+      "step": 39730
+    },
+    {
+      "epoch": 1.9737757027912983,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006421019171550612,
+      "loss": 0.5885,
+      "step": 39740
+    },
+    {
+      "epoch": 1.9742723750869176,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0006420621833714116,
+      "loss": 0.5902,
+      "step": 39750
+    },
+    {
+      "epoch": 1.974769047382537,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.000642022449587762,
+      "loss": 0.5694,
+      "step": 39760
+    },
+    {
+      "epoch": 1.9752657196781565,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006419827158041125,
+      "loss": 0.6038,
+      "step": 39770
+    },
+    {
+      "epoch": 1.9757623919737757,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0006419429820204629,
+      "loss": 0.5931,
+      "step": 39780
+    },
+    {
+      "epoch": 1.976259064269395,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0006419032482368134,
+      "loss": 0.5722,
+      "step": 39790
+    },
+    {
+      "epoch": 1.9767557365650144,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006418635144531639,
+      "loss": 0.5876,
+      "step": 39800
+    },
+    {
+      "epoch": 1.9772524088606338,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006418237806695143,
+      "loss": 0.5602,
+      "step": 39810
+    },
+    {
+      "epoch": 1.977749081156253,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0006417840468858647,
+      "loss": 0.5556,
+      "step": 39820
+    },
+    {
+      "epoch": 1.9782457534518725,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006417443131022152,
+      "loss": 0.5809,
+      "step": 39830
+    },
+    {
+      "epoch": 1.978742425747492,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006417045793185656,
+      "loss": 0.5805,
+      "step": 39840
+    },
+    {
+      "epoch": 1.9792390980431112,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006416648455349161,
+      "loss": 0.5848,
+      "step": 39850
+    },
+    {
+      "epoch": 1.9797357703387304,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006416251117512666,
+      "loss": 0.5994,
+      "step": 39860
+    },
+    {
+      "epoch": 1.9802324426343498,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.000641585377967617,
+      "loss": 0.5487,
+      "step": 39870
+    },
+    {
+      "epoch": 1.9807291149299693,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006415456441839674,
+      "loss": 0.5817,
+      "step": 39880
+    },
+    {
+      "epoch": 1.9812257872255885,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.000641505910400318,
+      "loss": 0.5628,
+      "step": 39890
+    },
+    {
+      "epoch": 1.981722459521208,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006414661766166684,
+      "loss": 0.5559,
+      "step": 39900
+    },
+    {
+      "epoch": 1.9822191318168274,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0006414264428330188,
+      "loss": 0.579,
+      "step": 39910
+    },
+    {
+      "epoch": 1.9827158041124466,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0006413867090493692,
+      "loss": 0.5864,
+      "step": 39920
+    },
+    {
+      "epoch": 1.9832124764080659,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0006413469752657197,
+      "loss": 0.5798,
+      "step": 39930
+    },
+    {
+      "epoch": 1.9837091487036853,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006413072414820702,
+      "loss": 0.589,
+      "step": 39940
+    },
+    {
+      "epoch": 1.9842058209993048,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006412675076984206,
+      "loss": 0.6079,
+      "step": 39950
+    },
+    {
+      "epoch": 1.984702493294924,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0006412277739147711,
+      "loss": 0.5823,
+      "step": 39960
+    },
+    {
+      "epoch": 1.9851991655905432,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0006411880401311215,
+      "loss": 0.5776,
+      "step": 39970
+    },
+    {
+      "epoch": 1.9856958378861627,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0006411483063474719,
+      "loss": 0.5796,
+      "step": 39980
+    },
+    {
+      "epoch": 1.9861925101817821,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006411085725638225,
+      "loss": 0.5989,
+      "step": 39990
+    },
+    {
+      "epoch": 1.9866891824774013,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006410688387801729,
+      "loss": 0.5959,
+      "step": 40000
+    },
+    {
+      "epoch": 1.9871858547730208,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0006410291049965233,
+      "loss": 0.5618,
+      "step": 40010
+    },
+    {
+      "epoch": 1.9876825270686402,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0006409893712128738,
+      "loss": 0.5903,
+      "step": 40020
+    },
+    {
+      "epoch": 1.9881791993642595,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0006409496374292242,
+      "loss": 0.5609,
+      "step": 40030
+    },
+    {
+      "epoch": 1.9886758716598787,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0006409099036455747,
+      "loss": 0.5721,
+      "step": 40040
+    },
+    {
+      "epoch": 1.9891725439554981,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006408701698619252,
+      "loss": 0.578,
+      "step": 40050
+    },
+    {
+      "epoch": 1.9896692162511176,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006408304360782756,
+      "loss": 0.5972,
+      "step": 40060
+    },
+    {
+      "epoch": 1.9901658885467368,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000640790702294626,
+      "loss": 0.5915,
+      "step": 40070
+    },
+    {
+      "epoch": 1.9906625608423563,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006407509685109765,
+      "loss": 0.597,
+      "step": 40080
+    },
+    {
+      "epoch": 1.9911592331379757,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.000640711234727327,
+      "loss": 0.5964,
+      "step": 40090
+    },
+    {
+      "epoch": 1.991655905433595,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0006406715009436774,
+      "loss": 0.5867,
+      "step": 40100
+    },
+    {
+      "epoch": 1.9921525777292142,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0006406317671600278,
+      "loss": 0.5844,
+      "step": 40110
+    },
+    {
+      "epoch": 1.9926492500248336,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0006405920333763783,
+      "loss": 0.6189,
+      "step": 40120
+    },
+    {
+      "epoch": 1.993145922320453,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006405522995927287,
+      "loss": 0.5775,
+      "step": 40130
+    },
+    {
+      "epoch": 1.9936425946160723,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006405125658090792,
+      "loss": 0.5647,
+      "step": 40140
+    },
+    {
+      "epoch": 1.9941392669116915,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006404728320254297,
+      "loss": 0.5857,
+      "step": 40150
+    },
+    {
+      "epoch": 1.994635939207311,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006404330982417801,
+      "loss": 0.5828,
+      "step": 40160
+    },
+    {
+      "epoch": 1.9951326115029304,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006403933644581305,
+      "loss": 0.5727,
+      "step": 40170
+    },
+    {
+      "epoch": 1.9956292837985496,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.000640353630674481,
+      "loss": 0.5982,
+      "step": 40180
+    },
+    {
+      "epoch": 1.996125956094169,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0006403138968908315,
+      "loss": 0.5826,
+      "step": 40190
+    },
+    {
+      "epoch": 1.9966226283897885,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0006402741631071819,
+      "loss": 0.5778,
+      "step": 40200
+    },
+    {
+      "epoch": 1.9971193006854078,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0006402344293235324,
+      "loss": 0.5487,
+      "step": 40210
+    },
+    {
+      "epoch": 1.997615972981027,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0006401946955398828,
+      "loss": 0.5894,
+      "step": 40220
+    },
+    {
+      "epoch": 1.9981126452766464,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006401549617562332,
+      "loss": 0.6065,
+      "step": 40230
+    },
+    {
+      "epoch": 1.998609317572266,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006401152279725838,
+      "loss": 0.5924,
+      "step": 40240
+    },
+    {
+      "epoch": 1.9991059898678851,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0006400754941889342,
+      "loss": 0.5868,
+      "step": 40250
+    },
+    {
+      "epoch": 1.9996026621635046,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006400357604052846,
+      "loss": 0.5794,
+      "step": 40260
+    },
+    {
+      "epoch": 2.000099334459124,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006399960266216351,
+      "loss": 0.5783,
+      "step": 40270
+    },
+    {
+      "epoch": 2.0005960067547433,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006399562928379855,
+      "loss": 0.5969,
+      "step": 40280
+    },
+    {
+      "epoch": 2.0010926790503625,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.000639916559054336,
+      "loss": 0.5477,
+      "step": 40290
+    },
+    {
+      "epoch": 2.0015893513459817,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0006398768252706864,
+      "loss": 0.5871,
+      "step": 40300
+    },
+    {
+      "epoch": 2.0020860236416014,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0006398370914870369,
+      "loss": 0.5876,
+      "step": 40310
+    },
+    {
+      "epoch": 2.0025826959372206,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0006397973577033874,
+      "loss": 0.5657,
+      "step": 40320
+    },
+    {
+      "epoch": 2.00307936823284,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006397576239197377,
+      "loss": 0.5332,
+      "step": 40330
+    },
+    {
+      "epoch": 2.0035760405284595,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0006397178901360883,
+      "loss": 0.5807,
+      "step": 40340
+    },
+    {
+      "epoch": 2.0040727128240787,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006396781563524388,
+      "loss": 0.5744,
+      "step": 40350
+    },
+    {
+      "epoch": 2.004569385119698,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006396384225687891,
+      "loss": 0.5867,
+      "step": 40360
+    },
+    {
+      "epoch": 2.005066057415317,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006395986887851396,
+      "loss": 0.583,
+      "step": 40370
+    },
+    {
+      "epoch": 2.005562729710937,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00063955895500149,
+      "loss": 0.5734,
+      "step": 40380
+    },
+    {
+      "epoch": 2.006059402006556,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006395192212178404,
+      "loss": 0.567,
+      "step": 40390
+    },
+    {
+      "epoch": 2.0065560743021753,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.000639479487434191,
+      "loss": 0.5576,
+      "step": 40400
+    },
+    {
+      "epoch": 2.007052746597795,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006394397536505414,
+      "loss": 0.606,
+      "step": 40410
+    },
+    {
+      "epoch": 2.007549418893414,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006394000198668918,
+      "loss": 0.5705,
+      "step": 40420
+    },
+    {
+      "epoch": 2.0080460911890334,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0006393602860832423,
+      "loss": 0.5697,
+      "step": 40430
+    },
+    {
+      "epoch": 2.0085427634846527,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0006393205522995927,
+      "loss": 0.5761,
+      "step": 40440
+    },
+    {
+      "epoch": 2.0090394357802723,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0006392808185159432,
+      "loss": 0.5834,
+      "step": 40450
+    },
+    {
+      "epoch": 2.0095361080758916,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006392410847322937,
+      "loss": 0.5585,
+      "step": 40460
+    },
+    {
+      "epoch": 2.010032780371511,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006392013509486441,
+      "loss": 0.5593,
+      "step": 40470
+    },
+    {
+      "epoch": 2.01052945266713,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006391616171649946,
+      "loss": 0.553,
+      "step": 40480
+    },
+    {
+      "epoch": 2.0110261249627497,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.000639121883381345,
+      "loss": 0.571,
+      "step": 40490
+    },
+    {
+      "epoch": 2.011522797258369,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0006390821495976955,
+      "loss": 0.5681,
+      "step": 40500
+    },
+    {
+      "epoch": 2.012019469553988,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.000639042415814046,
+      "loss": 0.5786,
+      "step": 40510
+    },
+    {
+      "epoch": 2.012516141849608,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006390026820303963,
+      "loss": 0.5697,
+      "step": 40520
+    },
+    {
+      "epoch": 2.013012814145227,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006389629482467468,
+      "loss": 0.5741,
+      "step": 40530
+    },
+    {
+      "epoch": 2.0135094864408463,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006389232144630974,
+      "loss": 0.5523,
+      "step": 40540
+    },
+    {
+      "epoch": 2.0140061587364655,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006388834806794477,
+      "loss": 0.5956,
+      "step": 40550
+    },
+    {
+      "epoch": 2.014502831032085,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0006388437468957982,
+      "loss": 0.5698,
+      "step": 40560
+    },
+    {
+      "epoch": 2.0149995033277044,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006388040131121486,
+      "loss": 0.5914,
+      "step": 40570
+    },
+    {
+      "epoch": 2.0154961756233236,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.000638764279328499,
+      "loss": 0.5779,
+      "step": 40580
+    },
+    {
+      "epoch": 2.0159928479189433,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0006387245455448495,
+      "loss": 0.5761,
+      "step": 40590
+    },
+    {
+      "epoch": 2.0164895202145625,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0006386848117612,
+      "loss": 0.5418,
+      "step": 40600
+    },
+    {
+      "epoch": 2.0169861925101817,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0006386450779775504,
+      "loss": 0.5969,
+      "step": 40610
+    },
+    {
+      "epoch": 2.017482864805801,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006386053441939009,
+      "loss": 0.5575,
+      "step": 40620
+    },
+    {
+      "epoch": 2.0179795371014206,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0006385656104102513,
+      "loss": 0.5786,
+      "step": 40630
+    },
+    {
+      "epoch": 2.01847620939704,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0006385258766266019,
+      "loss": 0.5354,
+      "step": 40640
+    },
+    {
+      "epoch": 2.018972881692659,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0006384861428429523,
+      "loss": 0.608,
+      "step": 40650
+    },
+    {
+      "epoch": 2.0194695539882783,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0006384464090593027,
+      "loss": 0.5686,
+      "step": 40660
+    },
+    {
+      "epoch": 2.019966226283898,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006384066752756532,
+      "loss": 0.5938,
+      "step": 40670
+    },
+    {
+      "epoch": 2.020462898579517,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006383669414920036,
+      "loss": 0.5727,
+      "step": 40680
+    },
+    {
+      "epoch": 2.0209595708751364,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.000638327207708354,
+      "loss": 0.5695,
+      "step": 40690
+    },
+    {
+      "epoch": 2.021456243170756,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0006382874739247046,
+      "loss": 0.5923,
+      "step": 40700
+    },
+    {
+      "epoch": 2.0219529154663753,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006382477401410549,
+      "loss": 0.5913,
+      "step": 40710
+    },
+    {
+      "epoch": 2.0224495877619946,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006382080063574054,
+      "loss": 0.5862,
+      "step": 40720
+    },
+    {
+      "epoch": 2.022946260057614,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006381682725737559,
+      "loss": 0.6107,
+      "step": 40730
+    },
+    {
+      "epoch": 2.0234429323532335,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006381285387901062,
+      "loss": 0.6058,
+      "step": 40740
+    },
+    {
+      "epoch": 2.0239396046488527,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0006380888050064568,
+      "loss": 0.5606,
+      "step": 40750
+    },
+    {
+      "epoch": 2.024436276944472,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0006380490712228073,
+      "loss": 0.572,
+      "step": 40760
+    },
+    {
+      "epoch": 2.0249329492400916,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006380093374391577,
+      "loss": 0.5759,
+      "step": 40770
+    },
+    {
+      "epoch": 2.025429621535711,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006379696036555081,
+      "loss": 0.6079,
+      "step": 40780
+    },
+    {
+      "epoch": 2.02592629383133,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006379298698718585,
+      "loss": 0.5937,
+      "step": 40790
+    },
+    {
+      "epoch": 2.0264229661269493,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006378901360882091,
+      "loss": 0.5953,
+      "step": 40800
+    },
+    {
+      "epoch": 2.026919638422569,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006378504023045595,
+      "loss": 0.5788,
+      "step": 40810
+    },
+    {
+      "epoch": 2.027416310718188,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006378106685209099,
+      "loss": 0.5634,
+      "step": 40820
+    },
+    {
+      "epoch": 2.0279129830138074,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0006377709347372604,
+      "loss": 0.5917,
+      "step": 40830
+    },
+    {
+      "epoch": 2.0284096553094266,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0006377312009536108,
+      "loss": 0.5721,
+      "step": 40840
+    },
+    {
+      "epoch": 2.0289063276050463,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0006376914671699613,
+      "loss": 0.5689,
+      "step": 40850
+    },
+    {
+      "epoch": 2.0294029999006655,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006376517333863118,
+      "loss": 0.5732,
+      "step": 40860
+    },
+    {
+      "epoch": 2.0298996721962848,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0006376119996026622,
+      "loss": 0.5649,
+      "step": 40870
+    },
+    {
+      "epoch": 2.0303963444919044,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006375722658190126,
+      "loss": 0.5461,
+      "step": 40880
+    },
+    {
+      "epoch": 2.0308930167875237,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006375325320353631,
+      "loss": 0.5322,
+      "step": 40890
+    },
+    {
+      "epoch": 2.031389689083143,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0006374927982517135,
+      "loss": 0.6145,
+      "step": 40900
+    },
+    {
+      "epoch": 2.031886361378762,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.000637453064468064,
+      "loss": 0.5766,
+      "step": 40910
+    },
+    {
+      "epoch": 2.032383033674382,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006374133306844145,
+      "loss": 0.563,
+      "step": 40920
+    },
+    {
+      "epoch": 2.032879705970001,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006373735969007649,
+      "loss": 0.5614,
+      "step": 40930
+    },
+    {
+      "epoch": 2.0333763782656202,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006373338631171153,
+      "loss": 0.5522,
+      "step": 40940
+    },
+    {
+      "epoch": 2.03387305056124,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006372941293334659,
+      "loss": 0.5635,
+      "step": 40950
+    },
+    {
+      "epoch": 2.034369722856859,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006372543955498163,
+      "loss": 0.5879,
+      "step": 40960
+    },
+    {
+      "epoch": 2.0348663951524784,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006372146617661667,
+      "loss": 0.5443,
+      "step": 40970
+    },
+    {
+      "epoch": 2.0353630674480976,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006371749279825171,
+      "loss": 0.5672,
+      "step": 40980
+    },
+    {
+      "epoch": 2.0358597397437173,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006371351941988676,
+      "loss": 0.5633,
+      "step": 40990
+    },
+    {
+      "epoch": 2.0363564120393365,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0006370954604152181,
+      "loss": 0.588,
+      "step": 41000
+    },
+    {
+      "epoch": 2.0368530843349557,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0006370557266315685,
+      "loss": 0.5558,
+      "step": 41010
+    },
+    {
+      "epoch": 2.037349756630575,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.000637015992847919,
+      "loss": 0.5913,
+      "step": 41020
+    },
+    {
+      "epoch": 2.0378464289261946,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006369762590642694,
+      "loss": 0.5494,
+      "step": 41030
+    },
+    {
+      "epoch": 2.038343101221814,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0006369365252806198,
+      "loss": 0.5775,
+      "step": 41040
+    },
+    {
+      "epoch": 2.038839773517433,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006368967914969704,
+      "loss": 0.5988,
+      "step": 41050
+    },
+    {
+      "epoch": 2.0393364458130527,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0006368570577133208,
+      "loss": 0.5296,
+      "step": 41060
+    },
+    {
+      "epoch": 2.039833118108672,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0006368173239296712,
+      "loss": 0.5901,
+      "step": 41070
+    },
+    {
+      "epoch": 2.040329790404291,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006367775901460217,
+      "loss": 0.559,
+      "step": 41080
+    },
+    {
+      "epoch": 2.0408264626999104,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006367378563623721,
+      "loss": 0.5933,
+      "step": 41090
+    },
+    {
+      "epoch": 2.04132313499553,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006366981225787226,
+      "loss": 0.5288,
+      "step": 41100
+    },
+    {
+      "epoch": 2.0418198072911493,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006366583887950731,
+      "loss": 0.5744,
+      "step": 41110
+    },
+    {
+      "epoch": 2.0423164795867685,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0006366186550114235,
+      "loss": 0.5789,
+      "step": 41120
+    },
+    {
+      "epoch": 2.042813151882388,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0006365789212277739,
+      "loss": 0.5652,
+      "step": 41130
+    },
+    {
+      "epoch": 2.0433098241780074,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0006365391874441244,
+      "loss": 0.5516,
+      "step": 41140
+    },
+    {
+      "epoch": 2.0438064964736267,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0006364994536604749,
+      "loss": 0.5759,
+      "step": 41150
+    },
+    {
+      "epoch": 2.044303168769246,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006364597198768253,
+      "loss": 0.5565,
+      "step": 41160
+    },
+    {
+      "epoch": 2.0447998410648656,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0006364199860931757,
+      "loss": 0.5778,
+      "step": 41170
+    },
+    {
+      "epoch": 2.045296513360485,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0006363802523095262,
+      "loss": 0.5653,
+      "step": 41180
+    },
+    {
+      "epoch": 2.045793185656104,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0006363405185258766,
+      "loss": 0.5975,
+      "step": 41190
+    },
+    {
+      "epoch": 2.0462898579517232,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006363007847422271,
+      "loss": 0.5602,
+      "step": 41200
+    },
+    {
+      "epoch": 2.046786530247343,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006362610509585776,
+      "loss": 0.5594,
+      "step": 41210
+    },
+    {
+      "epoch": 2.047283202542962,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0006362213171749281,
+      "loss": 0.5369,
+      "step": 41220
+    },
+    {
+      "epoch": 2.0477798748385814,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0006361815833912784,
+      "loss": 0.5881,
+      "step": 41230
+    },
+    {
+      "epoch": 2.048276547134201,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0006361418496076289,
+      "loss": 0.5492,
+      "step": 41240
+    },
+    {
+      "epoch": 2.0487732194298203,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006361021158239794,
+      "loss": 0.5637,
+      "step": 41250
+    },
+    {
+      "epoch": 2.0492698917254395,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0006360623820403298,
+      "loss": 0.5624,
+      "step": 41260
+    },
+    {
+      "epoch": 2.0497665640210587,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0006360226482566803,
+      "loss": 0.5966,
+      "step": 41270
+    },
+    {
+      "epoch": 2.0502632363166784,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006359829144730307,
+      "loss": 0.6032,
+      "step": 41280
+    },
+    {
+      "epoch": 2.0507599086122976,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0006359431806893811,
+      "loss": 0.594,
+      "step": 41290
+    },
+    {
+      "epoch": 2.051256580907917,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006359034469057317,
+      "loss": 0.5837,
+      "step": 41300
+    },
+    {
+      "epoch": 2.0517532532035365,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0006358637131220821,
+      "loss": 0.5417,
+      "step": 41310
+    },
+    {
+      "epoch": 2.0522499254991557,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0006358239793384325,
+      "loss": 0.6023,
+      "step": 41320
+    },
+    {
+      "epoch": 2.052746597794775,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.000635784245554783,
+      "loss": 0.5314,
+      "step": 41330
+    },
+    {
+      "epoch": 2.053243270090394,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0006357445117711334,
+      "loss": 0.5968,
+      "step": 41340
+    },
+    {
+      "epoch": 2.053739942386014,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0006357047779874839,
+      "loss": 0.558,
+      "step": 41350
+    },
+    {
+      "epoch": 2.054236614681633,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0006356650442038344,
+      "loss": 0.5716,
+      "step": 41360
+    },
+    {
+      "epoch": 2.0547332869772523,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006356253104201848,
+      "loss": 0.5497,
+      "step": 41370
+    },
+    {
+      "epoch": 2.0552299592728716,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006355855766365353,
+      "loss": 0.5591,
+      "step": 41380
+    },
+    {
+      "epoch": 2.0557266315684912,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0006355458428528856,
+      "loss": 0.5912,
+      "step": 41390
+    },
+    {
+      "epoch": 2.0562233038641105,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006355061090692362,
+      "loss": 0.5788,
+      "step": 41400
+    },
+    {
+      "epoch": 2.0567199761597297,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0006354663752855867,
+      "loss": 0.5456,
+      "step": 41410
+    },
+    {
+      "epoch": 2.0572166484553494,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.000635426641501937,
+      "loss": 0.5691,
+      "step": 41420
+    },
+    {
+      "epoch": 2.0577133207509686,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0006353869077182875,
+      "loss": 0.5643,
+      "step": 41430
+    },
+    {
+      "epoch": 2.058209993046588,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0006353471739346379,
+      "loss": 0.5555,
+      "step": 41440
+    },
+    {
+      "epoch": 2.058706665342207,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006353074401509884,
+      "loss": 0.5833,
+      "step": 41450
+    },
+    {
+      "epoch": 2.0592033376378267,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006352677063673389,
+      "loss": 0.5867,
+      "step": 41460
+    },
+    {
+      "epoch": 2.059700009933446,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006352279725836893,
+      "loss": 0.5708,
+      "step": 41470
+    },
+    {
+      "epoch": 2.060196682229065,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0006351882388000397,
+      "loss": 0.592,
+      "step": 41480
+    },
+    {
+      "epoch": 2.060693354524685,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006351485050163902,
+      "loss": 0.5738,
+      "step": 41490
+    },
+    {
+      "epoch": 2.061190026820304,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006351087712327407,
+      "loss": 0.5879,
+      "step": 41500
+    },
+    {
+      "epoch": 2.0616866991159233,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006350690374490912,
+      "loss": 0.5452,
+      "step": 41510
+    },
+    {
+      "epoch": 2.0621833714115425,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006350293036654416,
+      "loss": 0.5565,
+      "step": 41520
+    },
+    {
+      "epoch": 2.062680043707162,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.000634989569881792,
+      "loss": 0.5858,
+      "step": 41530
+    },
+    {
+      "epoch": 2.0631767160027814,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0006349498360981425,
+      "loss": 0.5948,
+      "step": 41540
+    },
+    {
+      "epoch": 2.0636733882984006,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.000634910102314493,
+      "loss": 0.5653,
+      "step": 41550
+    },
+    {
+      "epoch": 2.06417006059402,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006348703685308434,
+      "loss": 0.5673,
+      "step": 41560
+    },
+    {
+      "epoch": 2.0646667328896395,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006348306347471939,
+      "loss": 0.5766,
+      "step": 41570
+    },
+    {
+      "epoch": 2.0651634051852588,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006347909009635442,
+      "loss": 0.591,
+      "step": 41580
+    },
+    {
+      "epoch": 2.065660077480878,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0006347511671798947,
+      "loss": 0.5991,
+      "step": 41590
+    },
+    {
+      "epoch": 2.0661567497764977,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0006347114333962453,
+      "loss": 0.5761,
+      "step": 41600
+    },
+    {
+      "epoch": 2.066653422072117,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006346716996125956,
+      "loss": 0.5686,
+      "step": 41610
+    },
+    {
+      "epoch": 2.067150094367736,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0006346319658289461,
+      "loss": 0.5431,
+      "step": 41620
+    },
+    {
+      "epoch": 2.0676467666633553,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006345922320452966,
+      "loss": 0.5931,
+      "step": 41630
+    },
+    {
+      "epoch": 2.068143438958975,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0006345524982616469,
+      "loss": 0.5624,
+      "step": 41640
+    },
+    {
+      "epoch": 2.0686401112545942,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006345127644779975,
+      "loss": 0.6122,
+      "step": 41650
+    },
+    {
+      "epoch": 2.0691367835502135,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006344730306943479,
+      "loss": 0.5739,
+      "step": 41660
+    },
+    {
+      "epoch": 2.0696334558458327,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0006344332969106984,
+      "loss": 0.5955,
+      "step": 41670
+    },
+    {
+      "epoch": 2.0701301281414524,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0006343935631270488,
+      "loss": 0.5844,
+      "step": 41680
+    },
+    {
+      "epoch": 2.0706268004370716,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0006343538293433992,
+      "loss": 0.5685,
+      "step": 41690
+    },
+    {
+      "epoch": 2.071123472732691,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006343140955597498,
+      "loss": 0.5632,
+      "step": 41700
+    },
+    {
+      "epoch": 2.0716201450283105,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006342743617761002,
+      "loss": 0.5264,
+      "step": 41710
+    },
+    {
+      "epoch": 2.0721168173239297,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0006342346279924506,
+      "loss": 0.5794,
+      "step": 41720
+    },
+    {
+      "epoch": 2.072613489619549,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006341948942088011,
+      "loss": 0.579,
+      "step": 41730
+    },
+    {
+      "epoch": 2.073110161915168,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006341551604251515,
+      "loss": 0.6261,
+      "step": 41740
+    },
+    {
+      "epoch": 2.073606834210788,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.000634115426641502,
+      "loss": 0.5718,
+      "step": 41750
+    },
+    {
+      "epoch": 2.074103506506407,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0006340756928578525,
+      "loss": 0.5665,
+      "step": 41760
+    },
+    {
+      "epoch": 2.0746001788020263,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006340359590742028,
+      "loss": 0.5789,
+      "step": 41770
+    },
+    {
+      "epoch": 2.075096851097646,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006339962252905533,
+      "loss": 0.5854,
+      "step": 41780
+    },
+    {
+      "epoch": 2.075593523393265,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0006339564915069038,
+      "loss": 0.5833,
+      "step": 41790
+    },
+    {
+      "epoch": 2.0760901956888844,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006339167577232541,
+      "loss": 0.5486,
+      "step": 41800
+    },
+    {
+      "epoch": 2.0765868679845036,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0006338770239396047,
+      "loss": 0.5727,
+      "step": 41810
+    },
+    {
+      "epoch": 2.0770835402801233,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0006338372901559552,
+      "loss": 0.5696,
+      "step": 41820
+    },
+    {
+      "epoch": 2.0775802125757425,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006337975563723056,
+      "loss": 0.5609,
+      "step": 41830
+    },
+    {
+      "epoch": 2.0780768848713618,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.000633757822588656,
+      "loss": 0.5891,
+      "step": 41840
+    },
+    {
+      "epoch": 2.0785735571669814,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006337180888050064,
+      "loss": 0.5951,
+      "step": 41850
+    },
+    {
+      "epoch": 2.0790702294626007,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.000633678355021357,
+      "loss": 0.5704,
+      "step": 41860
+    },
+    {
+      "epoch": 2.07956690175822,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006336386212377074,
+      "loss": 0.5804,
+      "step": 41870
+    },
+    {
+      "epoch": 2.080063574053839,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006335988874540578,
+      "loss": 0.5651,
+      "step": 41880
+    },
+    {
+      "epoch": 2.080560246349459,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0006335591536704083,
+      "loss": 0.5728,
+      "step": 41890
+    },
+    {
+      "epoch": 2.081056918645078,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006335194198867587,
+      "loss": 0.5829,
+      "step": 41900
+    },
+    {
+      "epoch": 2.0815535909406973,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0006334796861031092,
+      "loss": 0.5777,
+      "step": 41910
+    },
+    {
+      "epoch": 2.0820502632363165,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0006334399523194597,
+      "loss": 0.5489,
+      "step": 41920
+    },
+    {
+      "epoch": 2.082546935531936,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0006334002185358101,
+      "loss": 0.5876,
+      "step": 41930
+    },
+    {
+      "epoch": 2.0830436078275554,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006333604847521605,
+      "loss": 0.5451,
+      "step": 41940
+    },
+    {
+      "epoch": 2.0835402801231746,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.000633320750968511,
+      "loss": 0.6034,
+      "step": 41950
+    },
+    {
+      "epoch": 2.0840369524187943,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0006332810171848615,
+      "loss": 0.5812,
+      "step": 41960
+    },
+    {
+      "epoch": 2.0845336247144135,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006332412834012119,
+      "loss": 0.5701,
+      "step": 41970
+    },
+    {
+      "epoch": 2.0850302970100327,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0006332015496175624,
+      "loss": 0.5569,
+      "step": 41980
+    },
+    {
+      "epoch": 2.085526969305652,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006331618158339128,
+      "loss": 0.6005,
+      "step": 41990
+    },
+    {
+      "epoch": 2.0860236416012716,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0006331220820502632,
+      "loss": 0.5837,
+      "step": 42000
+    },
+    {
+      "epoch": 2.086520313896891,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0006330823482666138,
+      "loss": 0.576,
+      "step": 42010
+    },
+    {
+      "epoch": 2.08701698619251,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0006330426144829642,
+      "loss": 0.5952,
+      "step": 42020
+    },
+    {
+      "epoch": 2.0875136584881293,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0006330028806993146,
+      "loss": 0.6144,
+      "step": 42030
+    },
+    {
+      "epoch": 2.088010330783749,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.000632963146915665,
+      "loss": 0.5758,
+      "step": 42040
+    },
+    {
+      "epoch": 2.088507003079368,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006329234131320155,
+      "loss": 0.5794,
+      "step": 42050
+    },
+    {
+      "epoch": 2.0890036753749874,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.000632883679348366,
+      "loss": 0.5563,
+      "step": 42060
+    },
+    {
+      "epoch": 2.089500347670607,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006328439455647164,
+      "loss": 0.5668,
+      "step": 42070
+    },
+    {
+      "epoch": 2.0899970199662263,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006328042117810669,
+      "loss": 0.5914,
+      "step": 42080
+    },
+    {
+      "epoch": 2.0904936922618456,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006327644779974173,
+      "loss": 0.5852,
+      "step": 42090
+    },
+    {
+      "epoch": 2.090990364557465,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006327247442137677,
+      "loss": 0.559,
+      "step": 42100
+    },
+    {
+      "epoch": 2.0914870368530845,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0006326850104301183,
+      "loss": 0.5842,
+      "step": 42110
+    },
+    {
+      "epoch": 2.0919837091487037,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006326452766464687,
+      "loss": 0.5544,
+      "step": 42120
+    },
+    {
+      "epoch": 2.092480381444323,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0006326055428628191,
+      "loss": 0.5769,
+      "step": 42130
+    },
+    {
+      "epoch": 2.0929770537399426,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006325658090791696,
+      "loss": 0.5735,
+      "step": 42140
+    },
+    {
+      "epoch": 2.093473726035562,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00063252607529552,
+      "loss": 0.5723,
+      "step": 42150
+    },
+    {
+      "epoch": 2.093970398331181,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006324863415118705,
+      "loss": 0.5741,
+      "step": 42160
+    },
+    {
+      "epoch": 2.0944670706268003,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.000632446607728221,
+      "loss": 0.5562,
+      "step": 42170
+    },
+    {
+      "epoch": 2.09496374292242,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006324068739445714,
+      "loss": 0.5773,
+      "step": 42180
+    },
+    {
+      "epoch": 2.095460415218039,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006323671401609218,
+      "loss": 0.587,
+      "step": 42190
+    },
+    {
+      "epoch": 2.0959570875136584,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0006323274063772723,
+      "loss": 0.5982,
+      "step": 42200
+    },
+    {
+      "epoch": 2.096453759809278,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0006322876725936228,
+      "loss": 0.5631,
+      "step": 42210
+    },
+    {
+      "epoch": 2.0969504321048973,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006322479388099732,
+      "loss": 0.5632,
+      "step": 42220
+    },
+    {
+      "epoch": 2.0974471044005165,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006322082050263236,
+      "loss": 0.5785,
+      "step": 42230
+    },
+    {
+      "epoch": 2.0979437766961357,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006321684712426741,
+      "loss": 0.575,
+      "step": 42240
+    },
+    {
+      "epoch": 2.0984404489917554,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0006321287374590245,
+      "loss": 0.5626,
+      "step": 42250
+    },
+    {
+      "epoch": 2.0989371212873746,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.000632089003675375,
+      "loss": 0.6085,
+      "step": 42260
+    },
+    {
+      "epoch": 2.099433793582994,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0006320492698917255,
+      "loss": 0.5885,
+      "step": 42270
+    },
+    {
+      "epoch": 2.099930465878613,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.000632009536108076,
+      "loss": 0.5616,
+      "step": 42280
+    },
+    {
+      "epoch": 2.1004271381742328,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0006319698023244263,
+      "loss": 0.5764,
+      "step": 42290
+    },
+    {
+      "epoch": 2.100923810469852,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006319300685407768,
+      "loss": 0.5754,
+      "step": 42300
+    },
+    {
+      "epoch": 2.101420482765471,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006318903347571273,
+      "loss": 0.5673,
+      "step": 42310
+    },
+    {
+      "epoch": 2.101917155061091,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0006318506009734777,
+      "loss": 0.5716,
+      "step": 42320
+    },
+    {
+      "epoch": 2.10241382735671,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0006318108671898282,
+      "loss": 0.5877,
+      "step": 42330
+    },
+    {
+      "epoch": 2.1029104996523293,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006317711334061786,
+      "loss": 0.5538,
+      "step": 42340
+    },
+    {
+      "epoch": 2.1034071719479486,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.000631731399622529,
+      "loss": 0.5831,
+      "step": 42350
+    },
+    {
+      "epoch": 2.1039038442435682,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006316916658388796,
+      "loss": 0.5802,
+      "step": 42360
+    },
+    {
+      "epoch": 2.1044005165391875,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00063165193205523,
+      "loss": 0.5466,
+      "step": 42370
+    },
+    {
+      "epoch": 2.1048971888348067,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0006316121982715804,
+      "loss": 0.5852,
+      "step": 42380
+    },
+    {
+      "epoch": 2.105393861130426,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0006315724644879309,
+      "loss": 0.5649,
+      "step": 42390
+    },
+    {
+      "epoch": 2.1058905334260456,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006315327307042813,
+      "loss": 0.608,
+      "step": 42400
+    },
+    {
+      "epoch": 2.106387205721665,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006314929969206319,
+      "loss": 0.6046,
+      "step": 42410
+    },
+    {
+      "epoch": 2.106883878017284,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006314532631369823,
+      "loss": 0.5915,
+      "step": 42420
+    },
+    {
+      "epoch": 2.1073805503129037,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006314135293533327,
+      "loss": 0.579,
+      "step": 42430
+    },
+    {
+      "epoch": 2.107877222608523,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0006313737955696832,
+      "loss": 0.5746,
+      "step": 42440
+    },
+    {
+      "epoch": 2.108373894904142,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006313340617860335,
+      "loss": 0.5734,
+      "step": 42450
+    },
+    {
+      "epoch": 2.1088705671997614,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006312943280023841,
+      "loss": 0.5612,
+      "step": 42460
+    },
+    {
+      "epoch": 2.109367239495381,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0006312545942187346,
+      "loss": 0.5787,
+      "step": 42470
+    },
+    {
+      "epoch": 2.1098639117910003,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0006312148604350849,
+      "loss": 0.5601,
+      "step": 42480
+    },
+    {
+      "epoch": 2.1103605840866195,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006311751266514354,
+      "loss": 0.5422,
+      "step": 42490
+    },
+    {
+      "epoch": 2.110857256382239,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0006311353928677858,
+      "loss": 0.592,
+      "step": 42500
+    },
+    {
+      "epoch": 2.1113539286778584,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006310956590841363,
+      "loss": 0.5701,
+      "step": 42510
+    },
+    {
+      "epoch": 2.1118506009734777,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006310559253004868,
+      "loss": 0.5951,
+      "step": 42520
+    },
+    {
+      "epoch": 2.112347273269097,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006310161915168372,
+      "loss": 0.5869,
+      "step": 42530
+    },
+    {
+      "epoch": 2.1128439455647166,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006309764577331876,
+      "loss": 0.579,
+      "step": 42540
+    },
+    {
+      "epoch": 2.113340617860336,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0006309367239495381,
+      "loss": 0.5473,
+      "step": 42550
+    },
+    {
+      "epoch": 2.113837290155955,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006308969901658886,
+      "loss": 0.6045,
+      "step": 42560
+    },
+    {
+      "epoch": 2.1143339624515747,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006308572563822391,
+      "loss": 0.5556,
+      "step": 42570
+    },
+    {
+      "epoch": 2.114830634747194,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0006308175225985895,
+      "loss": 0.6007,
+      "step": 42580
+    },
+    {
+      "epoch": 2.115327307042813,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006307777888149399,
+      "loss": 0.533,
+      "step": 42590
+    },
+    {
+      "epoch": 2.1158239793384324,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006307380550312904,
+      "loss": 0.548,
+      "step": 42600
+    },
+    {
+      "epoch": 2.116320651634052,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006306983212476409,
+      "loss": 0.562,
+      "step": 42610
+    },
+    {
+      "epoch": 2.1168173239296713,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0006306585874639913,
+      "loss": 0.5785,
+      "step": 42620
+    },
+    {
+      "epoch": 2.1173139962252905,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0006306188536803418,
+      "loss": 0.5775,
+      "step": 42630
+    },
+    {
+      "epoch": 2.1178106685209097,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0006305791198966921,
+      "loss": 0.5739,
+      "step": 42640
+    },
+    {
+      "epoch": 2.1183073408165294,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0006305393861130426,
+      "loss": 0.5826,
+      "step": 42650
+    },
+    {
+      "epoch": 2.1188040131121486,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0006304996523293932,
+      "loss": 0.5619,
+      "step": 42660
+    },
+    {
+      "epoch": 2.119300685407768,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006304599185457435,
+      "loss": 0.567,
+      "step": 42670
+    },
+    {
+      "epoch": 2.1197973577033875,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.000630420184762094,
+      "loss": 0.5476,
+      "step": 42680
+    },
+    {
+      "epoch": 2.1202940299990067,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006303804509784445,
+      "loss": 0.5572,
+      "step": 42690
+    },
+    {
+      "epoch": 2.120790702294626,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006303407171947948,
+      "loss": 0.5884,
+      "step": 42700
+    },
+    {
+      "epoch": 2.121287374590245,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0006303009834111454,
+      "loss": 0.591,
+      "step": 42710
+    },
+    {
+      "epoch": 2.121784046885865,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0006302612496274958,
+      "loss": 0.543,
+      "step": 42720
+    },
+    {
+      "epoch": 2.122280719181484,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006302215158438463,
+      "loss": 0.5653,
+      "step": 42730
+    },
+    {
+      "epoch": 2.1227773914771033,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0006301817820601967,
+      "loss": 0.6003,
+      "step": 42740
+    },
+    {
+      "epoch": 2.1232740637727225,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0006301420482765471,
+      "loss": 0.58,
+      "step": 42750
+    },
+    {
+      "epoch": 2.123770736068342,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0006301023144928977,
+      "loss": 0.5922,
+      "step": 42760
+    },
+    {
+      "epoch": 2.1242674083639614,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006300625807092481,
+      "loss": 0.6117,
+      "step": 42770
+    },
+    {
+      "epoch": 2.1247640806595807,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006300228469255985,
+      "loss": 0.576,
+      "step": 42780
+    },
+    {
+      "epoch": 2.1252607529552003,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.000629983113141949,
+      "loss": 0.5785,
+      "step": 42790
+    },
+    {
+      "epoch": 2.1257574252508196,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0006299433793582994,
+      "loss": 0.5766,
+      "step": 42800
+    },
+    {
+      "epoch": 2.126254097546439,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006299036455746499,
+      "loss": 0.5651,
+      "step": 42810
+    },
+    {
+      "epoch": 2.126750769842058,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006298639117910004,
+      "loss": 0.5559,
+      "step": 42820
+    },
+    {
+      "epoch": 2.1272474421376777,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006298241780073507,
+      "loss": 0.5722,
+      "step": 42830
+    },
+    {
+      "epoch": 2.127744114433297,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0006297844442237012,
+      "loss": 0.5456,
+      "step": 42840
+    },
+    {
+      "epoch": 2.128240786728916,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006297447104400517,
+      "loss": 0.5748,
+      "step": 42850
+    },
+    {
+      "epoch": 2.128737459024536,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0006297049766564022,
+      "loss": 0.6068,
+      "step": 42860
+    },
+    {
+      "epoch": 2.129234131320155,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006296652428727526,
+      "loss": 0.5902,
+      "step": 42870
+    },
+    {
+      "epoch": 2.1297308036157743,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006296255090891031,
+      "loss": 0.5661,
+      "step": 42880
+    },
+    {
+      "epoch": 2.1302274759113935,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0006295857753054535,
+      "loss": 0.5611,
+      "step": 42890
+    },
+    {
+      "epoch": 2.130724148207013,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0006295460415218039,
+      "loss": 0.5864,
+      "step": 42900
+    },
+    {
+      "epoch": 2.1312208205026324,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0006295063077381543,
+      "loss": 0.5897,
+      "step": 42910
+    },
+    {
+      "epoch": 2.1317174927982516,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006294665739545049,
+      "loss": 0.5606,
+      "step": 42920
+    },
+    {
+      "epoch": 2.1322141650938713,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0006294268401708553,
+      "loss": 0.5695,
+      "step": 42930
+    },
+    {
+      "epoch": 2.1327108373894905,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006293871063872057,
+      "loss": 0.5523,
+      "step": 42940
+    },
+    {
+      "epoch": 2.1332075096851097,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0006293473726035562,
+      "loss": 0.5766,
+      "step": 42950
+    },
+    {
+      "epoch": 2.133704181980729,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006293076388199067,
+      "loss": 0.5649,
+      "step": 42960
+    },
+    {
+      "epoch": 2.1342008542763486,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006292679050362571,
+      "loss": 0.5848,
+      "step": 42970
+    },
+    {
+      "epoch": 2.134697526571968,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006292281712526076,
+      "loss": 0.5745,
+      "step": 42980
+    },
+    {
+      "epoch": 2.135194198867587,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.000629188437468958,
+      "loss": 0.5442,
+      "step": 42990
+    },
+    {
+      "epoch": 2.1356908711632063,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006291487036853084,
+      "loss": 0.5622,
+      "step": 43000
+    },
+    {
+      "epoch": 2.136187543458826,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000629108969901659,
+      "loss": 0.5485,
+      "step": 43010
+    },
+    {
+      "epoch": 2.1366842157544452,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006290692361180094,
+      "loss": 0.5648,
+      "step": 43020
+    },
+    {
+      "epoch": 2.1371808880500645,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006290295023343598,
+      "loss": 0.5578,
+      "step": 43030
+    },
+    {
+      "epoch": 2.137677560345684,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006289897685507103,
+      "loss": 0.5757,
+      "step": 43040
+    },
+    {
+      "epoch": 2.1381742326413034,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006289500347670607,
+      "loss": 0.5711,
+      "step": 43050
+    },
+    {
+      "epoch": 2.1386709049369226,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0006289103009834111,
+      "loss": 0.5715,
+      "step": 43060
+    },
+    {
+      "epoch": 2.139167577232542,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0006288705671997617,
+      "loss": 0.5418,
+      "step": 43070
+    },
+    {
+      "epoch": 2.1396642495281615,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0006288308334161121,
+      "loss": 0.5691,
+      "step": 43080
+    },
+    {
+      "epoch": 2.1401609218237807,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0006287910996324625,
+      "loss": 0.5663,
+      "step": 43090
+    },
+    {
+      "epoch": 2.1406575941194,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006287513658488129,
+      "loss": 0.5602,
+      "step": 43100
+    },
+    {
+      "epoch": 2.141154266415019,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0006287116320651635,
+      "loss": 0.5615,
+      "step": 43110
+    },
+    {
+      "epoch": 2.141650938710639,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006286718982815139,
+      "loss": 0.5414,
+      "step": 43120
+    },
+    {
+      "epoch": 2.142147611006258,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0006286321644978643,
+      "loss": 0.5916,
+      "step": 43130
+    },
+    {
+      "epoch": 2.1426442833018773,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0006285924307142148,
+      "loss": 0.5707,
+      "step": 43140
+    },
+    {
+      "epoch": 2.143140955597497,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006285526969305652,
+      "loss": 0.5878,
+      "step": 43150
+    },
+    {
+      "epoch": 2.143637627893116,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006285129631469156,
+      "loss": 0.5597,
+      "step": 43160
+    },
+    {
+      "epoch": 2.1441343001887354,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006284732293632662,
+      "loss": 0.5935,
+      "step": 43170
+    },
+    {
+      "epoch": 2.1446309724843546,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006284334955796166,
+      "loss": 0.5761,
+      "step": 43180
+    },
+    {
+      "epoch": 2.1451276447799743,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.000628393761795967,
+      "loss": 0.573,
+      "step": 43190
+    },
+    {
+      "epoch": 2.1456243170755935,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0006283540280123175,
+      "loss": 0.6216,
+      "step": 43200
+    },
+    {
+      "epoch": 2.1461209893712128,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.000628314294228668,
+      "loss": 0.5643,
+      "step": 43210
+    },
+    {
+      "epoch": 2.1466176616668324,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006282745604450184,
+      "loss": 0.5654,
+      "step": 43220
+    },
+    {
+      "epoch": 2.1471143339624517,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006282348266613689,
+      "loss": 0.586,
+      "step": 43230
+    },
+    {
+      "epoch": 2.147611006258071,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006281950928777193,
+      "loss": 0.5753,
+      "step": 43240
+    },
+    {
+      "epoch": 2.14810767855369,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006281553590940697,
+      "loss": 0.5611,
+      "step": 43250
+    },
+    {
+      "epoch": 2.14860435084931,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006281156253104203,
+      "loss": 0.5534,
+      "step": 43260
+    },
+    {
+      "epoch": 2.149101023144929,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0006280758915267707,
+      "loss": 0.5512,
+      "step": 43270
+    },
+    {
+      "epoch": 2.1495976954405482,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0006280361577431211,
+      "loss": 0.6057,
+      "step": 43280
+    },
+    {
+      "epoch": 2.150094367736168,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0006279964239594716,
+      "loss": 0.6012,
+      "step": 43290
+    },
+    {
+      "epoch": 2.150591040031787,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.000627956690175822,
+      "loss": 0.5826,
+      "step": 43300
+    },
+    {
+      "epoch": 2.1510877123274064,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006279169563921726,
+      "loss": 0.5674,
+      "step": 43310
+    },
+    {
+      "epoch": 2.1515843846230256,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006278772226085229,
+      "loss": 0.5706,
+      "step": 43320
+    },
+    {
+      "epoch": 2.1520810569186453,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0006278374888248734,
+      "loss": 0.5826,
+      "step": 43330
+    },
+    {
+      "epoch": 2.1525777292142645,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0006277977550412239,
+      "loss": 0.56,
+      "step": 43340
+    },
+    {
+      "epoch": 2.1530744015098837,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006277580212575742,
+      "loss": 0.6019,
+      "step": 43350
+    },
+    {
+      "epoch": 2.153571073805503,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006277182874739247,
+      "loss": 0.5726,
+      "step": 43360
+    },
+    {
+      "epoch": 2.1540677461011226,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006276785536902752,
+      "loss": 0.5966,
+      "step": 43370
+    },
+    {
+      "epoch": 2.154564418396742,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0006276388199066256,
+      "loss": 0.5368,
+      "step": 43380
+    },
+    {
+      "epoch": 2.155061090692361,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006275990861229761,
+      "loss": 0.5404,
+      "step": 43390
+    },
+    {
+      "epoch": 2.1555577629879803,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006275593523393265,
+      "loss": 0.5567,
+      "step": 43400
+    },
+    {
+      "epoch": 2.1560544352836,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006275196185556769,
+      "loss": 0.5695,
+      "step": 43410
+    },
+    {
+      "epoch": 2.156551107579219,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006274798847720275,
+      "loss": 0.5826,
+      "step": 43420
+    },
+    {
+      "epoch": 2.1570477798748384,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0006274401509883779,
+      "loss": 0.5905,
+      "step": 43430
+    },
+    {
+      "epoch": 2.157544452170458,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006274004172047283,
+      "loss": 0.572,
+      "step": 43440
+    },
+    {
+      "epoch": 2.1580411244660773,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0006273606834210788,
+      "loss": 0.583,
+      "step": 43450
+    },
+    {
+      "epoch": 2.1585377967616965,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0006273209496374292,
+      "loss": 0.5851,
+      "step": 43460
+    },
+    {
+      "epoch": 2.1590344690573158,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0006272812158537798,
+      "loss": 0.5956,
+      "step": 43470
+    },
+    {
+      "epoch": 2.1595311413529354,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006272414820701302,
+      "loss": 0.5583,
+      "step": 43480
+    },
+    {
+      "epoch": 2.1600278136485547,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006272017482864806,
+      "loss": 0.6028,
+      "step": 43490
+    },
+    {
+      "epoch": 2.160524485944174,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0006271620145028311,
+      "loss": 0.5814,
+      "step": 43500
+    },
+    {
+      "epoch": 2.1610211582397936,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0006271222807191814,
+      "loss": 0.553,
+      "step": 43510
+    },
+    {
+      "epoch": 2.161517830535413,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.000627082546935532,
+      "loss": 0.5377,
+      "step": 43520
+    },
+    {
+      "epoch": 2.162014502831032,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0006270428131518825,
+      "loss": 0.5696,
+      "step": 43530
+    },
+    {
+      "epoch": 2.1625111751266513,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006270030793682328,
+      "loss": 0.534,
+      "step": 43540
+    },
+    {
+      "epoch": 2.163007847422271,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006269633455845833,
+      "loss": 0.5761,
+      "step": 43550
+    },
+    {
+      "epoch": 2.16350451971789,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006269236118009338,
+      "loss": 0.5962,
+      "step": 43560
+    },
+    {
+      "epoch": 2.1640011920135094,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006268838780172842,
+      "loss": 0.5698,
+      "step": 43570
+    },
+    {
+      "epoch": 2.164497864309129,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006268441442336347,
+      "loss": 0.5536,
+      "step": 43580
+    },
+    {
+      "epoch": 2.1649945366047483,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006268044104499851,
+      "loss": 0.5618,
+      "step": 43590
+    },
+    {
+      "epoch": 2.1654912089003675,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0006267646766663356,
+      "loss": 0.5572,
+      "step": 43600
+    },
+    {
+      "epoch": 2.1659878811959867,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.000626724942882686,
+      "loss": 0.5328,
+      "step": 43610
+    },
+    {
+      "epoch": 2.1664845534916064,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0006266852090990365,
+      "loss": 0.5961,
+      "step": 43620
+    },
+    {
+      "epoch": 2.1669812257872256,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.000626645475315387,
+      "loss": 0.5593,
+      "step": 43630
+    },
+    {
+      "epoch": 2.167477898082845,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006266057415317374,
+      "loss": 0.5712,
+      "step": 43640
+    },
+    {
+      "epoch": 2.167974570378464,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0006265660077480878,
+      "loss": 0.5823,
+      "step": 43650
+    },
+    {
+      "epoch": 2.1684712426740838,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006265262739644383,
+      "loss": 0.5414,
+      "step": 43660
+    },
+    {
+      "epoch": 2.168967914969703,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006264865401807888,
+      "loss": 0.5812,
+      "step": 43670
+    },
+    {
+      "epoch": 2.169464587265322,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0006264468063971392,
+      "loss": 0.5957,
+      "step": 43680
+    },
+    {
+      "epoch": 2.169961259560942,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006264070726134897,
+      "loss": 0.5713,
+      "step": 43690
+    },
+    {
+      "epoch": 2.170457931856561,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00062636733882984,
+      "loss": 0.5753,
+      "step": 43700
+    },
+    {
+      "epoch": 2.1709546041521803,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006263276050461905,
+      "loss": 0.5447,
+      "step": 43710
+    },
+    {
+      "epoch": 2.1714512764477996,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006262878712625411,
+      "loss": 0.5614,
+      "step": 43720
+    },
+    {
+      "epoch": 2.1719479487434192,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006262481374788914,
+      "loss": 0.5699,
+      "step": 43730
+    },
+    {
+      "epoch": 2.1724446210390385,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006262084036952419,
+      "loss": 0.5488,
+      "step": 43740
+    },
+    {
+      "epoch": 2.1729412933346577,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0006261686699115924,
+      "loss": 0.5825,
+      "step": 43750
+    },
+    {
+      "epoch": 2.173437965630277,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006261289361279428,
+      "loss": 0.5778,
+      "step": 43760
+    },
+    {
+      "epoch": 2.1739346379258966,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0006260892023442933,
+      "loss": 0.5551,
+      "step": 43770
+    },
+    {
+      "epoch": 2.174431310221516,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006260494685606437,
+      "loss": 0.5708,
+      "step": 43780
+    },
+    {
+      "epoch": 2.174927982517135,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0006260097347769942,
+      "loss": 0.5614,
+      "step": 43790
+    },
+    {
+      "epoch": 2.1754246548127547,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006259700009933446,
+      "loss": 0.5647,
+      "step": 43800
+    },
+    {
+      "epoch": 2.175921327108374,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.000625930267209695,
+      "loss": 0.5616,
+      "step": 43810
+    },
+    {
+      "epoch": 2.176417999403993,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006258905334260456,
+      "loss": 0.5607,
+      "step": 43820
+    },
+    {
+      "epoch": 2.1769146716996124,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.000625850799642396,
+      "loss": 0.568,
+      "step": 43830
+    },
+    {
+      "epoch": 2.177411343995232,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006258110658587464,
+      "loss": 0.5598,
+      "step": 43840
+    },
+    {
+      "epoch": 2.1779080162908513,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006257713320750969,
+      "loss": 0.5484,
+      "step": 43850
+    },
+    {
+      "epoch": 2.1784046885864705,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006257315982914473,
+      "loss": 0.572,
+      "step": 43860
+    },
+    {
+      "epoch": 2.17890136088209,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006256918645077978,
+      "loss": 0.5755,
+      "step": 43870
+    },
+    {
+      "epoch": 2.1793980331777094,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006256521307241483,
+      "loss": 0.5544,
+      "step": 43880
+    },
+    {
+      "epoch": 2.1798947054733286,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006256123969404987,
+      "loss": 0.5519,
+      "step": 43890
+    },
+    {
+      "epoch": 2.180391377768948,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006255726631568491,
+      "loss": 0.5935,
+      "step": 43900
+    },
+    {
+      "epoch": 2.1808880500645675,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0006255329293731996,
+      "loss": 0.562,
+      "step": 43910
+    },
+    {
+      "epoch": 2.1813847223601868,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0006254931955895501,
+      "loss": 0.562,
+      "step": 43920
+    },
+    {
+      "epoch": 2.181881394655806,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006254534618059005,
+      "loss": 0.5664,
+      "step": 43930
+    },
+    {
+      "epoch": 2.1823780669514257,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.000625413728022251,
+      "loss": 0.5742,
+      "step": 43940
+    },
+    {
+      "epoch": 2.182874739247045,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006253739942386014,
+      "loss": 0.5476,
+      "step": 43950
+    },
+    {
+      "epoch": 2.183371411542664,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006253342604549518,
+      "loss": 0.6065,
+      "step": 43960
+    },
+    {
+      "epoch": 2.1838680838382833,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0006252945266713023,
+      "loss": 0.5835,
+      "step": 43970
+    },
+    {
+      "epoch": 2.184364756133903,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0006252547928876528,
+      "loss": 0.5412,
+      "step": 43980
+    },
+    {
+      "epoch": 2.1848614284295222,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0006252150591040032,
+      "loss": 0.5947,
+      "step": 43990
+    },
+    {
+      "epoch": 2.1853581007251415,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006251753253203536,
+      "loss": 0.57,
+      "step": 44000
+    },
+    {
+      "epoch": 2.1858547730207607,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0006251355915367041,
+      "loss": 0.5734,
+      "step": 44010
+    },
+    {
+      "epoch": 2.1863514453163804,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006250958577530546,
+      "loss": 0.5586,
+      "step": 44020
+    },
+    {
+      "epoch": 2.1868481176119996,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.000625056123969405,
+      "loss": 0.5857,
+      "step": 44030
+    },
+    {
+      "epoch": 2.187344789907619,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0006250163901857555,
+      "loss": 0.5902,
+      "step": 44040
+    },
+    {
+      "epoch": 2.1878414622032385,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006249766564021059,
+      "loss": 0.5731,
+      "step": 44050
+    },
+    {
+      "epoch": 2.1883381344988577,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006249369226184563,
+      "loss": 0.5886,
+      "step": 44060
+    },
+    {
+      "epoch": 2.188834806794477,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0006248971888348069,
+      "loss": 0.5806,
+      "step": 44070
+    },
+    {
+      "epoch": 2.189331479090096,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006248574550511573,
+      "loss": 0.5795,
+      "step": 44080
+    },
+    {
+      "epoch": 2.189828151385716,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006248177212675077,
+      "loss": 0.5496,
+      "step": 44090
+    },
+    {
+      "epoch": 2.190324823681335,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006247779874838582,
+      "loss": 0.551,
+      "step": 44100
+    },
+    {
+      "epoch": 2.1908214959769543,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006247382537002086,
+      "loss": 0.5759,
+      "step": 44110
+    },
+    {
+      "epoch": 2.1913181682725735,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.000624698519916559,
+      "loss": 0.5475,
+      "step": 44120
+    },
+    {
+      "epoch": 2.191814840568193,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0006246587861329096,
+      "loss": 0.542,
+      "step": 44130
+    },
+    {
+      "epoch": 2.1923115128638124,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00062461905234926,
+      "loss": 0.5988,
+      "step": 44140
+    },
+    {
+      "epoch": 2.1928081851594317,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0006245793185656104,
+      "loss": 0.583,
+      "step": 44150
+    },
+    {
+      "epoch": 2.1933048574550513,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0006245395847819609,
+      "loss": 0.5564,
+      "step": 44160
+    },
+    {
+      "epoch": 2.1938015297506706,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0006244998509983114,
+      "loss": 0.5518,
+      "step": 44170
+    },
+    {
+      "epoch": 2.19429820204629,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006244601172146618,
+      "loss": 0.523,
+      "step": 44180
+    },
+    {
+      "epoch": 2.194794874341909,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006244203834310122,
+      "loss": 0.5687,
+      "step": 44190
+    },
+    {
+      "epoch": 2.1952915466375287,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006243806496473627,
+      "loss": 0.5617,
+      "step": 44200
+    },
+    {
+      "epoch": 2.195788218933148,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006243409158637132,
+      "loss": 0.5695,
+      "step": 44210
+    },
+    {
+      "epoch": 2.196284891228767,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006243011820800635,
+      "loss": 0.5669,
+      "step": 44220
+    },
+    {
+      "epoch": 2.196781563524387,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006242614482964141,
+      "loss": 0.5852,
+      "step": 44230
+    },
+    {
+      "epoch": 2.197278235820006,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0006242217145127645,
+      "loss": 0.5893,
+      "step": 44240
+    },
+    {
+      "epoch": 2.1977749081156253,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0006241819807291149,
+      "loss": 0.5775,
+      "step": 44250
+    },
+    {
+      "epoch": 2.1982715804112445,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006241422469454654,
+      "loss": 0.5627,
+      "step": 44260
+    },
+    {
+      "epoch": 2.198768252706864,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0006241025131618159,
+      "loss": 0.5642,
+      "step": 44270
+    },
+    {
+      "epoch": 2.1992649250024834,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006240627793781663,
+      "loss": 0.54,
+      "step": 44280
+    },
+    {
+      "epoch": 2.1997615972981026,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006240230455945168,
+      "loss": 0.5426,
+      "step": 44290
+    },
+    {
+      "epoch": 2.2002582695937223,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006239833118108672,
+      "loss": 0.5992,
+      "step": 44300
+    },
+    {
+      "epoch": 2.2007549418893415,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006239435780272176,
+      "loss": 0.5477,
+      "step": 44310
+    },
+    {
+      "epoch": 2.2012516141849607,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0006239038442435682,
+      "loss": 0.5361,
+      "step": 44320
+    },
+    {
+      "epoch": 2.20174828648058,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0006238641104599186,
+      "loss": 0.5513,
+      "step": 44330
+    },
+    {
+      "epoch": 2.2022449587761996,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.000623824376676269,
+      "loss": 0.5877,
+      "step": 44340
+    },
+    {
+      "epoch": 2.202741631071819,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006237846428926195,
+      "loss": 0.5765,
+      "step": 44350
+    },
+    {
+      "epoch": 2.203238303367438,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006237449091089699,
+      "loss": 0.5669,
+      "step": 44360
+    },
+    {
+      "epoch": 2.2037349756630573,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006237051753253205,
+      "loss": 0.5862,
+      "step": 44370
+    },
+    {
+      "epoch": 2.204231647958677,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0006236654415416708,
+      "loss": 0.5514,
+      "step": 44380
+    },
+    {
+      "epoch": 2.204728320254296,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006236257077580213,
+      "loss": 0.572,
+      "step": 44390
+    },
+    {
+      "epoch": 2.2052249925499154,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0006235859739743718,
+      "loss": 0.5707,
+      "step": 44400
+    },
+    {
+      "epoch": 2.205721664845535,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0006235462401907221,
+      "loss": 0.5504,
+      "step": 44410
+    },
+    {
+      "epoch": 2.2062183371411543,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006235065064070727,
+      "loss": 0.5581,
+      "step": 44420
+    },
+    {
+      "epoch": 2.2067150094367736,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006234667726234232,
+      "loss": 0.5712,
+      "step": 44430
+    },
+    {
+      "epoch": 2.207211681732393,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0006234270388397735,
+      "loss": 0.5782,
+      "step": 44440
+    },
+    {
+      "epoch": 2.2077083540280125,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.000623387305056124,
+      "loss": 0.593,
+      "step": 44450
+    },
+    {
+      "epoch": 2.2082050263236317,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0006233475712724744,
+      "loss": 0.5679,
+      "step": 44460
+    },
+    {
+      "epoch": 2.208701698619251,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0006233078374888248,
+      "loss": 0.5535,
+      "step": 44470
+    },
+    {
+      "epoch": 2.20919837091487,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006232681037051754,
+      "loss": 0.5747,
+      "step": 44480
+    },
+    {
+      "epoch": 2.20969504321049,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0006232283699215258,
+      "loss": 0.5539,
+      "step": 44490
+    },
+    {
+      "epoch": 2.210191715506109,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0006231886361378763,
+      "loss": 0.5745,
+      "step": 44500
+    },
+    {
+      "epoch": 2.2106883878017283,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006231489023542267,
+      "loss": 0.5689,
+      "step": 44510
+    },
+    {
+      "epoch": 2.211185060097348,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006231091685705771,
+      "loss": 0.5573,
+      "step": 44520
+    },
+    {
+      "epoch": 2.211681732392967,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0006230694347869277,
+      "loss": 0.5812,
+      "step": 44530
+    },
+    {
+      "epoch": 2.2121784046885864,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006230297010032781,
+      "loss": 0.5662,
+      "step": 44540
+    },
+    {
+      "epoch": 2.2126750769842056,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006229899672196285,
+      "loss": 0.5748,
+      "step": 44550
+    },
+    {
+      "epoch": 2.2131717492798253,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.000622950233435979,
+      "loss": 0.5514,
+      "step": 44560
+    },
+    {
+      "epoch": 2.2136684215754445,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0006229104996523293,
+      "loss": 0.5866,
+      "step": 44570
+    },
+    {
+      "epoch": 2.2141650938710637,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006228707658686799,
+      "loss": 0.598,
+      "step": 44580
+    },
+    {
+      "epoch": 2.2146617661666834,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0006228310320850304,
+      "loss": 0.5792,
+      "step": 44590
+    },
+    {
+      "epoch": 2.2151584384623026,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006227912983013807,
+      "loss": 0.572,
+      "step": 44600
+    },
+    {
+      "epoch": 2.215655110757922,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0006227515645177312,
+      "loss": 0.5688,
+      "step": 44610
+    },
+    {
+      "epoch": 2.216151783053541,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006227118307340818,
+      "loss": 0.5636,
+      "step": 44620
+    },
+    {
+      "epoch": 2.2166484553491608,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006226720969504321,
+      "loss": 0.5957,
+      "step": 44630
+    },
+    {
+      "epoch": 2.21714512764478,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006226323631667826,
+      "loss": 0.5581,
+      "step": 44640
+    },
+    {
+      "epoch": 2.2176417999403992,
+      "grad_norm": 0.125,
+      "learning_rate": 0.000622592629383133,
+      "loss": 0.5876,
+      "step": 44650
+    },
+    {
+      "epoch": 2.218138472236019,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0006225528955994835,
+      "loss": 0.5619,
+      "step": 44660
+    },
+    {
+      "epoch": 2.218635144531638,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.000622513161815834,
+      "loss": 0.567,
+      "step": 44670
+    },
+    {
+      "epoch": 2.2191318168272574,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0006224734280321844,
+      "loss": 0.5615,
+      "step": 44680
+    },
+    {
+      "epoch": 2.2196284891228766,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006224336942485349,
+      "loss": 0.5507,
+      "step": 44690
+    },
+    {
+      "epoch": 2.2201251614184963,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006223939604648853,
+      "loss": 0.5586,
+      "step": 44700
+    },
+    {
+      "epoch": 2.2206218337141155,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006223542266812357,
+      "loss": 0.5518,
+      "step": 44710
+    },
+    {
+      "epoch": 2.2211185060097347,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006223144928975862,
+      "loss": 0.5889,
+      "step": 44720
+    },
+    {
+      "epoch": 2.221615178305354,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006222747591139367,
+      "loss": 0.5527,
+      "step": 44730
+    },
+    {
+      "epoch": 2.2221118506009736,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006222350253302871,
+      "loss": 0.5946,
+      "step": 44740
+    },
+    {
+      "epoch": 2.222608522896593,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006221952915466376,
+      "loss": 0.5737,
+      "step": 44750
+    },
+    {
+      "epoch": 2.223105195192212,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.000622155557762988,
+      "loss": 0.5618,
+      "step": 44760
+    },
+    {
+      "epoch": 2.2236018674878317,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0006221158239793384,
+      "loss": 0.5826,
+      "step": 44770
+    },
+    {
+      "epoch": 2.224098539783451,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.000622076090195689,
+      "loss": 0.5718,
+      "step": 44780
+    },
+    {
+      "epoch": 2.22459521207907,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0006220363564120393,
+      "loss": 0.5569,
+      "step": 44790
+    },
+    {
+      "epoch": 2.2250918843746894,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006219966226283898,
+      "loss": 0.5731,
+      "step": 44800
+    },
+    {
+      "epoch": 2.225588556670309,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006219568888447403,
+      "loss": 0.5907,
+      "step": 44810
+    },
+    {
+      "epoch": 2.2260852289659283,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006219171550610907,
+      "loss": 0.5298,
+      "step": 44820
+    },
+    {
+      "epoch": 2.2265819012615475,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006218774212774412,
+      "loss": 0.5683,
+      "step": 44830
+    },
+    {
+      "epoch": 2.2270785735571668,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006218376874937916,
+      "loss": 0.5793,
+      "step": 44840
+    },
+    {
+      "epoch": 2.2275752458527864,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0006217979537101421,
+      "loss": 0.5538,
+      "step": 44850
+    },
+    {
+      "epoch": 2.2280719181484057,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006217582199264925,
+      "loss": 0.5721,
+      "step": 44860
+    },
+    {
+      "epoch": 2.228568590444025,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0006217184861428429,
+      "loss": 0.5709,
+      "step": 44870
+    },
+    {
+      "epoch": 2.2290652627396446,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0006216787523591935,
+      "loss": 0.5901,
+      "step": 44880
+    },
+    {
+      "epoch": 2.229561935035264,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0006216390185755439,
+      "loss": 0.5396,
+      "step": 44890
+    },
+    {
+      "epoch": 2.230058607330883,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006215992847918943,
+      "loss": 0.5628,
+      "step": 44900
+    },
+    {
+      "epoch": 2.2305552796265022,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006215595510082448,
+      "loss": 0.5887,
+      "step": 44910
+    },
+    {
+      "epoch": 2.231051951922122,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006215198172245952,
+      "loss": 0.5642,
+      "step": 44920
+    },
+    {
+      "epoch": 2.231548624217741,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0006214800834409457,
+      "loss": 0.5751,
+      "step": 44930
+    },
+    {
+      "epoch": 2.2320452965133604,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0006214403496572962,
+      "loss": 0.572,
+      "step": 44940
+    },
+    {
+      "epoch": 2.23254196880898,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006214006158736466,
+      "loss": 0.5656,
+      "step": 44950
+    },
+    {
+      "epoch": 2.2330386411045993,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.000621360882089997,
+      "loss": 0.5807,
+      "step": 44960
+    },
+    {
+      "epoch": 2.2335353134002185,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0006213211483063475,
+      "loss": 0.5496,
+      "step": 44970
+    },
+    {
+      "epoch": 2.2340319856958377,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.000621281414522698,
+      "loss": 0.5722,
+      "step": 44980
+    },
+    {
+      "epoch": 2.2345286579914574,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0006212416807390484,
+      "loss": 0.5721,
+      "step": 44990
+    },
+    {
+      "epoch": 2.2350253302870766,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0006212019469553989,
+      "loss": 0.5698,
+      "step": 45000
+    },
+    {
+      "epoch": 2.235522002582696,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006211622131717493,
+      "loss": 0.5797,
+      "step": 45010
+    },
+    {
+      "epoch": 2.2360186748783155,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006211224793880997,
+      "loss": 0.5671,
+      "step": 45020
+    },
+    {
+      "epoch": 2.2365153471739347,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006210827456044503,
+      "loss": 0.5678,
+      "step": 45030
+    },
+    {
+      "epoch": 2.237012019469554,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0006210430118208007,
+      "loss": 0.5494,
+      "step": 45040
+    },
+    {
+      "epoch": 2.237508691765173,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006210032780371511,
+      "loss": 0.5662,
+      "step": 45050
+    },
+    {
+      "epoch": 2.238005364060793,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0006209635442535015,
+      "loss": 0.5825,
+      "step": 45060
+    },
+    {
+      "epoch": 2.238502036356412,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.000620923810469852,
+      "loss": 0.5997,
+      "step": 45070
+    },
+    {
+      "epoch": 2.2389987086520313,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006208840766862025,
+      "loss": 0.5766,
+      "step": 45080
+    },
+    {
+      "epoch": 2.2394953809476505,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0006208443429025529,
+      "loss": 0.5571,
+      "step": 45090
+    },
+    {
+      "epoch": 2.23999205324327,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0006208046091189034,
+      "loss": 0.5639,
+      "step": 45100
+    },
+    {
+      "epoch": 2.2404887255388894,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006207648753352538,
+      "loss": 0.5667,
+      "step": 45110
+    },
+    {
+      "epoch": 2.2409853978345087,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006207251415516042,
+      "loss": 0.5826,
+      "step": 45120
+    },
+    {
+      "epoch": 2.2414820701301283,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006206854077679548,
+      "loss": 0.5695,
+      "step": 45130
+    },
+    {
+      "epoch": 2.2419787424257476,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0006206456739843052,
+      "loss": 0.5393,
+      "step": 45140
+    },
+    {
+      "epoch": 2.242475414721367,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006206059402006556,
+      "loss": 0.5668,
+      "step": 45150
+    },
+    {
+      "epoch": 2.242972087016986,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006205662064170061,
+      "loss": 0.5789,
+      "step": 45160
+    },
+    {
+      "epoch": 2.2434687593126057,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0006205264726333565,
+      "loss": 0.5663,
+      "step": 45170
+    },
+    {
+      "epoch": 2.243965431608225,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.000620486738849707,
+      "loss": 0.5535,
+      "step": 45180
+    },
+    {
+      "epoch": 2.244462103903844,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0006204470050660575,
+      "loss": 0.6223,
+      "step": 45190
+    },
+    {
+      "epoch": 2.2449587761994634,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006204072712824079,
+      "loss": 0.5743,
+      "step": 45200
+    },
+    {
+      "epoch": 2.245455448495083,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006203675374987583,
+      "loss": 0.57,
+      "step": 45210
+    },
+    {
+      "epoch": 2.2459521207907023,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0006203278037151088,
+      "loss": 0.5516,
+      "step": 45220
+    },
+    {
+      "epoch": 2.2464487930863215,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0006202880699314593,
+      "loss": 0.5767,
+      "step": 45230
+    },
+    {
+      "epoch": 2.246945465381941,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006202483361478097,
+      "loss": 0.558,
+      "step": 45240
+    },
+    {
+      "epoch": 2.2474421376775604,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0006202086023641601,
+      "loss": 0.5542,
+      "step": 45250
+    },
+    {
+      "epoch": 2.2479388099731796,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006201688685805106,
+      "loss": 0.5535,
+      "step": 45260
+    },
+    {
+      "epoch": 2.248435482268799,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0006201291347968611,
+      "loss": 0.5627,
+      "step": 45270
+    },
+    {
+      "epoch": 2.2489321545644185,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0006200894010132115,
+      "loss": 0.5727,
+      "step": 45280
+    },
+    {
+      "epoch": 2.2494288268600378,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.000620049667229562,
+      "loss": 0.5784,
+      "step": 45290
+    },
+    {
+      "epoch": 2.249925499155657,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0006200099334459124,
+      "loss": 0.5787,
+      "step": 45300
+    },
+    {
+      "epoch": 2.2504221714512767,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0006199701996622628,
+      "loss": 0.5643,
+      "step": 45310
+    },
+    {
+      "epoch": 2.250918843746896,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0006199304658786133,
+      "loss": 0.5724,
+      "step": 45320
+    },
+    {
+      "epoch": 2.251415516042515,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006198907320949638,
+      "loss": 0.5998,
+      "step": 45330
+    },
+    {
+      "epoch": 2.2519121883381343,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0006198509983113142,
+      "loss": 0.5486,
+      "step": 45340
+    },
+    {
+      "epoch": 2.252408860633754,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006198112645276647,
+      "loss": 0.559,
+      "step": 45350
+    },
+    {
+      "epoch": 2.2529055329293732,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006197715307440151,
+      "loss": 0.5696,
+      "step": 45360
+    },
+    {
+      "epoch": 2.2534022052249925,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006197317969603655,
+      "loss": 0.5941,
+      "step": 45370
+    },
+    {
+      "epoch": 2.253898877520612,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006196920631767161,
+      "loss": 0.5569,
+      "step": 45380
+    },
+    {
+      "epoch": 2.2543955498162314,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0006196523293930665,
+      "loss": 0.5771,
+      "step": 45390
+    },
+    {
+      "epoch": 2.2548922221118506,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.000619612595609417,
+      "loss": 0.5813,
+      "step": 45400
+    },
+    {
+      "epoch": 2.25538889440747,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0006195728618257674,
+      "loss": 0.5656,
+      "step": 45410
+    },
+    {
+      "epoch": 2.2558855667030895,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0006195331280421178,
+      "loss": 0.5561,
+      "step": 45420
+    },
+    {
+      "epoch": 2.2563822389987087,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006194933942584684,
+      "loss": 0.5836,
+      "step": 45430
+    },
+    {
+      "epoch": 2.256878911294328,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006194536604748187,
+      "loss": 0.5693,
+      "step": 45440
+    },
+    {
+      "epoch": 2.257375583589947,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0006194139266911692,
+      "loss": 0.5708,
+      "step": 45450
+    },
+    {
+      "epoch": 2.257872255885567,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006193741929075197,
+      "loss": 0.5635,
+      "step": 45460
+    },
+    {
+      "epoch": 2.258368928181186,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.00061933445912387,
+      "loss": 0.5647,
+      "step": 45470
+    },
+    {
+      "epoch": 2.2588656004768053,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006192947253402206,
+      "loss": 0.5577,
+      "step": 45480
+    },
+    {
+      "epoch": 2.2593622727724245,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0006192549915565711,
+      "loss": 0.5508,
+      "step": 45490
+    },
+    {
+      "epoch": 2.259858945068044,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006192152577729214,
+      "loss": 0.5782,
+      "step": 45500
+    },
+    {
+      "epoch": 2.2603556173636634,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006191755239892719,
+      "loss": 0.5538,
+      "step": 45510
+    },
+    {
+      "epoch": 2.2608522896592826,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006191357902056223,
+      "loss": 0.5579,
+      "step": 45520
+    },
+    {
+      "epoch": 2.2613489619549023,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0006190960564219727,
+      "loss": 0.5628,
+      "step": 45530
+    },
+    {
+      "epoch": 2.2618456342505215,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006190563226383233,
+      "loss": 0.5516,
+      "step": 45540
+    },
+    {
+      "epoch": 2.2623423065461408,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006190165888546737,
+      "loss": 0.5491,
+      "step": 45550
+    },
+    {
+      "epoch": 2.26283897884176,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006189768550710242,
+      "loss": 0.5538,
+      "step": 45560
+    },
+    {
+      "epoch": 2.2633356511373797,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0006189371212873746,
+      "loss": 0.5748,
+      "step": 45570
+    },
+    {
+      "epoch": 2.263832323432999,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.000618897387503725,
+      "loss": 0.5923,
+      "step": 45580
+    },
+    {
+      "epoch": 2.264328995728618,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006188576537200756,
+      "loss": 0.5539,
+      "step": 45590
+    },
+    {
+      "epoch": 2.264825668024238,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.000618817919936426,
+      "loss": 0.5704,
+      "step": 45600
+    },
+    {
+      "epoch": 2.265322340319857,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0006187781861527764,
+      "loss": 0.561,
+      "step": 45610
+    },
+    {
+      "epoch": 2.2658190126154762,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006187384523691269,
+      "loss": 0.5745,
+      "step": 45620
+    },
+    {
+      "epoch": 2.2663156849110955,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006186987185854772,
+      "loss": 0.5669,
+      "step": 45630
+    },
+    {
+      "epoch": 2.266812357206715,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006186589848018278,
+      "loss": 0.5674,
+      "step": 45640
+    },
+    {
+      "epoch": 2.2673090295023344,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006186192510181783,
+      "loss": 0.5598,
+      "step": 45650
+    },
+    {
+      "epoch": 2.2678057017979536,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006185795172345286,
+      "loss": 0.5784,
+      "step": 45660
+    },
+    {
+      "epoch": 2.2683023740935733,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0006185397834508791,
+      "loss": 0.6022,
+      "step": 45670
+    },
+    {
+      "epoch": 2.2687990463891925,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006185000496672297,
+      "loss": 0.567,
+      "step": 45680
+    },
+    {
+      "epoch": 2.2692957186848117,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00061846031588358,
+      "loss": 0.5606,
+      "step": 45690
+    },
+    {
+      "epoch": 2.269792390980431,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0006184205820999305,
+      "loss": 0.5446,
+      "step": 45700
+    },
+    {
+      "epoch": 2.2702890632760506,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0006183808483162809,
+      "loss": 0.5699,
+      "step": 45710
+    },
+    {
+      "epoch": 2.27078573557167,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006183411145326314,
+      "loss": 0.5412,
+      "step": 45720
+    },
+    {
+      "epoch": 2.271282407867289,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0006183013807489819,
+      "loss": 0.5655,
+      "step": 45730
+    },
+    {
+      "epoch": 2.2717790801629087,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006182616469653323,
+      "loss": 0.5383,
+      "step": 45740
+    },
+    {
+      "epoch": 2.272275752458528,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0006182219131816828,
+      "loss": 0.565,
+      "step": 45750
+    },
+    {
+      "epoch": 2.272772424754147,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006181821793980332,
+      "loss": 0.5611,
+      "step": 45760
+    },
+    {
+      "epoch": 2.2732690970497664,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0006181424456143836,
+      "loss": 0.5672,
+      "step": 45770
+    },
+    {
+      "epoch": 2.273765769345386,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006181027118307342,
+      "loss": 0.6097,
+      "step": 45780
+    },
+    {
+      "epoch": 2.2742624416410053,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006180629780470846,
+      "loss": 0.5667,
+      "step": 45790
+    },
+    {
+      "epoch": 2.2747591139366246,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000618023244263435,
+      "loss": 0.5503,
+      "step": 45800
+    },
+    {
+      "epoch": 2.275255786232244,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006179835104797855,
+      "loss": 0.5652,
+      "step": 45810
+    },
+    {
+      "epoch": 2.2757524585278635,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006179437766961359,
+      "loss": 0.566,
+      "step": 45820
+    },
+    {
+      "epoch": 2.2762491308234827,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006179040429124863,
+      "loss": 0.5466,
+      "step": 45830
+    },
+    {
+      "epoch": 2.276745803119102,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006178643091288369,
+      "loss": 0.5661,
+      "step": 45840
+    },
+    {
+      "epoch": 2.277242475414721,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0006178245753451873,
+      "loss": 0.548,
+      "step": 45850
+    },
+    {
+      "epoch": 2.277739147710341,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006177848415615377,
+      "loss": 0.5579,
+      "step": 45860
+    },
+    {
+      "epoch": 2.27823582000596,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0006177451077778882,
+      "loss": 0.566,
+      "step": 45870
+    },
+    {
+      "epoch": 2.2787324923015793,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006177053739942387,
+      "loss": 0.5508,
+      "step": 45880
+    },
+    {
+      "epoch": 2.279229164597199,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006176656402105891,
+      "loss": 0.5312,
+      "step": 45890
+    },
+    {
+      "epoch": 2.279725836892818,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0006176259064269395,
+      "loss": 0.5519,
+      "step": 45900
+    },
+    {
+      "epoch": 2.2802225091884374,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00061758617264329,
+      "loss": 0.5566,
+      "step": 45910
+    },
+    {
+      "epoch": 2.2807191814840566,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0006175464388596404,
+      "loss": 0.5298,
+      "step": 45920
+    },
+    {
+      "epoch": 2.2812158537796763,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0006175067050759908,
+      "loss": 0.5844,
+      "step": 45930
+    },
+    {
+      "epoch": 2.2817125260752955,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006174669712923414,
+      "loss": 0.5758,
+      "step": 45940
+    },
+    {
+      "epoch": 2.2822091983709147,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006174272375086918,
+      "loss": 0.5557,
+      "step": 45950
+    },
+    {
+      "epoch": 2.2827058706665344,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0006173875037250422,
+      "loss": 0.5717,
+      "step": 45960
+    },
+    {
+      "epoch": 2.2832025429621536,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006173477699413927,
+      "loss": 0.5456,
+      "step": 45970
+    },
+    {
+      "epoch": 2.283699215257773,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0006173080361577431,
+      "loss": 0.5535,
+      "step": 45980
+    },
+    {
+      "epoch": 2.284195887553392,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006172683023740936,
+      "loss": 0.575,
+      "step": 45990
+    },
+    {
+      "epoch": 2.2846925598490118,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006172285685904441,
+      "loss": 0.53,
+      "step": 46000
+    },
+    {
+      "epoch": 2.285189232144631,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0006171888348067945,
+      "loss": 0.5599,
+      "step": 46010
+    },
+    {
+      "epoch": 2.28568590444025,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006171491010231449,
+      "loss": 0.5837,
+      "step": 46020
+    },
+    {
+      "epoch": 2.28618257673587,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006171093672394954,
+      "loss": 0.5603,
+      "step": 46030
+    },
+    {
+      "epoch": 2.286679249031489,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0006170696334558459,
+      "loss": 0.5613,
+      "step": 46040
+    },
+    {
+      "epoch": 2.2871759213271083,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0006170298996721963,
+      "loss": 0.5899,
+      "step": 46050
+    },
+    {
+      "epoch": 2.2876725936227276,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0006169901658885468,
+      "loss": 0.6059,
+      "step": 46060
+    },
+    {
+      "epoch": 2.2881692659183472,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006169504321048972,
+      "loss": 0.5521,
+      "step": 46070
+    },
+    {
+      "epoch": 2.2886659382139665,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0006169106983212476,
+      "loss": 0.5426,
+      "step": 46080
+    },
+    {
+      "epoch": 2.2891626105095857,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0006168709645375982,
+      "loss": 0.5622,
+      "step": 46090
+    },
+    {
+      "epoch": 2.2896592828052054,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0006168312307539486,
+      "loss": 0.5883,
+      "step": 46100
+    },
+    {
+      "epoch": 2.2901559551008246,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.000616791496970299,
+      "loss": 0.5705,
+      "step": 46110
+    },
+    {
+      "epoch": 2.290652627396444,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006167517631866494,
+      "loss": 0.5991,
+      "step": 46120
+    },
+    {
+      "epoch": 2.291149299692063,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.000616712029403,
+      "loss": 0.5995,
+      "step": 46130
+    },
+    {
+      "epoch": 2.2916459719876827,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0006166722956193505,
+      "loss": 0.611,
+      "step": 46140
+    },
+    {
+      "epoch": 2.292142644283302,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0006166325618357008,
+      "loss": 0.5577,
+      "step": 46150
+    },
+    {
+      "epoch": 2.292639316578921,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006165928280520513,
+      "loss": 0.5759,
+      "step": 46160
+    },
+    {
+      "epoch": 2.2931359888745404,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006165530942684017,
+      "loss": 0.5706,
+      "step": 46170
+    },
+    {
+      "epoch": 2.29363266117016,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006165133604847521,
+      "loss": 0.5883,
+      "step": 46180
+    },
+    {
+      "epoch": 2.2941293334657793,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0006164736267011027,
+      "loss": 0.5422,
+      "step": 46190
+    },
+    {
+      "epoch": 2.2946260057613985,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0006164338929174531,
+      "loss": 0.5701,
+      "step": 46200
+    },
+    {
+      "epoch": 2.2951226780570178,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0006163941591338035,
+      "loss": 0.5675,
+      "step": 46210
+    },
+    {
+      "epoch": 2.2956193503526374,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000616354425350154,
+      "loss": 0.5479,
+      "step": 46220
+    },
+    {
+      "epoch": 2.2961160226482566,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006163146915665044,
+      "loss": 0.607,
+      "step": 46230
+    },
+    {
+      "epoch": 2.296612694943876,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0006162749577828549,
+      "loss": 0.5768,
+      "step": 46240
+    },
+    {
+      "epoch": 2.2971093672394955,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0006162352239992054,
+      "loss": 0.5206,
+      "step": 46250
+    },
+    {
+      "epoch": 2.2976060395351148,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006161954902155558,
+      "loss": 0.5652,
+      "step": 46260
+    },
+    {
+      "epoch": 2.298102711830734,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006161557564319062,
+      "loss": 0.6084,
+      "step": 46270
+    },
+    {
+      "epoch": 2.2985993841263532,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006161160226482567,
+      "loss": 0.5947,
+      "step": 46280
+    },
+    {
+      "epoch": 2.299096056421973,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006160762888646072,
+      "loss": 0.5682,
+      "step": 46290
+    },
+    {
+      "epoch": 2.299592728717592,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006160365550809577,
+      "loss": 0.5654,
+      "step": 46300
+    },
+    {
+      "epoch": 2.3000894010132114,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.000615996821297308,
+      "loss": 0.6128,
+      "step": 46310
+    },
+    {
+      "epoch": 2.300586073308831,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0006159570875136585,
+      "loss": 0.5572,
+      "step": 46320
+    },
+    {
+      "epoch": 2.3010827456044503,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.000615917353730009,
+      "loss": 0.5714,
+      "step": 46330
+    },
+    {
+      "epoch": 2.3015794179000695,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006158776199463594,
+      "loss": 0.5847,
+      "step": 46340
+    },
+    {
+      "epoch": 2.3020760901956887,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006158378861627099,
+      "loss": 0.5539,
+      "step": 46350
+    },
+    {
+      "epoch": 2.3025727624913084,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0006157981523790604,
+      "loss": 0.5749,
+      "step": 46360
+    },
+    {
+      "epoch": 2.3030694347869276,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006157584185954107,
+      "loss": 0.5744,
+      "step": 46370
+    },
+    {
+      "epoch": 2.303566107082547,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006157186848117612,
+      "loss": 0.5616,
+      "step": 46380
+    },
+    {
+      "epoch": 2.3040627793781665,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006156789510281117,
+      "loss": 0.5513,
+      "step": 46390
+    },
+    {
+      "epoch": 2.3045594516737857,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006156392172444621,
+      "loss": 0.5751,
+      "step": 46400
+    },
+    {
+      "epoch": 2.305056123969405,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0006155994834608126,
+      "loss": 0.5758,
+      "step": 46410
+    },
+    {
+      "epoch": 2.305552796265024,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.000615559749677163,
+      "loss": 0.5471,
+      "step": 46420
+    },
+    {
+      "epoch": 2.306049468560644,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006155200158935134,
+      "loss": 0.5547,
+      "step": 46430
+    },
+    {
+      "epoch": 2.306546140856263,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.000615480282109864,
+      "loss": 0.5718,
+      "step": 46440
+    },
+    {
+      "epoch": 2.3070428131518823,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0006154405483262144,
+      "loss": 0.578,
+      "step": 46450
+    },
+    {
+      "epoch": 2.307539485447502,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0006154008145425649,
+      "loss": 0.5647,
+      "step": 46460
+    },
+    {
+      "epoch": 2.308036157743121,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006153610807589153,
+      "loss": 0.5614,
+      "step": 46470
+    },
+    {
+      "epoch": 2.3085328300387404,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0006153213469752657,
+      "loss": 0.542,
+      "step": 46480
+    },
+    {
+      "epoch": 2.3090295023343597,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0006152816131916163,
+      "loss": 0.5636,
+      "step": 46490
+    },
+    {
+      "epoch": 2.3095261746299793,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0006152418794079666,
+      "loss": 0.5703,
+      "step": 46500
+    },
+    {
+      "epoch": 2.3100228469255986,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0006152021456243171,
+      "loss": 0.5895,
+      "step": 46510
+    },
+    {
+      "epoch": 2.310519519221218,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006151624118406676,
+      "loss": 0.5679,
+      "step": 46520
+    },
+    {
+      "epoch": 2.311016191516837,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006151226780570179,
+      "loss": 0.5894,
+      "step": 46530
+    },
+    {
+      "epoch": 2.3115128638124567,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006150829442733685,
+      "loss": 0.5703,
+      "step": 46540
+    },
+    {
+      "epoch": 2.312009536108076,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000615043210489719,
+      "loss": 0.563,
+      "step": 46550
+    },
+    {
+      "epoch": 2.312506208403695,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0006150034767060693,
+      "loss": 0.5498,
+      "step": 46560
+    },
+    {
+      "epoch": 2.3130028806993144,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0006149637429224198,
+      "loss": 0.5776,
+      "step": 46570
+    },
+    {
+      "epoch": 2.313499552994934,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0006149240091387702,
+      "loss": 0.5543,
+      "step": 46580
+    },
+    {
+      "epoch": 2.3139962252905533,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0006148842753551208,
+      "loss": 0.5573,
+      "step": 46590
+    },
+    {
+      "epoch": 2.3144928975861725,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0006148445415714712,
+      "loss": 0.5547,
+      "step": 46600
+    },
+    {
+      "epoch": 2.314989569881792,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006148048077878216,
+      "loss": 0.5712,
+      "step": 46610
+    },
+    {
+      "epoch": 2.3154862421774114,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0006147650740041721,
+      "loss": 0.5789,
+      "step": 46620
+    },
+    {
+      "epoch": 2.3159829144730306,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0006147253402205225,
+      "loss": 0.5589,
+      "step": 46630
+    },
+    {
+      "epoch": 2.31647958676865,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.000614685606436873,
+      "loss": 0.5565,
+      "step": 46640
+    },
+    {
+      "epoch": 2.3169762590642695,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006146458726532235,
+      "loss": 0.5665,
+      "step": 46650
+    },
+    {
+      "epoch": 2.3174729313598887,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006146061388695739,
+      "loss": 0.5588,
+      "step": 46660
+    },
+    {
+      "epoch": 2.317969603655508,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0006145664050859243,
+      "loss": 0.556,
+      "step": 46670
+    },
+    {
+      "epoch": 2.3184662759511276,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0006145266713022748,
+      "loss": 0.5475,
+      "step": 46680
+    },
+    {
+      "epoch": 2.318962948246747,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0006144869375186253,
+      "loss": 0.5643,
+      "step": 46690
+    },
+    {
+      "epoch": 2.319459620542366,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006144472037349757,
+      "loss": 0.5823,
+      "step": 46700
+    },
+    {
+      "epoch": 2.3199562928379853,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006144074699513262,
+      "loss": 0.5479,
+      "step": 46710
+    },
+    {
+      "epoch": 2.320452965133605,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006143677361676765,
+      "loss": 0.5915,
+      "step": 46720
+    },
+    {
+      "epoch": 2.320949637429224,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.000614328002384027,
+      "loss": 0.5537,
+      "step": 46730
+    },
+    {
+      "epoch": 2.3214463097248434,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0006142882686003776,
+      "loss": 0.5458,
+      "step": 46740
+    },
+    {
+      "epoch": 2.321942982020463,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.000614248534816728,
+      "loss": 0.5543,
+      "step": 46750
+    },
+    {
+      "epoch": 2.3224396543160823,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0006142088010330784,
+      "loss": 0.5659,
+      "step": 46760
+    },
+    {
+      "epoch": 2.3229363266117016,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006141690672494288,
+      "loss": 0.5702,
+      "step": 46770
+    },
+    {
+      "epoch": 2.323432998907321,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006141293334657793,
+      "loss": 0.554,
+      "step": 46780
+    },
+    {
+      "epoch": 2.3239296712029405,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0006140895996821298,
+      "loss": 0.5765,
+      "step": 46790
+    },
+    {
+      "epoch": 2.3244263434985597,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0006140498658984802,
+      "loss": 0.5599,
+      "step": 46800
+    },
+    {
+      "epoch": 2.324923015794179,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0006140101321148307,
+      "loss": 0.5532,
+      "step": 46810
+    },
+    {
+      "epoch": 2.3254196880897986,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0006139703983311811,
+      "loss": 0.5837,
+      "step": 46820
+    },
+    {
+      "epoch": 2.325916360385418,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006139306645475315,
+      "loss": 0.5828,
+      "step": 46830
+    },
+    {
+      "epoch": 2.326413032681037,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0006138909307638821,
+      "loss": 0.5436,
+      "step": 46840
+    },
+    {
+      "epoch": 2.3269097049766563,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006138511969802325,
+      "loss": 0.5453,
+      "step": 46850
+    },
+    {
+      "epoch": 2.3274063772722755,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0006138114631965829,
+      "loss": 0.5653,
+      "step": 46860
+    },
+    {
+      "epoch": 2.327903049567895,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006137717294129334,
+      "loss": 0.5757,
+      "step": 46870
+    },
+    {
+      "epoch": 2.3283997218635144,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0006137319956292838,
+      "loss": 0.5588,
+      "step": 46880
+    },
+    {
+      "epoch": 2.3288963941591336,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006136922618456343,
+      "loss": 0.569,
+      "step": 46890
+    },
+    {
+      "epoch": 2.3293930664547533,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0006136525280619848,
+      "loss": 0.5593,
+      "step": 46900
+    },
+    {
+      "epoch": 2.3298897387503725,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006136127942783352,
+      "loss": 0.5966,
+      "step": 46910
+    },
+    {
+      "epoch": 2.3303864110459918,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0006135730604946856,
+      "loss": 0.5688,
+      "step": 46920
+    },
+    {
+      "epoch": 2.330883083341611,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006135333267110361,
+      "loss": 0.5858,
+      "step": 46930
+    },
+    {
+      "epoch": 2.3313797556372307,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0006134935929273866,
+      "loss": 0.5659,
+      "step": 46940
+    },
+    {
+      "epoch": 2.33187642793285,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.000613453859143737,
+      "loss": 0.5702,
+      "step": 46950
+    },
+    {
+      "epoch": 2.332373100228469,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0006134141253600875,
+      "loss": 0.548,
+      "step": 46960
+    },
+    {
+      "epoch": 2.332869772524089,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0006133743915764379,
+      "loss": 0.5671,
+      "step": 46970
+    },
+    {
+      "epoch": 2.333366444819708,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0006133346577927883,
+      "loss": 0.5613,
+      "step": 46980
+    },
+    {
+      "epoch": 2.3338631171153272,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0006132949240091387,
+      "loss": 0.6136,
+      "step": 46990
+    },
+    {
+      "epoch": 2.3343597894109465,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0006132551902254893,
+      "loss": 0.5489,
+      "step": 47000
+    },
+    {
+      "epoch": 2.334856461706566,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0006132154564418397,
+      "loss": 0.5452,
+      "step": 47010
+    },
+    {
+      "epoch": 2.3353531340021854,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0006131757226581901,
+      "loss": 0.5752,
+      "step": 47020
+    },
+    {
+      "epoch": 2.3358498062978046,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0006131359888745406,
+      "loss": 0.5623,
+      "step": 47030
+    },
+    {
+      "epoch": 2.3363464785934243,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.000613096255090891,
+      "loss": 0.5492,
+      "step": 47040
+    },
+    {
+      "epoch": 2.3368431508890435,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0006130565213072415,
+      "loss": 0.5686,
+      "step": 47050
+    },
+    {
+      "epoch": 2.3373398231846627,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.000613016787523592,
+      "loss": 0.5749,
+      "step": 47060
+    },
+    {
+      "epoch": 2.337836495480282,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0006129770537399424,
+      "loss": 0.6005,
+      "step": 47070
+    },
+    {
+      "epoch": 2.3383331677759016,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006129373199562928,
+      "loss": 0.581,
+      "step": 47080
+    },
+    {
+      "epoch": 2.338829840071521,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006128975861726434,
+      "loss": 0.5312,
+      "step": 47090
+    },
+    {
+      "epoch": 2.33932651236714,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006128578523889938,
+      "loss": 0.5676,
+      "step": 47100
+    },
+    {
+      "epoch": 2.3398231846627597,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0006128181186053442,
+      "loss": 0.5615,
+      "step": 47110
+    },
+    {
+      "epoch": 2.340319856958379,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006127783848216947,
+      "loss": 0.5737,
+      "step": 47120
+    },
+    {
+      "epoch": 2.340816529253998,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006127386510380451,
+      "loss": 0.5802,
+      "step": 47130
+    },
+    {
+      "epoch": 2.3413132015496174,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006126989172543955,
+      "loss": 0.5564,
+      "step": 47140
+    },
+    {
+      "epoch": 2.341809873845237,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0006126591834707461,
+      "loss": 0.563,
+      "step": 47150
+    },
+    {
+      "epoch": 2.3423065461408563,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006126194496870965,
+      "loss": 0.5529,
+      "step": 47160
+    },
+    {
+      "epoch": 2.3428032184364755,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0006125797159034469,
+      "loss": 0.5879,
+      "step": 47170
+    },
+    {
+      "epoch": 2.3432998907320948,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0006125399821197973,
+      "loss": 0.5619,
+      "step": 47180
+    },
+    {
+      "epoch": 2.3437965630277144,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0006125002483361478,
+      "loss": 0.5826,
+      "step": 47190
+    },
+    {
+      "epoch": 2.3442932353233337,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0006124605145524984,
+      "loss": 0.5169,
+      "step": 47200
+    },
+    {
+      "epoch": 2.344789907618953,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0006124207807688487,
+      "loss": 0.5398,
+      "step": 47210
+    },
+    {
+      "epoch": 2.345286579914572,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006123810469851992,
+      "loss": 0.5669,
+      "step": 47220
+    },
+    {
+      "epoch": 2.345783252210192,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0006123413132015497,
+      "loss": 0.5466,
+      "step": 47230
+    },
+    {
+      "epoch": 2.346279924505811,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006123015794179,
+      "loss": 0.5729,
+      "step": 47240
+    },
+    {
+      "epoch": 2.3467765968014302,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006122618456342506,
+      "loss": 0.5874,
+      "step": 47250
+    },
+    {
+      "epoch": 2.34727326909705,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.000612222111850601,
+      "loss": 0.5505,
+      "step": 47260
+    },
+    {
+      "epoch": 2.347769941392669,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0006121823780669514,
+      "loss": 0.5806,
+      "step": 47270
+    },
+    {
+      "epoch": 2.3482666136882884,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0006121426442833019,
+      "loss": 0.558,
+      "step": 47280
+    },
+    {
+      "epoch": 2.3487632859839076,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006121029104996523,
+      "loss": 0.5507,
+      "step": 47290
+    },
+    {
+      "epoch": 2.3492599582795273,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0006120631767160028,
+      "loss": 0.5562,
+      "step": 47300
+    },
+    {
+      "epoch": 2.3497566305751465,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006120234429323533,
+      "loss": 0.5989,
+      "step": 47310
+    },
+    {
+      "epoch": 2.3502533028707657,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0006119837091487037,
+      "loss": 0.577,
+      "step": 47320
+    },
+    {
+      "epoch": 2.3507499751663854,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006119439753650541,
+      "loss": 0.5982,
+      "step": 47330
+    },
+    {
+      "epoch": 2.3512466474620046,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0006119042415814046,
+      "loss": 0.5664,
+      "step": 47340
+    },
+    {
+      "epoch": 2.351743319757624,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0006118645077977551,
+      "loss": 0.557,
+      "step": 47350
+    },
+    {
+      "epoch": 2.352239992053243,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006118247740141056,
+      "loss": 0.5568,
+      "step": 47360
+    },
+    {
+      "epoch": 2.3527366643488627,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0006117850402304559,
+      "loss": 0.5472,
+      "step": 47370
+    },
+    {
+      "epoch": 2.353233336644482,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006117453064468064,
+      "loss": 0.5609,
+      "step": 47380
+    },
+    {
+      "epoch": 2.353730008940101,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.000611705572663157,
+      "loss": 0.5356,
+      "step": 47390
+    },
+    {
+      "epoch": 2.354226681235721,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006116658388795073,
+      "loss": 0.5859,
+      "step": 47400
+    },
+    {
+      "epoch": 2.35472335353134,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0006116261050958578,
+      "loss": 0.5539,
+      "step": 47410
+    },
+    {
+      "epoch": 2.3552200258269593,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0006115863713122083,
+      "loss": 0.5498,
+      "step": 47420
+    },
+    {
+      "epoch": 2.3557166981225786,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006115466375285586,
+      "loss": 0.5688,
+      "step": 47430
+    },
+    {
+      "epoch": 2.3562133704181982,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0006115069037449091,
+      "loss": 0.543,
+      "step": 47440
+    },
+    {
+      "epoch": 2.3567100427138175,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0006114671699612596,
+      "loss": 0.5701,
+      "step": 47450
+    },
+    {
+      "epoch": 2.3572067150094367,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.00061142743617761,
+      "loss": 0.5587,
+      "step": 47460
+    },
+    {
+      "epoch": 2.3577033873050564,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0006113877023939605,
+      "loss": 0.6082,
+      "step": 47470
+    },
+    {
+      "epoch": 2.3582000596006756,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006113479686103109,
+      "loss": 0.5627,
+      "step": 47480
+    },
+    {
+      "epoch": 2.358696731896295,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0006113082348266614,
+      "loss": 0.5644,
+      "step": 47490
+    },
+    {
+      "epoch": 2.359193404191914,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0006112685010430119,
+      "loss": 0.5654,
+      "step": 47500
+    },
+    {
+      "epoch": 2.3596900764875337,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006112287672593623,
+      "loss": 0.5713,
+      "step": 47510
+    },
+    {
+      "epoch": 2.360186748783153,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0006111890334757128,
+      "loss": 0.5781,
+      "step": 47520
+    },
+    {
+      "epoch": 2.360683421078772,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0006111492996920632,
+      "loss": 0.5249,
+      "step": 47530
+    },
+    {
+      "epoch": 2.3611800933743914,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0006111095659084136,
+      "loss": 0.5465,
+      "step": 47540
+    },
+    {
+      "epoch": 2.361676765670011,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006110698321247642,
+      "loss": 0.5634,
+      "step": 47550
+    },
+    {
+      "epoch": 2.3621734379656303,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006110300983411146,
+      "loss": 0.5259,
+      "step": 47560
+    },
+    {
+      "epoch": 2.3626701102612495,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.000610990364557465,
+      "loss": 0.5567,
+      "step": 47570
+    },
+    {
+      "epoch": 2.3631667825568687,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006109506307738155,
+      "loss": 0.5519,
+      "step": 47580
+    },
+    {
+      "epoch": 2.3636634548524884,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0006109108969901658,
+      "loss": 0.5779,
+      "step": 47590
+    },
+    {
+      "epoch": 2.3641601271481076,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006108711632065164,
+      "loss": 0.5892,
+      "step": 47600
+    },
+    {
+      "epoch": 2.364656799443727,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006108314294228669,
+      "loss": 0.5698,
+      "step": 47610
+    },
+    {
+      "epoch": 2.3651534717393465,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0006107916956392172,
+      "loss": 0.5859,
+      "step": 47620
+    },
+    {
+      "epoch": 2.3656501440349658,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006107519618555677,
+      "loss": 0.5683,
+      "step": 47630
+    },
+    {
+      "epoch": 2.366146816330585,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006107122280719181,
+      "loss": 0.5654,
+      "step": 47640
+    },
+    {
+      "epoch": 2.366643488626204,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006106724942882687,
+      "loss": 0.5919,
+      "step": 47650
+    },
+    {
+      "epoch": 2.367140160921824,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006106327605046191,
+      "loss": 0.5796,
+      "step": 47660
+    },
+    {
+      "epoch": 2.367636833217443,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0006105930267209695,
+      "loss": 0.5524,
+      "step": 47670
+    },
+    {
+      "epoch": 2.3681335055130623,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00061055329293732,
+      "loss": 0.5327,
+      "step": 47680
+    },
+    {
+      "epoch": 2.368630177808682,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0006105135591536704,
+      "loss": 0.5474,
+      "step": 47690
+    },
+    {
+      "epoch": 2.3691268501043012,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006104738253700209,
+      "loss": 0.5729,
+      "step": 47700
+    },
+    {
+      "epoch": 2.3696235223999205,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006104340915863714,
+      "loss": 0.5572,
+      "step": 47710
+    },
+    {
+      "epoch": 2.3701201946955397,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006103943578027218,
+      "loss": 0.5617,
+      "step": 47720
+    },
+    {
+      "epoch": 2.3706168669911594,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006103546240190722,
+      "loss": 0.5546,
+      "step": 47730
+    },
+    {
+      "epoch": 2.3711135392867786,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0006103148902354227,
+      "loss": 0.5584,
+      "step": 47740
+    },
+    {
+      "epoch": 2.371610211582398,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0006102751564517732,
+      "loss": 0.5744,
+      "step": 47750
+    },
+    {
+      "epoch": 2.3721068838780175,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0006102354226681236,
+      "loss": 0.5729,
+      "step": 47760
+    },
+    {
+      "epoch": 2.3726035561736367,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006101956888844741,
+      "loss": 0.5538,
+      "step": 47770
+    },
+    {
+      "epoch": 2.373100228469256,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006101559551008244,
+      "loss": 0.5581,
+      "step": 47780
+    },
+    {
+      "epoch": 2.373596900764875,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0006101162213171749,
+      "loss": 0.5644,
+      "step": 47790
+    },
+    {
+      "epoch": 2.374093573060495,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006100764875335255,
+      "loss": 0.5738,
+      "step": 47800
+    },
+    {
+      "epoch": 2.374590245356114,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006100367537498759,
+      "loss": 0.5686,
+      "step": 47810
+    },
+    {
+      "epoch": 2.3750869176517333,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006099970199662263,
+      "loss": 0.5489,
+      "step": 47820
+    },
+    {
+      "epoch": 2.375583589947353,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006099572861825768,
+      "loss": 0.5604,
+      "step": 47830
+    },
+    {
+      "epoch": 2.376080262242972,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0006099175523989272,
+      "loss": 0.5671,
+      "step": 47840
+    },
+    {
+      "epoch": 2.3765769345385914,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0006098778186152777,
+      "loss": 0.5772,
+      "step": 47850
+    },
+    {
+      "epoch": 2.3770736068342107,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0006098380848316281,
+      "loss": 0.592,
+      "step": 47860
+    },
+    {
+      "epoch": 2.3775702791298303,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006097983510479786,
+      "loss": 0.555,
+      "step": 47870
+    },
+    {
+      "epoch": 2.3780669514254495,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.000609758617264329,
+      "loss": 0.5402,
+      "step": 47880
+    },
+    {
+      "epoch": 2.3785636237210688,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0006097188834806794,
+      "loss": 0.5615,
+      "step": 47890
+    },
+    {
+      "epoch": 2.379060296016688,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00060967914969703,
+      "loss": 0.5759,
+      "step": 47900
+    },
+    {
+      "epoch": 2.3795569683123077,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0006096394159133804,
+      "loss": 0.5377,
+      "step": 47910
+    },
+    {
+      "epoch": 2.380053640607927,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0006095996821297308,
+      "loss": 0.5576,
+      "step": 47920
+    },
+    {
+      "epoch": 2.380550312903546,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006095599483460813,
+      "loss": 0.5403,
+      "step": 47930
+    },
+    {
+      "epoch": 2.3810469851991654,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0006095202145624317,
+      "loss": 0.5741,
+      "step": 47940
+    },
+    {
+      "epoch": 2.381543657494785,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006094804807787822,
+      "loss": 0.5711,
+      "step": 47950
+    },
+    {
+      "epoch": 2.3820403297904043,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0006094407469951327,
+      "loss": 0.5971,
+      "step": 47960
+    },
+    {
+      "epoch": 2.3825370020860235,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006094010132114831,
+      "loss": 0.5602,
+      "step": 47970
+    },
+    {
+      "epoch": 2.383033674381643,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006093612794278335,
+      "loss": 0.5781,
+      "step": 47980
+    },
+    {
+      "epoch": 2.3835303466772624,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.000609321545644184,
+      "loss": 0.5557,
+      "step": 47990
+    },
+    {
+      "epoch": 2.3840270189728816,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006092818118605345,
+      "loss": 0.5475,
+      "step": 48000
+    },
+    {
+      "epoch": 2.384523691268501,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006092420780768849,
+      "loss": 0.5776,
+      "step": 48010
+    },
+    {
+      "epoch": 2.3850203635641205,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0006092023442932354,
+      "loss": 0.5688,
+      "step": 48020
+    },
+    {
+      "epoch": 2.3855170358597397,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0006091626105095858,
+      "loss": 0.5748,
+      "step": 48030
+    },
+    {
+      "epoch": 2.386013708155359,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0006091228767259362,
+      "loss": 0.5595,
+      "step": 48040
+    },
+    {
+      "epoch": 2.3865103804509786,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006090831429422867,
+      "loss": 0.55,
+      "step": 48050
+    },
+    {
+      "epoch": 2.387007052746598,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0006090434091586372,
+      "loss": 0.5507,
+      "step": 48060
+    },
+    {
+      "epoch": 2.387503725042217,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0006090036753749876,
+      "loss": 0.5414,
+      "step": 48070
+    },
+    {
+      "epoch": 2.3880003973378363,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.000608963941591338,
+      "loss": 0.5573,
+      "step": 48080
+    },
+    {
+      "epoch": 2.388497069633456,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0006089242078076885,
+      "loss": 0.5717,
+      "step": 48090
+    },
+    {
+      "epoch": 2.388993741929075,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006088844740240391,
+      "loss": 0.5798,
+      "step": 48100
+    },
+    {
+      "epoch": 2.3894904142246944,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006088447402403894,
+      "loss": 0.5437,
+      "step": 48110
+    },
+    {
+      "epoch": 2.389987086520314,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0006088050064567399,
+      "loss": 0.5658,
+      "step": 48120
+    },
+    {
+      "epoch": 2.3904837588159333,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0006087652726730903,
+      "loss": 0.5929,
+      "step": 48130
+    },
+    {
+      "epoch": 2.3909804311115526,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006087255388894407,
+      "loss": 0.5686,
+      "step": 48140
+    },
+    {
+      "epoch": 2.391477103407172,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0006086858051057913,
+      "loss": 0.5695,
+      "step": 48150
+    },
+    {
+      "epoch": 2.3919737757027915,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0006086460713221417,
+      "loss": 0.5974,
+      "step": 48160
+    },
+    {
+      "epoch": 2.3924704479984107,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0006086063375384921,
+      "loss": 0.5581,
+      "step": 48170
+    },
+    {
+      "epoch": 2.39296712029403,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006085666037548426,
+      "loss": 0.5586,
+      "step": 48180
+    },
+    {
+      "epoch": 2.3934637925896496,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.000608526869971193,
+      "loss": 0.5454,
+      "step": 48190
+    },
+    {
+      "epoch": 2.393960464885269,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006084871361875435,
+      "loss": 0.5464,
+      "step": 48200
+    },
+    {
+      "epoch": 2.394457137180888,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.000608447402403894,
+      "loss": 0.5685,
+      "step": 48210
+    },
+    {
+      "epoch": 2.3949538094765073,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0006084076686202444,
+      "loss": 0.545,
+      "step": 48220
+    },
+    {
+      "epoch": 2.395450481772127,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0006083679348365949,
+      "loss": 0.5746,
+      "step": 48230
+    },
+    {
+      "epoch": 2.395947154067746,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006083282010529452,
+      "loss": 0.5444,
+      "step": 48240
+    },
+    {
+      "epoch": 2.3964438263633654,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006082884672692958,
+      "loss": 0.5705,
+      "step": 48250
+    },
+    {
+      "epoch": 2.3969404986589846,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0006082487334856463,
+      "loss": 0.5899,
+      "step": 48260
+    },
+    {
+      "epoch": 2.3974371709546043,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006082089997019966,
+      "loss": 0.546,
+      "step": 48270
+    },
+    {
+      "epoch": 2.3979338432502235,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0006081692659183471,
+      "loss": 0.5731,
+      "step": 48280
+    },
+    {
+      "epoch": 2.3984305155458427,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006081295321346976,
+      "loss": 0.5569,
+      "step": 48290
+    },
+    {
+      "epoch": 2.398927187841462,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.000608089798351048,
+      "loss": 0.5787,
+      "step": 48300
+    },
+    {
+      "epoch": 2.3994238601370816,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0006080500645673985,
+      "loss": 0.565,
+      "step": 48310
+    },
+    {
+      "epoch": 2.399920532432701,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006080103307837489,
+      "loss": 0.5364,
+      "step": 48320
+    },
+    {
+      "epoch": 2.40041720472832,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0006079705970000993,
+      "loss": 0.5913,
+      "step": 48330
+    },
+    {
+      "epoch": 2.4009138770239398,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0006079308632164498,
+      "loss": 0.5611,
+      "step": 48340
+    },
+    {
+      "epoch": 2.401410549319559,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006078911294328003,
+      "loss": 0.5501,
+      "step": 48350
+    },
+    {
+      "epoch": 2.4019072216151782,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006078513956491507,
+      "loss": 0.5486,
+      "step": 48360
+    },
+    {
+      "epoch": 2.4024038939107974,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006078116618655012,
+      "loss": 0.5591,
+      "step": 48370
+    },
+    {
+      "epoch": 2.402900566206417,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006077719280818516,
+      "loss": 0.5502,
+      "step": 48380
+    },
+    {
+      "epoch": 2.4033972385020363,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006077321942982021,
+      "loss": 0.554,
+      "step": 48390
+    },
+    {
+      "epoch": 2.4038939107976556,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0006076924605145526,
+      "loss": 0.608,
+      "step": 48400
+    },
+    {
+      "epoch": 2.4043905830932752,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.000607652726730903,
+      "loss": 0.5643,
+      "step": 48410
+    },
+    {
+      "epoch": 2.4048872553888945,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006076129929472535,
+      "loss": 0.5602,
+      "step": 48420
+    },
+    {
+      "epoch": 2.4053839276845137,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006075732591636039,
+      "loss": 0.5781,
+      "step": 48430
+    },
+    {
+      "epoch": 2.405880599980133,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0006075335253799543,
+      "loss": 0.5666,
+      "step": 48440
+    },
+    {
+      "epoch": 2.4063772722757526,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006074937915963049,
+      "loss": 0.5493,
+      "step": 48450
+    },
+    {
+      "epoch": 2.406873944571372,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0006074540578126552,
+      "loss": 0.5756,
+      "step": 48460
+    },
+    {
+      "epoch": 2.407370616866991,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0006074143240290057,
+      "loss": 0.548,
+      "step": 48470
+    },
+    {
+      "epoch": 2.4078672891626107,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0006073745902453562,
+      "loss": 0.5627,
+      "step": 48480
+    },
+    {
+      "epoch": 2.40836396145823,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006073348564617065,
+      "loss": 0.5822,
+      "step": 48490
+    },
+    {
+      "epoch": 2.408860633753849,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.000607295122678057,
+      "loss": 0.5511,
+      "step": 48500
+    },
+    {
+      "epoch": 2.4093573060494684,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006072553888944075,
+      "loss": 0.5496,
+      "step": 48510
+    },
+    {
+      "epoch": 2.409853978345088,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0006072156551107579,
+      "loss": 0.5663,
+      "step": 48520
+    },
+    {
+      "epoch": 2.4103506506407073,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006071759213271084,
+      "loss": 0.5617,
+      "step": 48530
+    },
+    {
+      "epoch": 2.4108473229363265,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0006071361875434588,
+      "loss": 0.5533,
+      "step": 48540
+    },
+    {
+      "epoch": 2.411343995231946,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0006070964537598094,
+      "loss": 0.5751,
+      "step": 48550
+    },
+    {
+      "epoch": 2.4118406675275654,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0006070567199761598,
+      "loss": 0.5347,
+      "step": 48560
+    },
+    {
+      "epoch": 2.4123373398231847,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006070169861925102,
+      "loss": 0.5762,
+      "step": 48570
+    },
+    {
+      "epoch": 2.412834012118804,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006069772524088607,
+      "loss": 0.5724,
+      "step": 48580
+    },
+    {
+      "epoch": 2.4133306844144236,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0006069375186252111,
+      "loss": 0.5843,
+      "step": 48590
+    },
+    {
+      "epoch": 2.413827356710043,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0006068977848415615,
+      "loss": 0.5724,
+      "step": 48600
+    },
+    {
+      "epoch": 2.414324029005662,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0006068580510579121,
+      "loss": 0.5487,
+      "step": 48610
+    },
+    {
+      "epoch": 2.4148207013012812,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0006068183172742625,
+      "loss": 0.547,
+      "step": 48620
+    },
+    {
+      "epoch": 2.415317373596901,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0006067785834906129,
+      "loss": 0.5484,
+      "step": 48630
+    },
+    {
+      "epoch": 2.41581404589252,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006067388497069634,
+      "loss": 0.565,
+      "step": 48640
+    },
+    {
+      "epoch": 2.4163107181881394,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006066991159233137,
+      "loss": 0.5757,
+      "step": 48650
+    },
+    {
+      "epoch": 2.4168073904837586,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006066593821396643,
+      "loss": 0.5451,
+      "step": 48660
+    },
+    {
+      "epoch": 2.4173040627793783,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006066196483560148,
+      "loss": 0.5837,
+      "step": 48670
+    },
+    {
+      "epoch": 2.4178007350749975,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006065799145723652,
+      "loss": 0.5708,
+      "step": 48680
+    },
+    {
+      "epoch": 2.4182974073706167,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006065401807887156,
+      "loss": 0.565,
+      "step": 48690
+    },
+    {
+      "epoch": 2.4187940796662364,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.000606500447005066,
+      "loss": 0.5489,
+      "step": 48700
+    },
+    {
+      "epoch": 2.4192907519618556,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006064607132214166,
+      "loss": 0.5792,
+      "step": 48710
+    },
+    {
+      "epoch": 2.419787424257475,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.000606420979437767,
+      "loss": 0.5766,
+      "step": 48720
+    },
+    {
+      "epoch": 2.420284096553094,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0006063812456541174,
+      "loss": 0.5633,
+      "step": 48730
+    },
+    {
+      "epoch": 2.4207807688487137,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0006063415118704679,
+      "loss": 0.5718,
+      "step": 48740
+    },
+    {
+      "epoch": 2.421277441144333,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006063017780868183,
+      "loss": 0.573,
+      "step": 48750
+    },
+    {
+      "epoch": 2.421774113439952,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006062620443031688,
+      "loss": 0.5696,
+      "step": 48760
+    },
+    {
+      "epoch": 2.422270785735572,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0006062223105195193,
+      "loss": 0.5642,
+      "step": 48770
+    },
+    {
+      "epoch": 2.422767458031191,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006061825767358697,
+      "loss": 0.5697,
+      "step": 48780
+    },
+    {
+      "epoch": 2.4232641303268103,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0006061428429522201,
+      "loss": 0.5572,
+      "step": 48790
+    },
+    {
+      "epoch": 2.4237608026224295,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006061031091685706,
+      "loss": 0.5505,
+      "step": 48800
+    },
+    {
+      "epoch": 2.424257474918049,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0006060633753849211,
+      "loss": 0.5675,
+      "step": 48810
+    },
+    {
+      "epoch": 2.4247541472136684,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006060236416012715,
+      "loss": 0.5403,
+      "step": 48820
+    },
+    {
+      "epoch": 2.4252508195092877,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.000605983907817622,
+      "loss": 0.5562,
+      "step": 48830
+    },
+    {
+      "epoch": 2.4257474918049073,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006059441740339724,
+      "loss": 0.5669,
+      "step": 48840
+    },
+    {
+      "epoch": 2.4262441641005266,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0006059044402503228,
+      "loss": 0.5864,
+      "step": 48850
+    },
+    {
+      "epoch": 2.426740836396146,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0006058647064666734,
+      "loss": 0.553,
+      "step": 48860
+    },
+    {
+      "epoch": 2.427237508691765,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006058249726830238,
+      "loss": 0.5676,
+      "step": 48870
+    },
+    {
+      "epoch": 2.4277341809873847,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006057852388993742,
+      "loss": 0.543,
+      "step": 48880
+    },
+    {
+      "epoch": 2.428230853283004,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006057455051157247,
+      "loss": 0.5484,
+      "step": 48890
+    },
+    {
+      "epoch": 2.428727525578623,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006057057713320751,
+      "loss": 0.5772,
+      "step": 48900
+    },
+    {
+      "epoch": 2.429224197874243,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0006056660375484256,
+      "loss": 0.5984,
+      "step": 48910
+    },
+    {
+      "epoch": 2.429720870169862,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.000605626303764776,
+      "loss": 0.574,
+      "step": 48920
+    },
+    {
+      "epoch": 2.4302175424654813,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006055865699811265,
+      "loss": 0.5479,
+      "step": 48930
+    },
+    {
+      "epoch": 2.4307142147611005,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006055468361974769,
+      "loss": 0.5385,
+      "step": 48940
+    },
+    {
+      "epoch": 2.4312108870567197,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006055071024138273,
+      "loss": 0.5857,
+      "step": 48950
+    },
+    {
+      "epoch": 2.4317075593523394,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006054673686301779,
+      "loss": 0.5407,
+      "step": 48960
+    },
+    {
+      "epoch": 2.4322042316479586,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0006054276348465283,
+      "loss": 0.577,
+      "step": 48970
+    },
+    {
+      "epoch": 2.432700903943578,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006053879010628787,
+      "loss": 0.5924,
+      "step": 48980
+    },
+    {
+      "epoch": 2.4331975762391975,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0006053481672792292,
+      "loss": 0.5802,
+      "step": 48990
+    },
+    {
+      "epoch": 2.4336942485348168,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0006053084334955796,
+      "loss": 0.5521,
+      "step": 49000
+    },
+    {
+      "epoch": 2.434190920830436,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006052686997119301,
+      "loss": 0.5648,
+      "step": 49010
+    },
+    {
+      "epoch": 2.434687593126055,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006052289659282806,
+      "loss": 0.58,
+      "step": 49020
+    },
+    {
+      "epoch": 2.435184265421675,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.000605189232144631,
+      "loss": 0.5418,
+      "step": 49030
+    },
+    {
+      "epoch": 2.435680937717294,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006051494983609814,
+      "loss": 0.5535,
+      "step": 49040
+    },
+    {
+      "epoch": 2.4361776100129133,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006051097645773319,
+      "loss": 0.5718,
+      "step": 49050
+    },
+    {
+      "epoch": 2.436674282308533,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0006050700307936824,
+      "loss": 0.5783,
+      "step": 49060
+    },
+    {
+      "epoch": 2.4371709546041522,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0006050302970100328,
+      "loss": 0.5711,
+      "step": 49070
+    },
+    {
+      "epoch": 2.4376676268997715,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0006049905632263833,
+      "loss": 0.5727,
+      "step": 49080
+    },
+    {
+      "epoch": 2.4381642991953907,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0006049508294427337,
+      "loss": 0.567,
+      "step": 49090
+    },
+    {
+      "epoch": 2.4386609714910104,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0006049110956590841,
+      "loss": 0.573,
+      "step": 49100
+    },
+    {
+      "epoch": 2.4391576437866296,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0006048713618754346,
+      "loss": 0.5678,
+      "step": 49110
+    },
+    {
+      "epoch": 2.439654316082249,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006048316280917851,
+      "loss": 0.5699,
+      "step": 49120
+    },
+    {
+      "epoch": 2.4401509883778685,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0006047918943081356,
+      "loss": 0.5472,
+      "step": 49130
+    },
+    {
+      "epoch": 2.4406476606734877,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0006047521605244859,
+      "loss": 0.5707,
+      "step": 49140
+    },
+    {
+      "epoch": 2.441144332969107,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0006047124267408364,
+      "loss": 0.5915,
+      "step": 49150
+    },
+    {
+      "epoch": 2.441641005264726,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.000604672692957187,
+      "loss": 0.5795,
+      "step": 49160
+    },
+    {
+      "epoch": 2.442137677560346,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0006046329591735373,
+      "loss": 0.5591,
+      "step": 49170
+    },
+    {
+      "epoch": 2.442634349855965,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006045932253898878,
+      "loss": 0.5705,
+      "step": 49180
+    },
+    {
+      "epoch": 2.4431310221515843,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0006045534916062382,
+      "loss": 0.5731,
+      "step": 49190
+    },
+    {
+      "epoch": 2.443627694447204,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0006045137578225886,
+      "loss": 0.5761,
+      "step": 49200
+    },
+    {
+      "epoch": 2.444124366742823,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006044740240389392,
+      "loss": 0.5612,
+      "step": 49210
+    },
+    {
+      "epoch": 2.4446210390384424,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0006044342902552896,
+      "loss": 0.5573,
+      "step": 49220
+    },
+    {
+      "epoch": 2.4451177113340616,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00060439455647164,
+      "loss": 0.5884,
+      "step": 49230
+    },
+    {
+      "epoch": 2.4456143836296813,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0006043548226879905,
+      "loss": 0.5599,
+      "step": 49240
+    },
+    {
+      "epoch": 2.4461110559253005,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006043150889043409,
+      "loss": 0.5611,
+      "step": 49250
+    },
+    {
+      "epoch": 2.4466077282209198,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0006042753551206914,
+      "loss": 0.5909,
+      "step": 49260
+    },
+    {
+      "epoch": 2.4471044005165394,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006042356213370419,
+      "loss": 0.5518,
+      "step": 49270
+    },
+    {
+      "epoch": 2.4476010728121587,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0006041958875533923,
+      "loss": 0.5707,
+      "step": 49280
+    },
+    {
+      "epoch": 2.448097745107778,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006041561537697428,
+      "loss": 0.5541,
+      "step": 49290
+    },
+    {
+      "epoch": 2.448594417403397,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006041164199860931,
+      "loss": 0.5475,
+      "step": 49300
+    },
+    {
+      "epoch": 2.4490910896990163,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006040766862024437,
+      "loss": 0.5545,
+      "step": 49310
+    },
+    {
+      "epoch": 2.449587761994636,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0006040369524187942,
+      "loss": 0.554,
+      "step": 49320
+    },
+    {
+      "epoch": 2.4500844342902552,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006039972186351445,
+      "loss": 0.553,
+      "step": 49330
+    },
+    {
+      "epoch": 2.4505811065858745,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.000603957484851495,
+      "loss": 0.578,
+      "step": 49340
+    },
+    {
+      "epoch": 2.451077778881494,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006039177510678455,
+      "loss": 0.5772,
+      "step": 49350
+    },
+    {
+      "epoch": 2.4515744511771134,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0006038780172841959,
+      "loss": 0.5496,
+      "step": 49360
+    },
+    {
+      "epoch": 2.4520711234727326,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0006038382835005464,
+      "loss": 0.5625,
+      "step": 49370
+    },
+    {
+      "epoch": 2.452567795768352,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006037985497168968,
+      "loss": 0.5499,
+      "step": 49380
+    },
+    {
+      "epoch": 2.4530644680639715,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006037588159332472,
+      "loss": 0.589,
+      "step": 49390
+    },
+    {
+      "epoch": 2.4535611403595907,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0006037190821495977,
+      "loss": 0.5583,
+      "step": 49400
+    },
+    {
+      "epoch": 2.45405781265521,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006036793483659482,
+      "loss": 0.5581,
+      "step": 49410
+    },
+    {
+      "epoch": 2.4545544849508296,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0006036396145822986,
+      "loss": 0.5714,
+      "step": 49420
+    },
+    {
+      "epoch": 2.455051157246449,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0006035998807986491,
+      "loss": 0.5595,
+      "step": 49430
+    },
+    {
+      "epoch": 2.455547829542068,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0006035601470149995,
+      "loss": 0.5451,
+      "step": 49440
+    },
+    {
+      "epoch": 2.4560445018376873,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00060352041323135,
+      "loss": 0.5666,
+      "step": 49450
+    },
+    {
+      "epoch": 2.456541174133307,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006034806794477005,
+      "loss": 0.6094,
+      "step": 49460
+    },
+    {
+      "epoch": 2.457037846428926,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0006034409456640509,
+      "loss": 0.5374,
+      "step": 49470
+    },
+    {
+      "epoch": 2.4575345187245454,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0006034012118804014,
+      "loss": 0.5574,
+      "step": 49480
+    },
+    {
+      "epoch": 2.458031191020165,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0006033614780967518,
+      "loss": 0.5301,
+      "step": 49490
+    },
+    {
+      "epoch": 2.4585278633157843,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0006033217443131022,
+      "loss": 0.5728,
+      "step": 49500
+    },
+    {
+      "epoch": 2.4590245356114036,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0006032820105294528,
+      "loss": 0.5635,
+      "step": 49510
+    },
+    {
+      "epoch": 2.4595212079070228,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006032422767458031,
+      "loss": 0.5471,
+      "step": 49520
+    },
+    {
+      "epoch": 2.4600178802026424,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0006032025429621536,
+      "loss": 0.5895,
+      "step": 49530
+    },
+    {
+      "epoch": 2.4605145524982617,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006031628091785041,
+      "loss": 0.5626,
+      "step": 49540
+    },
+    {
+      "epoch": 2.461011224793881,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0006031230753948544,
+      "loss": 0.5692,
+      "step": 49550
+    },
+    {
+      "epoch": 2.4615078970895006,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000603083341611205,
+      "loss": 0.5512,
+      "step": 49560
+    },
+    {
+      "epoch": 2.46200456938512,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0006030436078275554,
+      "loss": 0.5326,
+      "step": 49570
+    },
+    {
+      "epoch": 2.462501241680739,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0006030038740439059,
+      "loss": 0.5489,
+      "step": 49580
+    },
+    {
+      "epoch": 2.4629979139763583,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006029641402602563,
+      "loss": 0.5411,
+      "step": 49590
+    },
+    {
+      "epoch": 2.463494586271978,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006029244064766067,
+      "loss": 0.571,
+      "step": 49600
+    },
+    {
+      "epoch": 2.463991258567597,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0006028846726929573,
+      "loss": 0.5845,
+      "step": 49610
+    },
+    {
+      "epoch": 2.4644879308632164,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0006028449389093077,
+      "loss": 0.5634,
+      "step": 49620
+    },
+    {
+      "epoch": 2.4649846031588356,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0006028052051256581,
+      "loss": 0.5268,
+      "step": 49630
+    },
+    {
+      "epoch": 2.4654812754544553,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0006027654713420086,
+      "loss": 0.5525,
+      "step": 49640
+    },
+    {
+      "epoch": 2.4659779477500745,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.000602725737558359,
+      "loss": 0.5734,
+      "step": 49650
+    },
+    {
+      "epoch": 2.4664746200456937,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0006026860037747094,
+      "loss": 0.581,
+      "step": 49660
+    },
+    {
+      "epoch": 2.466971292341313,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.00060264626999106,
+      "loss": 0.5424,
+      "step": 49670
+    },
+    {
+      "epoch": 2.4674679646369326,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006026065362074104,
+      "loss": 0.5855,
+      "step": 49680
+    },
+    {
+      "epoch": 2.467964636932552,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006025668024237608,
+      "loss": 0.5456,
+      "step": 49690
+    },
+    {
+      "epoch": 2.468461309228171,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0006025270686401113,
+      "loss": 0.5672,
+      "step": 49700
+    },
+    {
+      "epoch": 2.4689579815237908,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0006024873348564616,
+      "loss": 0.5702,
+      "step": 49710
+    },
+    {
+      "epoch": 2.46945465381941,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0006024476010728122,
+      "loss": 0.5975,
+      "step": 49720
+    },
+    {
+      "epoch": 2.469951326115029,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0006024078672891627,
+      "loss": 0.5456,
+      "step": 49730
+    },
+    {
+      "epoch": 2.4704479984106484,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0006023681335055131,
+      "loss": 0.5342,
+      "step": 49740
+    },
+    {
+      "epoch": 2.470944670706268,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0006023283997218635,
+      "loss": 0.5671,
+      "step": 49750
+    },
+    {
+      "epoch": 2.4714413430018873,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0006022886659382141,
+      "loss": 0.5535,
+      "step": 49760
+    },
+    {
+      "epoch": 2.4719380152975066,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0006022489321545645,
+      "loss": 0.5873,
+      "step": 49770
+    },
+    {
+      "epoch": 2.4724346875931262,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0006022091983709149,
+      "loss": 0.5657,
+      "step": 49780
+    },
+    {
+      "epoch": 2.4729313598887455,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006021694645872653,
+      "loss": 0.5449,
+      "step": 49790
+    },
+    {
+      "epoch": 2.4734280321843647,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0006021297308036158,
+      "loss": 0.5494,
+      "step": 49800
+    },
+    {
+      "epoch": 2.473924704479984,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006020899970199662,
+      "loss": 0.5704,
+      "step": 49810
+    },
+    {
+      "epoch": 2.4744213767756036,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006020502632363167,
+      "loss": 0.5526,
+      "step": 49820
+    },
+    {
+      "epoch": 2.474918049071223,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0006020105294526672,
+      "loss": 0.5606,
+      "step": 49830
+    },
+    {
+      "epoch": 2.475414721366842,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0006019707956690176,
+      "loss": 0.588,
+      "step": 49840
+    },
+    {
+      "epoch": 2.4759113936624617,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.000601931061885368,
+      "loss": 0.5424,
+      "step": 49850
+    },
+    {
+      "epoch": 2.476408065958081,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0006018913281017186,
+      "loss": 0.5562,
+      "step": 49860
+    },
+    {
+      "epoch": 2.4769047382537,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.000601851594318069,
+      "loss": 0.545,
+      "step": 49870
+    },
+    {
+      "epoch": 2.4774014105493194,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0006018118605344194,
+      "loss": 0.5647,
+      "step": 49880
+    },
+    {
+      "epoch": 2.477898082844939,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0006017721267507699,
+      "loss": 0.5427,
+      "step": 49890
+    },
+    {
+      "epoch": 2.4783947551405583,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0006017323929671203,
+      "loss": 0.5381,
+      "step": 49900
+    },
+    {
+      "epoch": 2.4788914274361775,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0006016926591834707,
+      "loss": 0.555,
+      "step": 49910
+    },
+    {
+      "epoch": 2.479388099731797,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006016529253998213,
+      "loss": 0.5809,
+      "step": 49920
+    },
+    {
+      "epoch": 2.4798847720274164,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0006016131916161717,
+      "loss": 0.5339,
+      "step": 49930
+    },
+    {
+      "epoch": 2.4803814443230356,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006015734578325221,
+      "loss": 0.5717,
+      "step": 49940
+    },
+    {
+      "epoch": 2.480878116618655,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0006015337240488726,
+      "loss": 0.5455,
+      "step": 49950
+    },
+    {
+      "epoch": 2.4813747889142745,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.000601493990265223,
+      "loss": 0.5582,
+      "step": 49960
+    },
+    {
+      "epoch": 2.4818714612098938,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0006014542564815735,
+      "loss": 0.588,
+      "step": 49970
+    },
+    {
+      "epoch": 2.482368133505513,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0006014145226979239,
+      "loss": 0.5391,
+      "step": 49980
+    },
+    {
+      "epoch": 2.4828648058011322,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0006013747889142744,
+      "loss": 0.5665,
+      "step": 49990
+    },
+    {
+      "epoch": 2.483361478096752,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0006013350551306248,
+      "loss": 0.5499,
+      "step": 50000
+    },
+    {
+      "epoch": 2.483858150392371,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0006012953213469752,
+      "loss": 0.5816,
+      "step": 50010
+    },
+    {
+      "epoch": 2.4843548226879903,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006012555875633258,
+      "loss": 0.5779,
+      "step": 50020
+    },
+    {
+      "epoch": 2.4848514949836096,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006012158537796763,
+      "loss": 0.526,
+      "step": 50030
+    },
+    {
+      "epoch": 2.4853481672792292,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0006011761199960266,
+      "loss": 0.5614,
+      "step": 50040
+    },
+    {
+      "epoch": 2.4858448395748485,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0006011363862123771,
+      "loss": 0.5772,
+      "step": 50050
+    },
+    {
+      "epoch": 2.4863415118704677,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006010966524287275,
+      "loss": 0.5888,
+      "step": 50060
+    },
+    {
+      "epoch": 2.4868381841660874,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.000601056918645078,
+      "loss": 0.5385,
+      "step": 50070
+    },
+    {
+      "epoch": 2.4873348564617066,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0006010171848614285,
+      "loss": 0.6059,
+      "step": 50080
+    },
+    {
+      "epoch": 2.487831528757326,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0006009774510777789,
+      "loss": 0.5604,
+      "step": 50090
+    },
+    {
+      "epoch": 2.488328201052945,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0006009377172941293,
+      "loss": 0.5387,
+      "step": 50100
+    },
+    {
+      "epoch": 2.4888248733485647,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0006008979835104798,
+      "loss": 0.5603,
+      "step": 50110
+    },
+    {
+      "epoch": 2.489321545644184,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0006008582497268303,
+      "loss": 0.5659,
+      "step": 50120
+    },
+    {
+      "epoch": 2.489818217939803,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0006008185159431807,
+      "loss": 0.5607,
+      "step": 50130
+    },
+    {
+      "epoch": 2.490314890235423,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0006007787821595312,
+      "loss": 0.5308,
+      "step": 50140
+    },
+    {
+      "epoch": 2.490811562531042,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0006007390483758816,
+      "loss": 0.5595,
+      "step": 50150
+    },
+    {
+      "epoch": 2.4913082348266613,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.000600699314592232,
+      "loss": 0.5742,
+      "step": 50160
+    },
+    {
+      "epoch": 2.4918049071222805,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006006595808085825,
+      "loss": 0.5578,
+      "step": 50170
+    },
+    {
+      "epoch": 2.4923015794179,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.000600619847024933,
+      "loss": 0.5511,
+      "step": 50180
+    },
+    {
+      "epoch": 2.4927982517135194,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0006005801132412835,
+      "loss": 0.5548,
+      "step": 50190
+    },
+    {
+      "epoch": 2.4932949240091387,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0006005403794576338,
+      "loss": 0.5776,
+      "step": 50200
+    },
+    {
+      "epoch": 2.4937915963047583,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0006005006456739843,
+      "loss": 0.5829,
+      "step": 50210
+    },
+    {
+      "epoch": 2.4942882686003776,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0006004609118903349,
+      "loss": 0.5768,
+      "step": 50220
+    },
+    {
+      "epoch": 2.494784940895997,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0006004211781066852,
+      "loss": 0.5617,
+      "step": 50230
+    },
+    {
+      "epoch": 2.495281613191616,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0006003814443230357,
+      "loss": 0.5578,
+      "step": 50240
+    },
+    {
+      "epoch": 2.4957782854872357,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0006003417105393861,
+      "loss": 0.5624,
+      "step": 50250
+    },
+    {
+      "epoch": 2.496274957782855,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0006003019767557365,
+      "loss": 0.5755,
+      "step": 50260
+    },
+    {
+      "epoch": 2.496771630078474,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0006002622429720871,
+      "loss": 0.5617,
+      "step": 50270
+    },
+    {
+      "epoch": 2.497268302374094,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0006002225091884375,
+      "loss": 0.538,
+      "step": 50280
+    },
+    {
+      "epoch": 2.497764974669713,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0006001827754047879,
+      "loss": 0.5541,
+      "step": 50290
+    },
+    {
+      "epoch": 2.4982616469653323,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0006001430416211384,
+      "loss": 0.5606,
+      "step": 50300
+    },
+    {
+      "epoch": 2.4987583192609515,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0006001033078374888,
+      "loss": 0.5533,
+      "step": 50310
+    },
+    {
+      "epoch": 2.499254991556571,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0006000635740538393,
+      "loss": 0.5404,
+      "step": 50320
+    },
+    {
+      "epoch": 2.4997516638521904,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0006000238402701898,
+      "loss": 0.5438,
+      "step": 50330
+    },
+    {
+      "epoch": 2.5002483361478096,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005999841064865402,
+      "loss": 0.556,
+      "step": 50340
+    },
+    {
+      "epoch": 2.5007450084434293,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005999443727028907,
+      "loss": 0.589,
+      "step": 50350
+    },
+    {
+      "epoch": 2.5012416807390485,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005999046389192411,
+      "loss": 0.5908,
+      "step": 50360
+    },
+    {
+      "epoch": 2.5017383530346677,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005998649051355916,
+      "loss": 0.5258,
+      "step": 50370
+    },
+    {
+      "epoch": 2.502235025330287,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005998251713519421,
+      "loss": 0.5685,
+      "step": 50380
+    },
+    {
+      "epoch": 2.502731697625906,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005997854375682924,
+      "loss": 0.5652,
+      "step": 50390
+    },
+    {
+      "epoch": 2.503228369921526,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005997457037846429,
+      "loss": 0.5472,
+      "step": 50400
+    },
+    {
+      "epoch": 2.503725042217145,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005997059700009934,
+      "loss": 0.5512,
+      "step": 50410
+    },
+    {
+      "epoch": 2.5042217145127643,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005996662362173438,
+      "loss": 0.5654,
+      "step": 50420
+    },
+    {
+      "epoch": 2.504718386808384,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005996265024336943,
+      "loss": 0.5999,
+      "step": 50430
+    },
+    {
+      "epoch": 2.505215059104003,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005995867686500447,
+      "loss": 0.555,
+      "step": 50440
+    },
+    {
+      "epoch": 2.5057117313996224,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005995470348663951,
+      "loss": 0.5657,
+      "step": 50450
+    },
+    {
+      "epoch": 2.5062084036952417,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005995073010827456,
+      "loss": 0.5458,
+      "step": 50460
+    },
+    {
+      "epoch": 2.5067050759908613,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005994675672990961,
+      "loss": 0.565,
+      "step": 50470
+    },
+    {
+      "epoch": 2.5072017482864806,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005994278335154466,
+      "loss": 0.566,
+      "step": 50480
+    },
+    {
+      "epoch": 2.5076984205821,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.000599388099731797,
+      "loss": 0.5603,
+      "step": 50490
+    },
+    {
+      "epoch": 2.5081950928777195,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005993483659481474,
+      "loss": 0.544,
+      "step": 50500
+    },
+    {
+      "epoch": 2.5086917651733387,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005993086321644979,
+      "loss": 0.5736,
+      "step": 50510
+    },
+    {
+      "epoch": 2.509188437468958,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005992688983808484,
+      "loss": 0.5321,
+      "step": 50520
+    },
+    {
+      "epoch": 2.509685109764577,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005992291645971988,
+      "loss": 0.5619,
+      "step": 50530
+    },
+    {
+      "epoch": 2.510181782060197,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005991894308135493,
+      "loss": 0.5561,
+      "step": 50540
+    },
+    {
+      "epoch": 2.510678454355816,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005991496970298997,
+      "loss": 0.5537,
+      "step": 50550
+    },
+    {
+      "epoch": 2.5111751266514353,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005991099632462501,
+      "loss": 0.5722,
+      "step": 50560
+    },
+    {
+      "epoch": 2.511671798947055,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005990702294626007,
+      "loss": 0.5708,
+      "step": 50570
+    },
+    {
+      "epoch": 2.512168471242674,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.000599030495678951,
+      "loss": 0.5803,
+      "step": 50580
+    },
+    {
+      "epoch": 2.5126651435382934,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005989907618953015,
+      "loss": 0.5337,
+      "step": 50590
+    },
+    {
+      "epoch": 2.5131618158339126,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.000598951028111652,
+      "loss": 0.5547,
+      "step": 50600
+    },
+    {
+      "epoch": 2.5136584881295323,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005989112943280023,
+      "loss": 0.5613,
+      "step": 50610
+    },
+    {
+      "epoch": 2.5141551604251515,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005988715605443529,
+      "loss": 0.5465,
+      "step": 50620
+    },
+    {
+      "epoch": 2.5146518327207708,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005988318267607034,
+      "loss": 0.5655,
+      "step": 50630
+    },
+    {
+      "epoch": 2.5151485050163904,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005987920929770538,
+      "loss": 0.5754,
+      "step": 50640
+    },
+    {
+      "epoch": 2.5156451773120097,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005987523591934042,
+      "loss": 0.5651,
+      "step": 50650
+    },
+    {
+      "epoch": 2.516141849607629,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005987126254097546,
+      "loss": 0.5524,
+      "step": 50660
+    },
+    {
+      "epoch": 2.516638521903248,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005986728916261052,
+      "loss": 0.5469,
+      "step": 50670
+    },
+    {
+      "epoch": 2.5171351941988673,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005986331578424556,
+      "loss": 0.5544,
+      "step": 50680
+    },
+    {
+      "epoch": 2.517631866494487,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.000598593424058806,
+      "loss": 0.5957,
+      "step": 50690
+    },
+    {
+      "epoch": 2.5181285387901062,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005985536902751565,
+      "loss": 0.5565,
+      "step": 50700
+    },
+    {
+      "epoch": 2.518625211085726,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005985139564915069,
+      "loss": 0.574,
+      "step": 50710
+    },
+    {
+      "epoch": 2.519121883381345,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005984742227078574,
+      "loss": 0.5723,
+      "step": 50720
+    },
+    {
+      "epoch": 2.5196185556769644,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005984344889242079,
+      "loss": 0.5467,
+      "step": 50730
+    },
+    {
+      "epoch": 2.5201152279725836,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005983947551405583,
+      "loss": 0.5421,
+      "step": 50740
+    },
+    {
+      "epoch": 2.520611900268203,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005983550213569087,
+      "loss": 0.5228,
+      "step": 50750
+    },
+    {
+      "epoch": 2.5211085725638225,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005983152875732592,
+      "loss": 0.5707,
+      "step": 50760
+    },
+    {
+      "epoch": 2.5216052448594417,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005982755537896097,
+      "loss": 0.5645,
+      "step": 50770
+    },
+    {
+      "epoch": 2.522101917155061,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005982358200059601,
+      "loss": 0.563,
+      "step": 50780
+    },
+    {
+      "epoch": 2.5225985894506806,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005981960862223106,
+      "loss": 0.5781,
+      "step": 50790
+    },
+    {
+      "epoch": 2.5230952617463,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.000598156352438661,
+      "loss": 0.5269,
+      "step": 50800
+    },
+    {
+      "epoch": 2.523591934041919,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005981166186550114,
+      "loss": 0.55,
+      "step": 50810
+    },
+    {
+      "epoch": 2.5240886063375383,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.000598076884871362,
+      "loss": 0.5676,
+      "step": 50820
+    },
+    {
+      "epoch": 2.524585278633158,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005980371510877124,
+      "loss": 0.5254,
+      "step": 50830
+    },
+    {
+      "epoch": 2.525081950928777,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005979974173040628,
+      "loss": 0.5673,
+      "step": 50840
+    },
+    {
+      "epoch": 2.5255786232243964,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005979576835204132,
+      "loss": 0.5848,
+      "step": 50850
+    },
+    {
+      "epoch": 2.526075295520016,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005979179497367637,
+      "loss": 0.5343,
+      "step": 50860
+    },
+    {
+      "epoch": 2.5265719678156353,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005978782159531142,
+      "loss": 0.5898,
+      "step": 50870
+    },
+    {
+      "epoch": 2.5270686401112545,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005978384821694646,
+      "loss": 0.5488,
+      "step": 50880
+    },
+    {
+      "epoch": 2.5275653124068738,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005977987483858151,
+      "loss": 0.5552,
+      "step": 50890
+    },
+    {
+      "epoch": 2.5280619847024934,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005977590146021655,
+      "loss": 0.5616,
+      "step": 50900
+    },
+    {
+      "epoch": 2.5285586569981127,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005977192808185159,
+      "loss": 0.5752,
+      "step": 50910
+    },
+    {
+      "epoch": 2.529055329293732,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005976795470348665,
+      "loss": 0.5545,
+      "step": 50920
+    },
+    {
+      "epoch": 2.5295520015893516,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005976398132512169,
+      "loss": 0.5562,
+      "step": 50930
+    },
+    {
+      "epoch": 2.530048673884971,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005976000794675673,
+      "loss": 0.56,
+      "step": 50940
+    },
+    {
+      "epoch": 2.53054534618059,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005975603456839178,
+      "loss": 0.5614,
+      "step": 50950
+    },
+    {
+      "epoch": 2.5310420184762092,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005975206119002682,
+      "loss": 0.5647,
+      "step": 50960
+    },
+    {
+      "epoch": 2.5315386907718285,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005974808781166186,
+      "loss": 0.5495,
+      "step": 50970
+    },
+    {
+      "epoch": 2.532035363067448,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005974411443329692,
+      "loss": 0.5878,
+      "step": 50980
+    },
+    {
+      "epoch": 2.5325320353630674,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005974014105493196,
+      "loss": 0.5465,
+      "step": 50990
+    },
+    {
+      "epoch": 2.533028707658687,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00059736167676567,
+      "loss": 0.5815,
+      "step": 51000
+    },
+    {
+      "epoch": 2.5335253799543063,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005973219429820205,
+      "loss": 0.5638,
+      "step": 51010
+    },
+    {
+      "epoch": 2.5340220522499255,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.000597282209198371,
+      "loss": 0.5659,
+      "step": 51020
+    },
+    {
+      "epoch": 2.5345187245455447,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005972424754147214,
+      "loss": 0.5737,
+      "step": 51030
+    },
+    {
+      "epoch": 2.535015396841164,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005972027416310718,
+      "loss": 0.5632,
+      "step": 51040
+    },
+    {
+      "epoch": 2.5355120691367836,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005971630078474223,
+      "loss": 0.568,
+      "step": 51050
+    },
+    {
+      "epoch": 2.536008741432403,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005971232740637727,
+      "loss": 0.5503,
+      "step": 51060
+    },
+    {
+      "epoch": 2.5365054137280225,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005970835402801231,
+      "loss": 0.5756,
+      "step": 51070
+    },
+    {
+      "epoch": 2.5370020860236417,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005970438064964737,
+      "loss": 0.5414,
+      "step": 51080
+    },
+    {
+      "epoch": 2.537498758319261,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005970040727128242,
+      "loss": 0.5714,
+      "step": 51090
+    },
+    {
+      "epoch": 2.53799543061488,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005969643389291745,
+      "loss": 0.5618,
+      "step": 51100
+    },
+    {
+      "epoch": 2.5384921029104994,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.000596924605145525,
+      "loss": 0.551,
+      "step": 51110
+    },
+    {
+      "epoch": 2.538988775206119,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005968848713618754,
+      "loss": 0.5455,
+      "step": 51120
+    },
+    {
+      "epoch": 2.5394854475017383,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005968451375782259,
+      "loss": 0.5642,
+      "step": 51130
+    },
+    {
+      "epoch": 2.5399821197973576,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005968054037945764,
+      "loss": 0.5395,
+      "step": 51140
+    },
+    {
+      "epoch": 2.5404787920929772,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005967656700109268,
+      "loss": 0.5503,
+      "step": 51150
+    },
+    {
+      "epoch": 2.5409754643885964,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005967259362272772,
+      "loss": 0.5227,
+      "step": 51160
+    },
+    {
+      "epoch": 2.5414721366842157,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005966862024436278,
+      "loss": 0.5776,
+      "step": 51170
+    },
+    {
+      "epoch": 2.541968808979835,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005966464686599782,
+      "loss": 0.5362,
+      "step": 51180
+    },
+    {
+      "epoch": 2.5424654812754546,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005966067348763286,
+      "loss": 0.5623,
+      "step": 51190
+    },
+    {
+      "epoch": 2.542962153571074,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005965670010926791,
+      "loss": 0.5747,
+      "step": 51200
+    },
+    {
+      "epoch": 2.543458825866693,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005965272673090295,
+      "loss": 0.5386,
+      "step": 51210
+    },
+    {
+      "epoch": 2.5439554981623127,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.00059648753352538,
+      "loss": 0.5459,
+      "step": 51220
+    },
+    {
+      "epoch": 2.544452170457932,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005964477997417305,
+      "loss": 0.5909,
+      "step": 51230
+    },
+    {
+      "epoch": 2.544948842753551,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005964080659580809,
+      "loss": 0.5465,
+      "step": 51240
+    },
+    {
+      "epoch": 2.5454455150491704,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005963683321744314,
+      "loss": 0.5811,
+      "step": 51250
+    },
+    {
+      "epoch": 2.54594218734479,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005963285983907817,
+      "loss": 0.5755,
+      "step": 51260
+    },
+    {
+      "epoch": 2.5464388596404093,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005962888646071322,
+      "loss": 0.5701,
+      "step": 51270
+    },
+    {
+      "epoch": 2.5469355319360285,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005962491308234828,
+      "loss": 0.5748,
+      "step": 51280
+    },
+    {
+      "epoch": 2.547432204231648,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005962093970398331,
+      "loss": 0.5806,
+      "step": 51290
+    },
+    {
+      "epoch": 2.5479288765272674,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005961696632561836,
+      "loss": 0.5596,
+      "step": 51300
+    },
+    {
+      "epoch": 2.5484255488228866,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.000596129929472534,
+      "loss": 0.5824,
+      "step": 51310
+    },
+    {
+      "epoch": 2.548922221118506,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005960901956888844,
+      "loss": 0.5651,
+      "step": 51320
+    },
+    {
+      "epoch": 2.549418893414125,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.000596050461905235,
+      "loss": 0.5594,
+      "step": 51330
+    },
+    {
+      "epoch": 2.5499155657097448,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005960107281215854,
+      "loss": 0.5409,
+      "step": 51340
+    },
+    {
+      "epoch": 2.550412238005364,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005959709943379358,
+      "loss": 0.5479,
+      "step": 51350
+    },
+    {
+      "epoch": 2.5509089103009837,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005959312605542863,
+      "loss": 0.5584,
+      "step": 51360
+    },
+    {
+      "epoch": 2.551405582596603,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005958915267706367,
+      "loss": 0.568,
+      "step": 51370
+    },
+    {
+      "epoch": 2.551902254892222,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005958517929869873,
+      "loss": 0.5619,
+      "step": 51380
+    },
+    {
+      "epoch": 2.5523989271878413,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005958120592033377,
+      "loss": 0.5573,
+      "step": 51390
+    },
+    {
+      "epoch": 2.5528955994834606,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005957723254196881,
+      "loss": 0.561,
+      "step": 51400
+    },
+    {
+      "epoch": 2.5533922717790802,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005957325916360386,
+      "loss": 0.5522,
+      "step": 51410
+    },
+    {
+      "epoch": 2.5538889440746995,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.000595692857852389,
+      "loss": 0.5504,
+      "step": 51420
+    },
+    {
+      "epoch": 2.554385616370319,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005956531240687395,
+      "loss": 0.5398,
+      "step": 51430
+    },
+    {
+      "epoch": 2.5548822886659384,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.00059561339028509,
+      "loss": 0.5599,
+      "step": 51440
+    },
+    {
+      "epoch": 2.5553789609615576,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005955736565014403,
+      "loss": 0.5635,
+      "step": 51450
+    },
+    {
+      "epoch": 2.555875633257177,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005955339227177908,
+      "loss": 0.5605,
+      "step": 51460
+    },
+    {
+      "epoch": 2.556372305552796,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005954941889341414,
+      "loss": 0.5695,
+      "step": 51470
+    },
+    {
+      "epoch": 2.5568689778484157,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005954544551504917,
+      "loss": 0.576,
+      "step": 51480
+    },
+    {
+      "epoch": 2.557365650144035,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005954147213668422,
+      "loss": 0.5342,
+      "step": 51490
+    },
+    {
+      "epoch": 2.557862322439654,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005953749875831927,
+      "loss": 0.5262,
+      "step": 51500
+    },
+    {
+      "epoch": 2.558358994735274,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.000595335253799543,
+      "loss": 0.5752,
+      "step": 51510
+    },
+    {
+      "epoch": 2.558855667030893,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0005952955200158935,
+      "loss": 0.5404,
+      "step": 51520
+    },
+    {
+      "epoch": 2.5593523393265123,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.000595255786232244,
+      "loss": 0.583,
+      "step": 51530
+    },
+    {
+      "epoch": 2.5598490116221315,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005952160524485945,
+      "loss": 0.5767,
+      "step": 51540
+    },
+    {
+      "epoch": 2.560345683917751,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005951763186649449,
+      "loss": 0.5468,
+      "step": 51550
+    },
+    {
+      "epoch": 2.5608423562133704,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005951365848812953,
+      "loss": 0.5662,
+      "step": 51560
+    },
+    {
+      "epoch": 2.5613390285089896,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005950968510976458,
+      "loss": 0.5559,
+      "step": 51570
+    },
+    {
+      "epoch": 2.5618357008046093,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005950571173139963,
+      "loss": 0.5359,
+      "step": 51580
+    },
+    {
+      "epoch": 2.5623323731002285,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005950173835303467,
+      "loss": 0.5524,
+      "step": 51590
+    },
+    {
+      "epoch": 2.5628290453958478,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005949776497466972,
+      "loss": 0.5748,
+      "step": 51600
+    },
+    {
+      "epoch": 2.563325717691467,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005949379159630476,
+      "loss": 0.589,
+      "step": 51610
+    },
+    {
+      "epoch": 2.5638223899870867,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.000594898182179398,
+      "loss": 0.5767,
+      "step": 51620
+    },
+    {
+      "epoch": 2.564319062282706,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005948584483957486,
+      "loss": 0.5685,
+      "step": 51630
+    },
+    {
+      "epoch": 2.564815734578325,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005948187146120989,
+      "loss": 0.5551,
+      "step": 51640
+    },
+    {
+      "epoch": 2.565312406873945,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005947789808284494,
+      "loss": 0.5697,
+      "step": 51650
+    },
+    {
+      "epoch": 2.565809079169564,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005947392470447999,
+      "loss": 0.5754,
+      "step": 51660
+    },
+    {
+      "epoch": 2.5663057514651832,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005946995132611503,
+      "loss": 0.5471,
+      "step": 51670
+    },
+    {
+      "epoch": 2.5668024237608025,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005946597794775008,
+      "loss": 0.5898,
+      "step": 51680
+    },
+    {
+      "epoch": 2.5672990960564217,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0005946200456938513,
+      "loss": 0.5595,
+      "step": 51690
+    },
+    {
+      "epoch": 2.5677957683520414,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005945803119102017,
+      "loss": 0.5411,
+      "step": 51700
+    },
+    {
+      "epoch": 2.5682924406476606,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005945405781265521,
+      "loss": 0.5632,
+      "step": 51710
+    },
+    {
+      "epoch": 2.5687891129432803,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005945008443429025,
+      "loss": 0.5642,
+      "step": 51720
+    },
+    {
+      "epoch": 2.5692857852388995,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005944611105592531,
+      "loss": 0.5499,
+      "step": 51730
+    },
+    {
+      "epoch": 2.5697824575345187,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005944213767756035,
+      "loss": 0.5764,
+      "step": 51740
+    },
+    {
+      "epoch": 2.570279129830138,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005943816429919539,
+      "loss": 0.53,
+      "step": 51750
+    },
+    {
+      "epoch": 2.570775802125757,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005943419092083044,
+      "loss": 0.6109,
+      "step": 51760
+    },
+    {
+      "epoch": 2.571272474421377,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005943021754246548,
+      "loss": 0.5751,
+      "step": 51770
+    },
+    {
+      "epoch": 2.571769146716996,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005942624416410053,
+      "loss": 0.5906,
+      "step": 51780
+    },
+    {
+      "epoch": 2.5722658190126158,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005942227078573558,
+      "loss": 0.5624,
+      "step": 51790
+    },
+    {
+      "epoch": 2.572762491308235,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005941829740737062,
+      "loss": 0.5324,
+      "step": 51800
+    },
+    {
+      "epoch": 2.573259163603854,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005941432402900566,
+      "loss": 0.5542,
+      "step": 51810
+    },
+    {
+      "epoch": 2.5737558358994734,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005941035065064071,
+      "loss": 0.5472,
+      "step": 51820
+    },
+    {
+      "epoch": 2.5742525081950927,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005940637727227576,
+      "loss": 0.5527,
+      "step": 51830
+    },
+    {
+      "epoch": 2.5747491804907123,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.000594024038939108,
+      "loss": 0.5812,
+      "step": 51840
+    },
+    {
+      "epoch": 2.5752458527863316,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005939843051554585,
+      "loss": 0.5717,
+      "step": 51850
+    },
+    {
+      "epoch": 2.575742525081951,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005939445713718089,
+      "loss": 0.5504,
+      "step": 51860
+    },
+    {
+      "epoch": 2.5762391973775705,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005939048375881593,
+      "loss": 0.5747,
+      "step": 51870
+    },
+    {
+      "epoch": 2.5767358696731897,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0005938651038045099,
+      "loss": 0.5129,
+      "step": 51880
+    },
+    {
+      "epoch": 2.577232541968809,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005938253700208603,
+      "loss": 0.5677,
+      "step": 51890
+    },
+    {
+      "epoch": 2.577729214264428,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005937856362372107,
+      "loss": 0.5708,
+      "step": 51900
+    },
+    {
+      "epoch": 2.578225886560048,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005937459024535611,
+      "loss": 0.5443,
+      "step": 51910
+    },
+    {
+      "epoch": 2.578722558855667,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005937061686699116,
+      "loss": 0.5586,
+      "step": 51920
+    },
+    {
+      "epoch": 2.5792192311512863,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005936664348862621,
+      "loss": 0.5766,
+      "step": 51930
+    },
+    {
+      "epoch": 2.579715903446906,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005936267011026125,
+      "loss": 0.579,
+      "step": 51940
+    },
+    {
+      "epoch": 2.580212575742525,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.000593586967318963,
+      "loss": 0.5805,
+      "step": 51950
+    },
+    {
+      "epoch": 2.5807092480381444,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005935472335353134,
+      "loss": 0.5551,
+      "step": 51960
+    },
+    {
+      "epoch": 2.5812059203337636,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005935074997516638,
+      "loss": 0.5564,
+      "step": 51970
+    },
+    {
+      "epoch": 2.5817025926293833,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005934677659680144,
+      "loss": 0.5701,
+      "step": 51980
+    },
+    {
+      "epoch": 2.5821992649250025,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005934280321843648,
+      "loss": 0.5799,
+      "step": 51990
+    },
+    {
+      "epoch": 2.5826959372206217,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005933882984007152,
+      "loss": 0.5812,
+      "step": 52000
+    },
+    {
+      "epoch": 2.5831926095162414,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005933485646170657,
+      "loss": 0.5707,
+      "step": 52010
+    },
+    {
+      "epoch": 2.5836892818118606,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005933088308334161,
+      "loss": 0.5802,
+      "step": 52020
+    },
+    {
+      "epoch": 2.58418595410748,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005932690970497666,
+      "loss": 0.5655,
+      "step": 52030
+    },
+    {
+      "epoch": 2.584682626403099,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005932293632661171,
+      "loss": 0.5621,
+      "step": 52040
+    },
+    {
+      "epoch": 2.5851792986987183,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005931896294824675,
+      "loss": 0.5851,
+      "step": 52050
+    },
+    {
+      "epoch": 2.585675970994338,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005931498956988179,
+      "loss": 0.5652,
+      "step": 52060
+    },
+    {
+      "epoch": 2.586172643289957,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005931101619151684,
+      "loss": 0.5645,
+      "step": 52070
+    },
+    {
+      "epoch": 2.586669315585577,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005930704281315189,
+      "loss": 0.5482,
+      "step": 52080
+    },
+    {
+      "epoch": 2.587165987881196,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005930306943478693,
+      "loss": 0.5524,
+      "step": 52090
+    },
+    {
+      "epoch": 2.5876626601768153,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005929909605642197,
+      "loss": 0.5543,
+      "step": 52100
+    },
+    {
+      "epoch": 2.5881593324724346,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005929512267805702,
+      "loss": 0.5657,
+      "step": 52110
+    },
+    {
+      "epoch": 2.588656004768054,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005929114929969207,
+      "loss": 0.5542,
+      "step": 52120
+    },
+    {
+      "epoch": 2.5891526770636735,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.000592871759213271,
+      "loss": 0.571,
+      "step": 52130
+    },
+    {
+      "epoch": 2.5896493493592927,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005928320254296216,
+      "loss": 0.5935,
+      "step": 52140
+    },
+    {
+      "epoch": 2.590146021654912,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005927922916459721,
+      "loss": 0.5556,
+      "step": 52150
+    },
+    {
+      "epoch": 2.5906426939505316,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005927525578623224,
+      "loss": 0.5633,
+      "step": 52160
+    },
+    {
+      "epoch": 2.591139366246151,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005927128240786729,
+      "loss": 0.5721,
+      "step": 52170
+    },
+    {
+      "epoch": 2.59163603854177,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005926730902950234,
+      "loss": 0.5669,
+      "step": 52180
+    },
+    {
+      "epoch": 2.5921327108373893,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005926333565113738,
+      "loss": 0.5699,
+      "step": 52190
+    },
+    {
+      "epoch": 2.592629383133009,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005925936227277243,
+      "loss": 0.5165,
+      "step": 52200
+    },
+    {
+      "epoch": 2.593126055428628,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005925538889440747,
+      "loss": 0.5432,
+      "step": 52210
+    },
+    {
+      "epoch": 2.5936227277242474,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005925141551604251,
+      "loss": 0.5869,
+      "step": 52220
+    },
+    {
+      "epoch": 2.594119400019867,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005924744213767757,
+      "loss": 0.5908,
+      "step": 52230
+    },
+    {
+      "epoch": 2.5946160723154863,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005924346875931261,
+      "loss": 0.5426,
+      "step": 52240
+    },
+    {
+      "epoch": 2.5951127446111055,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005923949538094765,
+      "loss": 0.5939,
+      "step": 52250
+    },
+    {
+      "epoch": 2.5956094169067248,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.000592355220025827,
+      "loss": 0.5533,
+      "step": 52260
+    },
+    {
+      "epoch": 2.5961060892023444,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005923154862421774,
+      "loss": 0.5361,
+      "step": 52270
+    },
+    {
+      "epoch": 2.5966027614979637,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.000592275752458528,
+      "loss": 0.5498,
+      "step": 52280
+    },
+    {
+      "epoch": 2.597099433793583,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005922360186748784,
+      "loss": 0.5545,
+      "step": 52290
+    },
+    {
+      "epoch": 2.5975961060892026,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005921962848912288,
+      "loss": 0.5817,
+      "step": 52300
+    },
+    {
+      "epoch": 2.5980927783848218,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005921565511075793,
+      "loss": 0.5931,
+      "step": 52310
+    },
+    {
+      "epoch": 2.598589450680441,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005921168173239296,
+      "loss": 0.5344,
+      "step": 52320
+    },
+    {
+      "epoch": 2.5990861229760602,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005920770835402802,
+      "loss": 0.5631,
+      "step": 52330
+    },
+    {
+      "epoch": 2.59958279527168,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005920373497566307,
+      "loss": 0.5554,
+      "step": 52340
+    },
+    {
+      "epoch": 2.600079467567299,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.000591997615972981,
+      "loss": 0.5733,
+      "step": 52350
+    },
+    {
+      "epoch": 2.6005761398629184,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005919578821893315,
+      "loss": 0.5849,
+      "step": 52360
+    },
+    {
+      "epoch": 2.601072812158538,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005919181484056819,
+      "loss": 0.5489,
+      "step": 52370
+    },
+    {
+      "epoch": 2.6015694844541573,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005918784146220323,
+      "loss": 0.5192,
+      "step": 52380
+    },
+    {
+      "epoch": 2.6020661567497765,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005918386808383829,
+      "loss": 0.5245,
+      "step": 52390
+    },
+    {
+      "epoch": 2.6025628290453957,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005917989470547333,
+      "loss": 0.5205,
+      "step": 52400
+    },
+    {
+      "epoch": 2.603059501341015,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005917592132710837,
+      "loss": 0.5696,
+      "step": 52410
+    },
+    {
+      "epoch": 2.6035561736366346,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005917194794874342,
+      "loss": 0.5737,
+      "step": 52420
+    },
+    {
+      "epoch": 2.604052845932254,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005916797457037846,
+      "loss": 0.5474,
+      "step": 52430
+    },
+    {
+      "epoch": 2.6045495182278735,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005916400119201352,
+      "loss": 0.5651,
+      "step": 52440
+    },
+    {
+      "epoch": 2.6050461905234927,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005916002781364856,
+      "loss": 0.5621,
+      "step": 52450
+    },
+    {
+      "epoch": 2.605542862819112,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.000591560544352836,
+      "loss": 0.5554,
+      "step": 52460
+    },
+    {
+      "epoch": 2.606039535114731,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005915208105691865,
+      "loss": 0.5747,
+      "step": 52470
+    },
+    {
+      "epoch": 2.6065362074103504,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.000591481076785537,
+      "loss": 0.5467,
+      "step": 52480
+    },
+    {
+      "epoch": 2.60703287970597,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005914413430018874,
+      "loss": 0.5722,
+      "step": 52490
+    },
+    {
+      "epoch": 2.6075295520015893,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005914016092182379,
+      "loss": 0.5646,
+      "step": 52500
+    },
+    {
+      "epoch": 2.6080262242972085,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005913618754345882,
+      "loss": 0.6004,
+      "step": 52510
+    },
+    {
+      "epoch": 2.608522896592828,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005913221416509387,
+      "loss": 0.5948,
+      "step": 52520
+    },
+    {
+      "epoch": 2.6090195688884474,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005912824078672893,
+      "loss": 0.5654,
+      "step": 52530
+    },
+    {
+      "epoch": 2.6095162411840667,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005912426740836396,
+      "loss": 0.5787,
+      "step": 52540
+    },
+    {
+      "epoch": 2.610012913479686,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005912029402999901,
+      "loss": 0.5808,
+      "step": 52550
+    },
+    {
+      "epoch": 2.6105095857753056,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005911632065163406,
+      "loss": 0.5478,
+      "step": 52560
+    },
+    {
+      "epoch": 2.611006258070925,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.000591123472732691,
+      "loss": 0.5723,
+      "step": 52570
+    },
+    {
+      "epoch": 2.611502930366544,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005910837389490414,
+      "loss": 0.541,
+      "step": 52580
+    },
+    {
+      "epoch": 2.6119996026621637,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005910440051653919,
+      "loss": 0.5594,
+      "step": 52590
+    },
+    {
+      "epoch": 2.612496274957783,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005910042713817424,
+      "loss": 0.5633,
+      "step": 52600
+    },
+    {
+      "epoch": 2.612992947253402,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005909645375980928,
+      "loss": 0.5304,
+      "step": 52610
+    },
+    {
+      "epoch": 2.6134896195490214,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005909248038144432,
+      "loss": 0.5781,
+      "step": 52620
+    },
+    {
+      "epoch": 2.613986291844641,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005908850700307938,
+      "loss": 0.525,
+      "step": 52630
+    },
+    {
+      "epoch": 2.6144829641402603,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005908453362471442,
+      "loss": 0.551,
+      "step": 52640
+    },
+    {
+      "epoch": 2.6149796364358795,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005908056024634946,
+      "loss": 0.5546,
+      "step": 52650
+    },
+    {
+      "epoch": 2.615476308731499,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005907658686798451,
+      "loss": 0.5552,
+      "step": 52660
+    },
+    {
+      "epoch": 2.6159729810271184,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005907261348961955,
+      "loss": 0.5338,
+      "step": 52670
+    },
+    {
+      "epoch": 2.6164696533227376,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005906864011125459,
+      "loss": 0.5431,
+      "step": 52680
+    },
+    {
+      "epoch": 2.616966325618357,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005906466673288965,
+      "loss": 0.5516,
+      "step": 52690
+    },
+    {
+      "epoch": 2.6174629979139765,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005906069335452468,
+      "loss": 0.5614,
+      "step": 52700
+    },
+    {
+      "epoch": 2.6179596702095957,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005905671997615973,
+      "loss": 0.5881,
+      "step": 52710
+    },
+    {
+      "epoch": 2.618456342505215,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005905274659779478,
+      "loss": 0.5574,
+      "step": 52720
+    },
+    {
+      "epoch": 2.6189530148008346,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005904877321942982,
+      "loss": 0.5601,
+      "step": 52730
+    },
+    {
+      "epoch": 2.619449687096454,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005904479984106487,
+      "loss": 0.5525,
+      "step": 52740
+    },
+    {
+      "epoch": 2.619946359392073,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005904082646269992,
+      "loss": 0.5587,
+      "step": 52750
+    },
+    {
+      "epoch": 2.6204430316876923,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005903685308433496,
+      "loss": 0.5613,
+      "step": 52760
+    },
+    {
+      "epoch": 2.6209397039833116,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005903287970597,
+      "loss": 0.53,
+      "step": 52770
+    },
+    {
+      "epoch": 2.6214363762789312,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005902890632760504,
+      "loss": 0.571,
+      "step": 52780
+    },
+    {
+      "epoch": 2.6219330485745505,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.000590249329492401,
+      "loss": 0.5459,
+      "step": 52790
+    },
+    {
+      "epoch": 2.62242972087017,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005902095957087514,
+      "loss": 0.5785,
+      "step": 52800
+    },
+    {
+      "epoch": 2.6229263931657893,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005901698619251018,
+      "loss": 0.5447,
+      "step": 52810
+    },
+    {
+      "epoch": 2.6234230654614086,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005901301281414523,
+      "loss": 0.5534,
+      "step": 52820
+    },
+    {
+      "epoch": 2.623919737757028,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005900903943578027,
+      "loss": 0.5597,
+      "step": 52830
+    },
+    {
+      "epoch": 2.624416410052647,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0005900506605741532,
+      "loss": 0.5841,
+      "step": 52840
+    },
+    {
+      "epoch": 2.6249130823482667,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005900109267905037,
+      "loss": 0.5437,
+      "step": 52850
+    },
+    {
+      "epoch": 2.625409754643886,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005899711930068541,
+      "loss": 0.5652,
+      "step": 52860
+    },
+    {
+      "epoch": 2.625906426939505,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005899314592232045,
+      "loss": 0.563,
+      "step": 52870
+    },
+    {
+      "epoch": 2.626403099235125,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.000589891725439555,
+      "loss": 0.5704,
+      "step": 52880
+    },
+    {
+      "epoch": 2.626899771530744,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005898519916559055,
+      "loss": 0.5478,
+      "step": 52890
+    },
+    {
+      "epoch": 2.6273964438263633,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005898122578722559,
+      "loss": 0.5751,
+      "step": 52900
+    },
+    {
+      "epoch": 2.6278931161219825,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005897725240886064,
+      "loss": 0.5358,
+      "step": 52910
+    },
+    {
+      "epoch": 2.628389788417602,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005897327903049568,
+      "loss": 0.5574,
+      "step": 52920
+    },
+    {
+      "epoch": 2.6288864607132214,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005896930565213072,
+      "loss": 0.5408,
+      "step": 52930
+    },
+    {
+      "epoch": 2.6293831330088406,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005896533227376578,
+      "loss": 0.5505,
+      "step": 52940
+    },
+    {
+      "epoch": 2.6298798053044603,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005896135889540082,
+      "loss": 0.5553,
+      "step": 52950
+    },
+    {
+      "epoch": 2.6303764776000795,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005895738551703586,
+      "loss": 0.5837,
+      "step": 52960
+    },
+    {
+      "epoch": 2.6308731498956988,
+      "grad_norm": 0.125,
+      "learning_rate": 0.000589534121386709,
+      "loss": 0.5853,
+      "step": 52970
+    },
+    {
+      "epoch": 2.631369822191318,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005894943876030595,
+      "loss": 0.5668,
+      "step": 52980
+    },
+    {
+      "epoch": 2.6318664944869377,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00058945465381941,
+      "loss": 0.5911,
+      "step": 52990
+    },
+    {
+      "epoch": 2.632363166782557,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005894149200357604,
+      "loss": 0.5381,
+      "step": 53000
+    },
+    {
+      "epoch": 2.632859839078176,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005893751862521109,
+      "loss": 0.543,
+      "step": 53010
+    },
+    {
+      "epoch": 2.633356511373796,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005893354524684614,
+      "loss": 0.5565,
+      "step": 53020
+    },
+    {
+      "epoch": 2.633853183669415,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005892957186848117,
+      "loss": 0.5544,
+      "step": 53030
+    },
+    {
+      "epoch": 2.6343498559650342,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005892559849011623,
+      "loss": 0.5533,
+      "step": 53040
+    },
+    {
+      "epoch": 2.6348465282606535,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005892162511175127,
+      "loss": 0.5711,
+      "step": 53050
+    },
+    {
+      "epoch": 2.635343200556273,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005891765173338631,
+      "loss": 0.5413,
+      "step": 53060
+    },
+    {
+      "epoch": 2.6358398728518924,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005891367835502136,
+      "loss": 0.5541,
+      "step": 53070
+    },
+    {
+      "epoch": 2.6363365451475116,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.000589097049766564,
+      "loss": 0.6096,
+      "step": 53080
+    },
+    {
+      "epoch": 2.6368332174431313,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005890573159829145,
+      "loss": 0.5926,
+      "step": 53090
+    },
+    {
+      "epoch": 2.6373298897387505,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.000589017582199265,
+      "loss": 0.5598,
+      "step": 53100
+    },
+    {
+      "epoch": 2.6378265620343697,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005889778484156154,
+      "loss": 0.5967,
+      "step": 53110
+    },
+    {
+      "epoch": 2.638323234329989,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005889381146319658,
+      "loss": 0.5706,
+      "step": 53120
+    },
+    {
+      "epoch": 2.638819906625608,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005888983808483163,
+      "loss": 0.5699,
+      "step": 53130
+    },
+    {
+      "epoch": 2.639316578921228,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0005888586470646668,
+      "loss": 0.5778,
+      "step": 53140
+    },
+    {
+      "epoch": 2.639813251216847,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005888189132810172,
+      "loss": 0.5371,
+      "step": 53150
+    },
+    {
+      "epoch": 2.6403099235124667,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005887791794973677,
+      "loss": 0.5268,
+      "step": 53160
+    },
+    {
+      "epoch": 2.640806595808086,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005887394457137181,
+      "loss": 0.5741,
+      "step": 53170
+    },
+    {
+      "epoch": 2.641303268103705,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005886997119300686,
+      "loss": 0.5917,
+      "step": 53180
+    },
+    {
+      "epoch": 2.6417999403993244,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.000588659978146419,
+      "loss": 0.5463,
+      "step": 53190
+    },
+    {
+      "epoch": 2.6422966126949436,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005886202443627695,
+      "loss": 0.5711,
+      "step": 53200
+    },
+    {
+      "epoch": 2.6427932849905633,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00058858051057912,
+      "loss": 0.5452,
+      "step": 53210
+    },
+    {
+      "epoch": 2.6432899572861825,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005885407767954703,
+      "loss": 0.5595,
+      "step": 53220
+    },
+    {
+      "epoch": 2.6437866295818018,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005885010430118208,
+      "loss": 0.5474,
+      "step": 53230
+    },
+    {
+      "epoch": 2.6442833018774214,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005884613092281713,
+      "loss": 0.5801,
+      "step": 53240
+    },
+    {
+      "epoch": 2.6447799741730407,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005884215754445217,
+      "loss": 0.5728,
+      "step": 53250
+    },
+    {
+      "epoch": 2.64527664646866,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005883818416608722,
+      "loss": 0.5775,
+      "step": 53260
+    },
+    {
+      "epoch": 2.645773318764279,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005883421078772226,
+      "loss": 0.5507,
+      "step": 53270
+    },
+    {
+      "epoch": 2.646269991059899,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.000588302374093573,
+      "loss": 0.5794,
+      "step": 53280
+    },
+    {
+      "epoch": 2.646766663355518,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005882626403099236,
+      "loss": 0.5759,
+      "step": 53290
+    },
+    {
+      "epoch": 2.6472633356511373,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.000588222906526274,
+      "loss": 0.5717,
+      "step": 53300
+    },
+    {
+      "epoch": 2.647760007946757,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005881831727426245,
+      "loss": 0.5721,
+      "step": 53310
+    },
+    {
+      "epoch": 2.648256680242376,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005881434389589749,
+      "loss": 0.5629,
+      "step": 53320
+    },
+    {
+      "epoch": 2.6487533525379954,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005881037051753253,
+      "loss": 0.5881,
+      "step": 53330
+    },
+    {
+      "epoch": 2.6492500248336146,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005880639713916759,
+      "loss": 0.5568,
+      "step": 53340
+    },
+    {
+      "epoch": 2.6497466971292343,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005880242376080263,
+      "loss": 0.5692,
+      "step": 53350
+    },
+    {
+      "epoch": 2.6502433694248535,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005879845038243767,
+      "loss": 0.5416,
+      "step": 53360
+    },
+    {
+      "epoch": 2.6507400417204727,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005879447700407272,
+      "loss": 0.5644,
+      "step": 53370
+    },
+    {
+      "epoch": 2.6512367140160924,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005879050362570775,
+      "loss": 0.5573,
+      "step": 53380
+    },
+    {
+      "epoch": 2.6517333863117116,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005878653024734281,
+      "loss": 0.5575,
+      "step": 53390
+    },
+    {
+      "epoch": 2.652230058607331,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005878255686897786,
+      "loss": 0.5463,
+      "step": 53400
+    },
+    {
+      "epoch": 2.65272673090295,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005877858349061289,
+      "loss": 0.5543,
+      "step": 53410
+    },
+    {
+      "epoch": 2.6532234031985693,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005877461011224794,
+      "loss": 0.5464,
+      "step": 53420
+    },
+    {
+      "epoch": 2.653720075494189,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005877063673388299,
+      "loss": 0.5422,
+      "step": 53430
+    },
+    {
+      "epoch": 2.654216747789808,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005876666335551802,
+      "loss": 0.5569,
+      "step": 53440
+    },
+    {
+      "epoch": 2.654713420085428,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005876268997715308,
+      "loss": 0.5385,
+      "step": 53450
+    },
+    {
+      "epoch": 2.655210092381047,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005875871659878812,
+      "loss": 0.5348,
+      "step": 53460
+    },
+    {
+      "epoch": 2.6557067646766663,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005875474322042317,
+      "loss": 0.5475,
+      "step": 53470
+    },
+    {
+      "epoch": 2.6562034369722856,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005875076984205821,
+      "loss": 0.5525,
+      "step": 53480
+    },
+    {
+      "epoch": 2.656700109267905,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005874679646369326,
+      "loss": 0.5494,
+      "step": 53490
+    },
+    {
+      "epoch": 2.6571967815635245,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005874282308532831,
+      "loss": 0.5446,
+      "step": 53500
+    },
+    {
+      "epoch": 2.6576934538591437,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005873884970696335,
+      "loss": 0.5354,
+      "step": 53510
+    },
+    {
+      "epoch": 2.6581901261547634,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005873487632859839,
+      "loss": 0.5875,
+      "step": 53520
+    },
+    {
+      "epoch": 2.6586867984503826,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005873090295023344,
+      "loss": 0.5827,
+      "step": 53530
+    },
+    {
+      "epoch": 2.659183470746002,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005872692957186849,
+      "loss": 0.5811,
+      "step": 53540
+    },
+    {
+      "epoch": 2.659680143041621,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005872295619350353,
+      "loss": 0.5338,
+      "step": 53550
+    },
+    {
+      "epoch": 2.6601768153372403,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005871898281513858,
+      "loss": 0.5841,
+      "step": 53560
+    },
+    {
+      "epoch": 2.66067348763286,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005871500943677361,
+      "loss": 0.5524,
+      "step": 53570
+    },
+    {
+      "epoch": 2.661170159928479,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005871103605840866,
+      "loss": 0.5401,
+      "step": 53580
+    },
+    {
+      "epoch": 2.6616668322240984,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005870706268004372,
+      "loss": 0.5596,
+      "step": 53590
+    },
+    {
+      "epoch": 2.662163504519718,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005870308930167875,
+      "loss": 0.5294,
+      "step": 53600
+    },
+    {
+      "epoch": 2.6626601768153373,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.000586991159233138,
+      "loss": 0.5505,
+      "step": 53610
+    },
+    {
+      "epoch": 2.6631568491109565,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005869514254494885,
+      "loss": 0.5808,
+      "step": 53620
+    },
+    {
+      "epoch": 2.6636535214065757,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005869116916658389,
+      "loss": 0.559,
+      "step": 53630
+    },
+    {
+      "epoch": 2.6641501937021954,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005868719578821894,
+      "loss": 0.5737,
+      "step": 53640
+    },
+    {
+      "epoch": 2.6646468659978146,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005868322240985398,
+      "loss": 0.5706,
+      "step": 53650
+    },
+    {
+      "epoch": 2.665143538293434,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005867924903148903,
+      "loss": 0.5827,
+      "step": 53660
+    },
+    {
+      "epoch": 2.6656402105890535,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005867527565312407,
+      "loss": 0.5591,
+      "step": 53670
+    },
+    {
+      "epoch": 2.6661368828846728,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005867130227475911,
+      "loss": 0.5814,
+      "step": 53680
+    },
+    {
+      "epoch": 2.666633555180292,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005866732889639417,
+      "loss": 0.5339,
+      "step": 53690
+    },
+    {
+      "epoch": 2.667130227475911,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005866335551802921,
+      "loss": 0.5854,
+      "step": 53700
+    },
+    {
+      "epoch": 2.667626899771531,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005865938213966425,
+      "loss": 0.5793,
+      "step": 53710
+    },
+    {
+      "epoch": 2.66812357206715,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.000586554087612993,
+      "loss": 0.5616,
+      "step": 53720
+    },
+    {
+      "epoch": 2.6686202443627693,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005865143538293434,
+      "loss": 0.5383,
+      "step": 53730
+    },
+    {
+      "epoch": 2.669116916658389,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005864746200456938,
+      "loss": 0.5441,
+      "step": 53740
+    },
+    {
+      "epoch": 2.6696135889540082,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005864348862620444,
+      "loss": 0.554,
+      "step": 53750
+    },
+    {
+      "epoch": 2.6701102612496275,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005863951524783948,
+      "loss": 0.5615,
+      "step": 53760
+    },
+    {
+      "epoch": 2.6706069335452467,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005863554186947452,
+      "loss": 0.5313,
+      "step": 53770
+    },
+    {
+      "epoch": 2.671103605840866,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005863156849110957,
+      "loss": 0.5787,
+      "step": 53780
+    },
+    {
+      "epoch": 2.6716002781364856,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005862759511274462,
+      "loss": 0.5651,
+      "step": 53790
+    },
+    {
+      "epoch": 2.672096950432105,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005862362173437966,
+      "loss": 0.5787,
+      "step": 53800
+    },
+    {
+      "epoch": 2.6725936227277245,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005861964835601471,
+      "loss": 0.554,
+      "step": 53810
+    },
+    {
+      "epoch": 2.6730902950233437,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005861567497764975,
+      "loss": 0.5671,
+      "step": 53820
+    },
+    {
+      "epoch": 2.673586967318963,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005861170159928479,
+      "loss": 0.5346,
+      "step": 53830
+    },
+    {
+      "epoch": 2.674083639614582,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005860772822091983,
+      "loss": 0.5534,
+      "step": 53840
+    },
+    {
+      "epoch": 2.6745803119102014,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005860375484255489,
+      "loss": 0.5411,
+      "step": 53850
+    },
+    {
+      "epoch": 2.675076984205821,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005859978146418993,
+      "loss": 0.5475,
+      "step": 53860
+    },
+    {
+      "epoch": 2.6755736565014403,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005859580808582497,
+      "loss": 0.5515,
+      "step": 53870
+    },
+    {
+      "epoch": 2.67607032879706,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005859183470746002,
+      "loss": 0.575,
+      "step": 53880
+    },
+    {
+      "epoch": 2.676567001092679,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005858786132909506,
+      "loss": 0.531,
+      "step": 53890
+    },
+    {
+      "epoch": 2.6770636733882984,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005858388795073011,
+      "loss": 0.5366,
+      "step": 53900
+    },
+    {
+      "epoch": 2.6775603456839177,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005857991457236516,
+      "loss": 0.5634,
+      "step": 53910
+    },
+    {
+      "epoch": 2.678057017979537,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.000585759411940002,
+      "loss": 0.542,
+      "step": 53920
+    },
+    {
+      "epoch": 2.6785536902751566,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005857196781563524,
+      "loss": 0.5662,
+      "step": 53930
+    },
+    {
+      "epoch": 2.679050362570776,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.000585679944372703,
+      "loss": 0.5822,
+      "step": 53940
+    },
+    {
+      "epoch": 2.679547034866395,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005856402105890534,
+      "loss": 0.5319,
+      "step": 53950
+    },
+    {
+      "epoch": 2.6800437071620147,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0005856004768054038,
+      "loss": 0.5512,
+      "step": 53960
+    },
+    {
+      "epoch": 2.680540379457634,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005855607430217543,
+      "loss": 0.5579,
+      "step": 53970
+    },
+    {
+      "epoch": 2.681037051753253,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005855210092381047,
+      "loss": 0.5863,
+      "step": 53980
+    },
+    {
+      "epoch": 2.6815337240488724,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005854812754544551,
+      "loss": 0.5027,
+      "step": 53990
+    },
+    {
+      "epoch": 2.682030396344492,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005854415416708057,
+      "loss": 0.5308,
+      "step": 54000
+    },
+    {
+      "epoch": 2.6825270686401113,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005854018078871561,
+      "loss": 0.5463,
+      "step": 54010
+    },
+    {
+      "epoch": 2.6830237409357305,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005853620741035065,
+      "loss": 0.5609,
+      "step": 54020
+    },
+    {
+      "epoch": 2.68352041323135,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.000585322340319857,
+      "loss": 0.5414,
+      "step": 54030
+    },
+    {
+      "epoch": 2.6840170855269694,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005852826065362074,
+      "loss": 0.556,
+      "step": 54040
+    },
+    {
+      "epoch": 2.6845137578225886,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005852428727525579,
+      "loss": 0.5399,
+      "step": 54050
+    },
+    {
+      "epoch": 2.685010430118208,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005852031389689083,
+      "loss": 0.5864,
+      "step": 54060
+    },
+    {
+      "epoch": 2.6855071024138275,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005851634051852588,
+      "loss": 0.5751,
+      "step": 54070
+    },
+    {
+      "epoch": 2.6860037747094467,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005851236714016093,
+      "loss": 0.5782,
+      "step": 54080
+    },
+    {
+      "epoch": 2.686500447005066,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005850839376179596,
+      "loss": 0.5686,
+      "step": 54090
+    },
+    {
+      "epoch": 2.6869971193006856,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005850442038343102,
+      "loss": 0.5474,
+      "step": 54100
+    },
+    {
+      "epoch": 2.687493791596305,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005850044700506606,
+      "loss": 0.5418,
+      "step": 54110
+    },
+    {
+      "epoch": 2.687990463891924,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.000584964736267011,
+      "loss": 0.541,
+      "step": 54120
+    },
+    {
+      "epoch": 2.6884871361875433,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005849250024833615,
+      "loss": 0.5756,
+      "step": 54130
+    },
+    {
+      "epoch": 2.6889838084831625,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005848852686997119,
+      "loss": 0.5503,
+      "step": 54140
+    },
+    {
+      "epoch": 2.689480480778782,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005848455349160624,
+      "loss": 0.551,
+      "step": 54150
+    },
+    {
+      "epoch": 2.6899771530744014,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005848058011324129,
+      "loss": 0.5432,
+      "step": 54160
+    },
+    {
+      "epoch": 2.690473825370021,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005847660673487633,
+      "loss": 0.5528,
+      "step": 54170
+    },
+    {
+      "epoch": 2.6909704976656403,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005847263335651137,
+      "loss": 0.5809,
+      "step": 54180
+    },
+    {
+      "epoch": 2.6914671699612596,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005846865997814642,
+      "loss": 0.5633,
+      "step": 54190
+    },
+    {
+      "epoch": 2.691963842256879,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005846468659978147,
+      "loss": 0.5667,
+      "step": 54200
+    },
+    {
+      "epoch": 2.692460514552498,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005846071322141652,
+      "loss": 0.5571,
+      "step": 54210
+    },
+    {
+      "epoch": 2.6929571868481177,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005845673984305156,
+      "loss": 0.5512,
+      "step": 54220
+    },
+    {
+      "epoch": 2.693453859143737,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.000584527664646866,
+      "loss": 0.5544,
+      "step": 54230
+    },
+    {
+      "epoch": 2.6939505314393566,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005844879308632165,
+      "loss": 0.5588,
+      "step": 54240
+    },
+    {
+      "epoch": 2.694447203734976,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0005844481970795669,
+      "loss": 0.5605,
+      "step": 54250
+    },
+    {
+      "epoch": 2.694943876030595,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005844084632959174,
+      "loss": 0.5892,
+      "step": 54260
+    },
+    {
+      "epoch": 2.6954405483262143,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005843687295122679,
+      "loss": 0.5496,
+      "step": 54270
+    },
+    {
+      "epoch": 2.6959372206218335,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005843289957286182,
+      "loss": 0.5694,
+      "step": 54280
+    },
+    {
+      "epoch": 2.696433892917453,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005842892619449687,
+      "loss": 0.5944,
+      "step": 54290
+    },
+    {
+      "epoch": 2.6969305652130724,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005842495281613193,
+      "loss": 0.5402,
+      "step": 54300
+    },
+    {
+      "epoch": 2.6974272375086916,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005842097943776696,
+      "loss": 0.5805,
+      "step": 54310
+    },
+    {
+      "epoch": 2.6979239098043113,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005841700605940201,
+      "loss": 0.5464,
+      "step": 54320
+    },
+    {
+      "epoch": 2.6984205820999305,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005841303268103705,
+      "loss": 0.5548,
+      "step": 54330
+    },
+    {
+      "epoch": 2.6989172543955497,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005840905930267209,
+      "loss": 0.5549,
+      "step": 54340
+    },
+    {
+      "epoch": 2.699413926691169,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005840508592430715,
+      "loss": 0.5741,
+      "step": 54350
+    },
+    {
+      "epoch": 2.6999105989867886,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005840111254594219,
+      "loss": 0.582,
+      "step": 54360
+    },
+    {
+      "epoch": 2.700407271282408,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005839713916757724,
+      "loss": 0.5709,
+      "step": 54370
+    },
+    {
+      "epoch": 2.700903943578027,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005839316578921228,
+      "loss": 0.5393,
+      "step": 54380
+    },
+    {
+      "epoch": 2.7014006158736468,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005838919241084732,
+      "loss": 0.5663,
+      "step": 54390
+    },
+    {
+      "epoch": 2.701897288169266,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005838521903248238,
+      "loss": 0.5497,
+      "step": 54400
+    },
+    {
+      "epoch": 2.7023939604648852,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005838124565411742,
+      "loss": 0.5285,
+      "step": 54410
+    },
+    {
+      "epoch": 2.7028906327605045,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005837727227575246,
+      "loss": 0.5538,
+      "step": 54420
+    },
+    {
+      "epoch": 2.703387305056124,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005837329889738751,
+      "loss": 0.5732,
+      "step": 54430
+    },
+    {
+      "epoch": 2.7038839773517434,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005836932551902254,
+      "loss": 0.5686,
+      "step": 54440
+    },
+    {
+      "epoch": 2.7043806496473626,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.000583653521406576,
+      "loss": 0.5449,
+      "step": 54450
+    },
+    {
+      "epoch": 2.7048773219429822,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005836137876229265,
+      "loss": 0.5793,
+      "step": 54460
+    },
+    {
+      "epoch": 2.7053739942386015,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005835740538392768,
+      "loss": 0.5356,
+      "step": 54470
+    },
+    {
+      "epoch": 2.7058706665342207,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005835343200556273,
+      "loss": 0.5417,
+      "step": 54480
+    },
+    {
+      "epoch": 2.70636733882984,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005834945862719778,
+      "loss": 0.5341,
+      "step": 54490
+    },
+    {
+      "epoch": 2.706864011125459,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005834548524883282,
+      "loss": 0.5406,
+      "step": 54500
+    },
+    {
+      "epoch": 2.707360683421079,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005834151187046787,
+      "loss": 0.586,
+      "step": 54510
+    },
+    {
+      "epoch": 2.707857355716698,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005833753849210291,
+      "loss": 0.5395,
+      "step": 54520
+    },
+    {
+      "epoch": 2.7083540280123177,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005833356511373796,
+      "loss": 0.5561,
+      "step": 54530
+    },
+    {
+      "epoch": 2.708850700307937,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.00058329591735373,
+      "loss": 0.5365,
+      "step": 54540
+    },
+    {
+      "epoch": 2.709347372603556,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005832561835700805,
+      "loss": 0.5619,
+      "step": 54550
+    },
+    {
+      "epoch": 2.7098440448991754,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.000583216449786431,
+      "loss": 0.5714,
+      "step": 54560
+    },
+    {
+      "epoch": 2.7103407171947946,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005831767160027814,
+      "loss": 0.5312,
+      "step": 54570
+    },
+    {
+      "epoch": 2.7108373894904143,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005831369822191318,
+      "loss": 0.5659,
+      "step": 54580
+    },
+    {
+      "epoch": 2.7113340617860335,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005830972484354823,
+      "loss": 0.5533,
+      "step": 54590
+    },
+    {
+      "epoch": 2.7118307340816528,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005830575146518328,
+      "loss": 0.5633,
+      "step": 54600
+    },
+    {
+      "epoch": 2.7123274063772724,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005830177808681832,
+      "loss": 0.5577,
+      "step": 54610
+    },
+    {
+      "epoch": 2.7128240786728917,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005829780470845337,
+      "loss": 0.572,
+      "step": 54620
+    },
+    {
+      "epoch": 2.713320750968511,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005829383133008841,
+      "loss": 0.5672,
+      "step": 54630
+    },
+    {
+      "epoch": 2.71381742326413,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005828985795172345,
+      "loss": 0.5401,
+      "step": 54640
+    },
+    {
+      "epoch": 2.71431409555975,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005828588457335851,
+      "loss": 0.5604,
+      "step": 54650
+    },
+    {
+      "epoch": 2.714810767855369,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005828191119499355,
+      "loss": 0.5605,
+      "step": 54660
+    },
+    {
+      "epoch": 2.7153074401509882,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005827793781662859,
+      "loss": 0.5624,
+      "step": 54670
+    },
+    {
+      "epoch": 2.715804112446608,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005827396443826364,
+      "loss": 0.5725,
+      "step": 54680
+    },
+    {
+      "epoch": 2.716300784742227,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005826999105989868,
+      "loss": 0.5552,
+      "step": 54690
+    },
+    {
+      "epoch": 2.7167974570378464,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005826601768153373,
+      "loss": 0.5755,
+      "step": 54700
+    },
+    {
+      "epoch": 2.7172941293334656,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005826204430316877,
+      "loss": 0.5701,
+      "step": 54710
+    },
+    {
+      "epoch": 2.7177908016290853,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005825807092480382,
+      "loss": 0.5731,
+      "step": 54720
+    },
+    {
+      "epoch": 2.7182874739247045,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005825409754643886,
+      "loss": 0.5941,
+      "step": 54730
+    },
+    {
+      "epoch": 2.7187841462203237,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.000582501241680739,
+      "loss": 0.5641,
+      "step": 54740
+    },
+    {
+      "epoch": 2.7192808185159434,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005824615078970896,
+      "loss": 0.5417,
+      "step": 54750
+    },
+    {
+      "epoch": 2.7197774908115626,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00058242177411344,
+      "loss": 0.5744,
+      "step": 54760
+    },
+    {
+      "epoch": 2.720274163107182,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005823820403297904,
+      "loss": 0.571,
+      "step": 54770
+    },
+    {
+      "epoch": 2.720770835402801,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005823423065461409,
+      "loss": 0.5549,
+      "step": 54780
+    },
+    {
+      "epoch": 2.7212675076984207,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005823025727624913,
+      "loss": 0.5888,
+      "step": 54790
+    },
+    {
+      "epoch": 2.72176417999404,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005822628389788418,
+      "loss": 0.5551,
+      "step": 54800
+    },
+    {
+      "epoch": 2.722260852289659,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0005822231051951923,
+      "loss": 0.5548,
+      "step": 54810
+    },
+    {
+      "epoch": 2.722757524585279,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005821833714115427,
+      "loss": 0.5363,
+      "step": 54820
+    },
+    {
+      "epoch": 2.723254196880898,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005821436376278931,
+      "loss": 0.5589,
+      "step": 54830
+    },
+    {
+      "epoch": 2.7237508691765173,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0005821039038442436,
+      "loss": 0.5889,
+      "step": 54840
+    },
+    {
+      "epoch": 2.7242475414721365,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005820641700605941,
+      "loss": 0.5374,
+      "step": 54850
+    },
+    {
+      "epoch": 2.7247442137677558,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005820244362769445,
+      "loss": 0.5606,
+      "step": 54860
+    },
+    {
+      "epoch": 2.7252408860633754,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.000581984702493295,
+      "loss": 0.5569,
+      "step": 54870
+    },
+    {
+      "epoch": 2.7257375583589947,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0005819449687096454,
+      "loss": 0.5761,
+      "step": 54880
+    },
+    {
+      "epoch": 2.7262342306546143,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005819052349259958,
+      "loss": 0.5551,
+      "step": 54890
+    },
+    {
+      "epoch": 2.7267309029502336,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005818655011423464,
+      "loss": 0.5721,
+      "step": 54900
+    },
+    {
+      "epoch": 2.727227575245853,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005818257673586968,
+      "loss": 0.5725,
+      "step": 54910
+    },
+    {
+      "epoch": 2.727724247541472,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005817860335750472,
+      "loss": 0.5593,
+      "step": 54920
+    },
+    {
+      "epoch": 2.7282209198370913,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005817462997913976,
+      "loss": 0.5551,
+      "step": 54930
+    },
+    {
+      "epoch": 2.728717592132711,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005817065660077481,
+      "loss": 0.5658,
+      "step": 54940
+    },
+    {
+      "epoch": 2.72921426442833,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005816668322240986,
+      "loss": 0.5853,
+      "step": 54950
+    },
+    {
+      "epoch": 2.7297109367239494,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.000581627098440449,
+      "loss": 0.5695,
+      "step": 54960
+    },
+    {
+      "epoch": 2.730207609019569,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005815873646567995,
+      "loss": 0.5309,
+      "step": 54970
+    },
+    {
+      "epoch": 2.7307042813151883,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005815476308731499,
+      "loss": 0.5484,
+      "step": 54980
+    },
+    {
+      "epoch": 2.7312009536108075,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005815078970895003,
+      "loss": 0.5488,
+      "step": 54990
+    },
+    {
+      "epoch": 2.7316976259064267,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005814681633058509,
+      "loss": 0.5306,
+      "step": 55000
+    },
+    {
+      "epoch": 2.7321942982020464,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005814284295222013,
+      "loss": 0.5647,
+      "step": 55010
+    },
+    {
+      "epoch": 2.7326909704976656,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005813886957385517,
+      "loss": 0.5721,
+      "step": 55020
+    },
+    {
+      "epoch": 2.733187642793285,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005813489619549022,
+      "loss": 0.5647,
+      "step": 55030
+    },
+    {
+      "epoch": 2.7336843150889045,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005813092281712526,
+      "loss": 0.5479,
+      "step": 55040
+    },
+    {
+      "epoch": 2.7341809873845238,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.000581269494387603,
+      "loss": 0.5689,
+      "step": 55050
+    },
+    {
+      "epoch": 2.734677659680143,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005812297606039536,
+      "loss": 0.535,
+      "step": 55060
+    },
+    {
+      "epoch": 2.735174331975762,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.000581190026820304,
+      "loss": 0.5574,
+      "step": 55070
+    },
+    {
+      "epoch": 2.735671004271382,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005811502930366544,
+      "loss": 0.5488,
+      "step": 55080
+    },
+    {
+      "epoch": 2.736167676567001,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005811105592530049,
+      "loss": 0.5373,
+      "step": 55090
+    },
+    {
+      "epoch": 2.7366643488626203,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005810708254693554,
+      "loss": 0.5725,
+      "step": 55100
+    },
+    {
+      "epoch": 2.73716102115824,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005810310916857059,
+      "loss": 0.5674,
+      "step": 55110
+    },
+    {
+      "epoch": 2.7376576934538592,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005809913579020562,
+      "loss": 0.5623,
+      "step": 55120
+    },
+    {
+      "epoch": 2.7381543657494785,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005809516241184067,
+      "loss": 0.5248,
+      "step": 55130
+    },
+    {
+      "epoch": 2.7386510380450977,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005809118903347572,
+      "loss": 0.5304,
+      "step": 55140
+    },
+    {
+      "epoch": 2.7391477103407174,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005808721565511075,
+      "loss": 0.5724,
+      "step": 55150
+    },
+    {
+      "epoch": 2.7396443826363366,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005808324227674581,
+      "loss": 0.5655,
+      "step": 55160
+    },
+    {
+      "epoch": 2.740141054931956,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005807926889838085,
+      "loss": 0.5197,
+      "step": 55170
+    },
+    {
+      "epoch": 2.7406377272275755,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005807529552001589,
+      "loss": 0.551,
+      "step": 55180
+    },
+    {
+      "epoch": 2.7411343995231947,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005807132214165094,
+      "loss": 0.5737,
+      "step": 55190
+    },
+    {
+      "epoch": 2.741631071818814,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005806734876328598,
+      "loss": 0.5701,
+      "step": 55200
+    },
+    {
+      "epoch": 2.742127744114433,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005806337538492103,
+      "loss": 0.5167,
+      "step": 55210
+    },
+    {
+      "epoch": 2.7426244164100524,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005805940200655608,
+      "loss": 0.5687,
+      "step": 55220
+    },
+    {
+      "epoch": 2.743121088705672,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005805542862819112,
+      "loss": 0.5623,
+      "step": 55230
+    },
+    {
+      "epoch": 2.7436177610012913,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005805145524982616,
+      "loss": 0.5348,
+      "step": 55240
+    },
+    {
+      "epoch": 2.744114433296911,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005804748187146121,
+      "loss": 0.549,
+      "step": 55250
+    },
+    {
+      "epoch": 2.74461110559253,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005804350849309626,
+      "loss": 0.5745,
+      "step": 55260
+    },
+    {
+      "epoch": 2.7451077778881494,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005803953511473131,
+      "loss": 0.5593,
+      "step": 55270
+    },
+    {
+      "epoch": 2.7456044501837686,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005803556173636635,
+      "loss": 0.5544,
+      "step": 55280
+    },
+    {
+      "epoch": 2.746101122479388,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005803158835800139,
+      "loss": 0.5419,
+      "step": 55290
+    },
+    {
+      "epoch": 2.7465977947750075,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005802761497963645,
+      "loss": 0.6107,
+      "step": 55300
+    },
+    {
+      "epoch": 2.7470944670706268,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005802364160127148,
+      "loss": 0.5794,
+      "step": 55310
+    },
+    {
+      "epoch": 2.747591139366246,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005801966822290653,
+      "loss": 0.5429,
+      "step": 55320
+    },
+    {
+      "epoch": 2.7480878116618657,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005801569484454158,
+      "loss": 0.5616,
+      "step": 55330
+    },
+    {
+      "epoch": 2.748584483957485,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0005801172146617661,
+      "loss": 0.5419,
+      "step": 55340
+    },
+    {
+      "epoch": 2.749081156253104,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005800774808781166,
+      "loss": 0.5558,
+      "step": 55350
+    },
+    {
+      "epoch": 2.7495778285487233,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005800377470944672,
+      "loss": 0.5258,
+      "step": 55360
+    },
+    {
+      "epoch": 2.750074500844343,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005799980133108175,
+      "loss": 0.5429,
+      "step": 55370
+    },
+    {
+      "epoch": 2.7505711731399622,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.000579958279527168,
+      "loss": 0.5463,
+      "step": 55380
+    },
+    {
+      "epoch": 2.7510678454355815,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005799185457435184,
+      "loss": 0.5475,
+      "step": 55390
+    },
+    {
+      "epoch": 2.751564517731201,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.000579878811959869,
+      "loss": 0.5423,
+      "step": 55400
+    },
+    {
+      "epoch": 2.7520611900268204,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005798390781762194,
+      "loss": 0.5488,
+      "step": 55410
+    },
+    {
+      "epoch": 2.7525578623224396,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005797993443925698,
+      "loss": 0.5434,
+      "step": 55420
+    },
+    {
+      "epoch": 2.753054534618059,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005797596106089203,
+      "loss": 0.5555,
+      "step": 55430
+    },
+    {
+      "epoch": 2.7535512069136785,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005797198768252707,
+      "loss": 0.5677,
+      "step": 55440
+    },
+    {
+      "epoch": 2.7540478792092977,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005796801430416211,
+      "loss": 0.5583,
+      "step": 55450
+    },
+    {
+      "epoch": 2.754544551504917,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005796404092579717,
+      "loss": 0.5538,
+      "step": 55460
+    },
+    {
+      "epoch": 2.7550412238005366,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005796006754743221,
+      "loss": 0.5444,
+      "step": 55470
+    },
+    {
+      "epoch": 2.755537896096156,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005795609416906725,
+      "loss": 0.5492,
+      "step": 55480
+    },
+    {
+      "epoch": 2.756034568391775,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.000579521207907023,
+      "loss": 0.5969,
+      "step": 55490
+    },
+    {
+      "epoch": 2.7565312406873943,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005794814741233733,
+      "loss": 0.5599,
+      "step": 55500
+    },
+    {
+      "epoch": 2.757027912983014,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005794417403397239,
+      "loss": 0.5281,
+      "step": 55510
+    },
+    {
+      "epoch": 2.757524585278633,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005794020065560744,
+      "loss": 0.565,
+      "step": 55520
+    },
+    {
+      "epoch": 2.7580212575742524,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005793622727724247,
+      "loss": 0.5562,
+      "step": 55530
+    },
+    {
+      "epoch": 2.758517929869872,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005793225389887752,
+      "loss": 0.5884,
+      "step": 55540
+    },
+    {
+      "epoch": 2.7590146021654913,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005792828052051257,
+      "loss": 0.5439,
+      "step": 55550
+    },
+    {
+      "epoch": 2.7595112744611106,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005792430714214762,
+      "loss": 0.5555,
+      "step": 55560
+    },
+    {
+      "epoch": 2.76000794675673,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005792033376378266,
+      "loss": 0.5547,
+      "step": 55570
+    },
+    {
+      "epoch": 2.760504619052349,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.000579163603854177,
+      "loss": 0.5592,
+      "step": 55580
+    },
+    {
+      "epoch": 2.7610012913479687,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005791238700705275,
+      "loss": 0.5555,
+      "step": 55590
+    },
+    {
+      "epoch": 2.761497963643588,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005790841362868779,
+      "loss": 0.5415,
+      "step": 55600
+    },
+    {
+      "epoch": 2.7619946359392076,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005790444025032284,
+      "loss": 0.5891,
+      "step": 55610
+    },
+    {
+      "epoch": 2.762491308234827,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005790046687195789,
+      "loss": 0.5656,
+      "step": 55620
+    },
+    {
+      "epoch": 2.762987980530446,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005789649349359293,
+      "loss": 0.5647,
+      "step": 55630
+    },
+    {
+      "epoch": 2.7634846528260653,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005789252011522797,
+      "loss": 0.5689,
+      "step": 55640
+    },
+    {
+      "epoch": 2.7639813251216845,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005788854673686302,
+      "loss": 0.5426,
+      "step": 55650
+    },
+    {
+      "epoch": 2.764477997417304,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005788457335849807,
+      "loss": 0.5417,
+      "step": 55660
+    },
+    {
+      "epoch": 2.7649746697129234,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005788059998013311,
+      "loss": 0.5529,
+      "step": 55670
+    },
+    {
+      "epoch": 2.7654713420085426,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005787662660176816,
+      "loss": 0.5507,
+      "step": 55680
+    },
+    {
+      "epoch": 2.7659680143041623,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.000578726532234032,
+      "loss": 0.5881,
+      "step": 55690
+    },
+    {
+      "epoch": 2.7664646865997815,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005786867984503824,
+      "loss": 0.5548,
+      "step": 55700
+    },
+    {
+      "epoch": 2.7669613588954007,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.000578647064666733,
+      "loss": 0.5443,
+      "step": 55710
+    },
+    {
+      "epoch": 2.76745803119102,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005786073308830834,
+      "loss": 0.5477,
+      "step": 55720
+    },
+    {
+      "epoch": 2.7679547034866396,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005785675970994338,
+      "loss": 0.5643,
+      "step": 55730
+    },
+    {
+      "epoch": 2.768451375782259,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005785278633157843,
+      "loss": 0.5501,
+      "step": 55740
+    },
+    {
+      "epoch": 2.768948048077878,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005784881295321347,
+      "loss": 0.5595,
+      "step": 55750
+    },
+    {
+      "epoch": 2.7694447203734978,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005784483957484852,
+      "loss": 0.5723,
+      "step": 55760
+    },
+    {
+      "epoch": 2.769941392669117,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005784086619648356,
+      "loss": 0.5765,
+      "step": 55770
+    },
+    {
+      "epoch": 2.770438064964736,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005783689281811861,
+      "loss": 0.5515,
+      "step": 55780
+    },
+    {
+      "epoch": 2.7709347372603554,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005783291943975365,
+      "loss": 0.5348,
+      "step": 55790
+    },
+    {
+      "epoch": 2.771431409555975,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005782894606138869,
+      "loss": 0.5507,
+      "step": 55800
+    },
+    {
+      "epoch": 2.7719280818515943,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005782497268302375,
+      "loss": 0.5576,
+      "step": 55810
+    },
+    {
+      "epoch": 2.7724247541472136,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005782099930465879,
+      "loss": 0.5549,
+      "step": 55820
+    },
+    {
+      "epoch": 2.7729214264428332,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005781702592629383,
+      "loss": 0.5687,
+      "step": 55830
+    },
+    {
+      "epoch": 2.7734180987384525,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005781305254792888,
+      "loss": 0.5809,
+      "step": 55840
+    },
+    {
+      "epoch": 2.7739147710340717,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005780907916956392,
+      "loss": 0.5701,
+      "step": 55850
+    },
+    {
+      "epoch": 2.774411443329691,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0005780510579119897,
+      "loss": 0.568,
+      "step": 55860
+    },
+    {
+      "epoch": 2.77490811562531,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005780113241283402,
+      "loss": 0.5715,
+      "step": 55870
+    },
+    {
+      "epoch": 2.77540478792093,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005779715903446906,
+      "loss": 0.5823,
+      "step": 55880
+    },
+    {
+      "epoch": 2.775901460216549,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.000577931856561041,
+      "loss": 0.5197,
+      "step": 55890
+    },
+    {
+      "epoch": 2.7763981325121687,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005778921227773915,
+      "loss": 0.6089,
+      "step": 55900
+    },
+    {
+      "epoch": 2.776894804807788,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.000577852388993742,
+      "loss": 0.5394,
+      "step": 55910
+    },
+    {
+      "epoch": 2.777391477103407,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005778126552100924,
+      "loss": 0.5576,
+      "step": 55920
+    },
+    {
+      "epoch": 2.7778881493990264,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005777729214264429,
+      "loss": 0.5577,
+      "step": 55930
+    },
+    {
+      "epoch": 2.7783848216946456,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005777331876427933,
+      "loss": 0.5627,
+      "step": 55940
+    },
+    {
+      "epoch": 2.7788814939902653,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005776934538591437,
+      "loss": 0.5685,
+      "step": 55950
+    },
+    {
+      "epoch": 2.7793781662858845,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005776537200754943,
+      "loss": 0.5755,
+      "step": 55960
+    },
+    {
+      "epoch": 2.779874838581504,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005776139862918447,
+      "loss": 0.5418,
+      "step": 55970
+    },
+    {
+      "epoch": 2.7803715108771234,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005775742525081951,
+      "loss": 0.5429,
+      "step": 55980
+    },
+    {
+      "epoch": 2.7808681831727426,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005775345187245455,
+      "loss": 0.5879,
+      "step": 55990
+    },
+    {
+      "epoch": 2.781364855468362,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.000577494784940896,
+      "loss": 0.5522,
+      "step": 56000
+    },
+    {
+      "epoch": 2.781861527763981,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005774550511572466,
+      "loss": 0.5791,
+      "step": 56010
+    },
+    {
+      "epoch": 2.7823582000596008,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005774153173735969,
+      "loss": 0.5753,
+      "step": 56020
+    },
+    {
+      "epoch": 2.78285487235522,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005773755835899474,
+      "loss": 0.5485,
+      "step": 56030
+    },
+    {
+      "epoch": 2.7833515446508392,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005773358498062978,
+      "loss": 0.5378,
+      "step": 56040
+    },
+    {
+      "epoch": 2.783848216946459,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005772961160226482,
+      "loss": 0.5962,
+      "step": 56050
+    },
+    {
+      "epoch": 2.784344889242078,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005772563822389988,
+      "loss": 0.5633,
+      "step": 56060
+    },
+    {
+      "epoch": 2.7848415615376974,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005772166484553492,
+      "loss": 0.5915,
+      "step": 56070
+    },
+    {
+      "epoch": 2.7853382338333166,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005771769146716996,
+      "loss": 0.565,
+      "step": 56080
+    },
+    {
+      "epoch": 2.7858349061289363,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005771371808880501,
+      "loss": 0.544,
+      "step": 56090
+    },
+    {
+      "epoch": 2.7863315784245555,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005770974471044005,
+      "loss": 0.5503,
+      "step": 56100
+    },
+    {
+      "epoch": 2.7868282507201747,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.000577057713320751,
+      "loss": 0.5926,
+      "step": 56110
+    },
+    {
+      "epoch": 2.7873249230157944,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005770179795371015,
+      "loss": 0.5907,
+      "step": 56120
+    },
+    {
+      "epoch": 2.7878215953114136,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005769782457534519,
+      "loss": 0.542,
+      "step": 56130
+    },
+    {
+      "epoch": 2.788318267607033,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005769385119698023,
+      "loss": 0.5577,
+      "step": 56140
+    },
+    {
+      "epoch": 2.788814939902652,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005768987781861528,
+      "loss": 0.5581,
+      "step": 56150
+    },
+    {
+      "epoch": 2.7893116121982717,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005768590444025033,
+      "loss": 0.5403,
+      "step": 56160
+    },
+    {
+      "epoch": 2.789808284493891,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005768193106188538,
+      "loss": 0.5795,
+      "step": 56170
+    },
+    {
+      "epoch": 2.79030495678951,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005767795768352041,
+      "loss": 0.529,
+      "step": 56180
+    },
+    {
+      "epoch": 2.79080162908513,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005767398430515546,
+      "loss": 0.5354,
+      "step": 56190
+    },
+    {
+      "epoch": 2.791298301380749,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005767001092679051,
+      "loss": 0.5647,
+      "step": 56200
+    },
+    {
+      "epoch": 2.7917949736763683,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005766603754842554,
+      "loss": 0.5636,
+      "step": 56210
+    },
+    {
+      "epoch": 2.7922916459719875,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.000576620641700606,
+      "loss": 0.5823,
+      "step": 56220
+    },
+    {
+      "epoch": 2.7927883182676068,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005765809079169565,
+      "loss": 0.5592,
+      "step": 56230
+    },
+    {
+      "epoch": 2.7932849905632264,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005765411741333068,
+      "loss": 0.5779,
+      "step": 56240
+    },
+    {
+      "epoch": 2.7937816628588457,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005765014403496573,
+      "loss": 0.5916,
+      "step": 56250
+    },
+    {
+      "epoch": 2.7942783351544653,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0005764617065660078,
+      "loss": 0.567,
+      "step": 56260
+    },
+    {
+      "epoch": 2.7947750074500846,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005764219727823582,
+      "loss": 0.5717,
+      "step": 56270
+    },
+    {
+      "epoch": 2.795271679745704,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005763822389987087,
+      "loss": 0.5514,
+      "step": 56280
+    },
+    {
+      "epoch": 2.795768352041323,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005763425052150591,
+      "loss": 0.5704,
+      "step": 56290
+    },
+    {
+      "epoch": 2.7962650243369422,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005763027714314096,
+      "loss": 0.5569,
+      "step": 56300
+    },
+    {
+      "epoch": 2.796761696632562,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00057626303764776,
+      "loss": 0.5818,
+      "step": 56310
+    },
+    {
+      "epoch": 2.797258368928181,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005762233038641105,
+      "loss": 0.5351,
+      "step": 56320
+    },
+    {
+      "epoch": 2.797755041223801,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.000576183570080461,
+      "loss": 0.5564,
+      "step": 56330
+    },
+    {
+      "epoch": 2.79825171351942,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0005761438362968114,
+      "loss": 0.577,
+      "step": 56340
+    },
+    {
+      "epoch": 2.7987483858150393,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005761041025131618,
+      "loss": 0.5136,
+      "step": 56350
+    },
+    {
+      "epoch": 2.7992450581106585,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005760643687295124,
+      "loss": 0.5576,
+      "step": 56360
+    },
+    {
+      "epoch": 2.7997417304062777,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005760246349458627,
+      "loss": 0.5479,
+      "step": 56370
+    },
+    {
+      "epoch": 2.8002384027018974,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005759849011622132,
+      "loss": 0.5534,
+      "step": 56380
+    },
+    {
+      "epoch": 2.8007350749975166,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005759451673785637,
+      "loss": 0.5432,
+      "step": 56390
+    },
+    {
+      "epoch": 2.801231747293136,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.000575905433594914,
+      "loss": 0.5506,
+      "step": 56400
+    },
+    {
+      "epoch": 2.8017284195887555,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005758656998112646,
+      "loss": 0.5494,
+      "step": 56410
+    },
+    {
+      "epoch": 2.8022250918843747,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005758259660276151,
+      "loss": 0.5985,
+      "step": 56420
+    },
+    {
+      "epoch": 2.802721764179994,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005757862322439654,
+      "loss": 0.5345,
+      "step": 56430
+    },
+    {
+      "epoch": 2.803218436475613,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005757464984603159,
+      "loss": 0.5927,
+      "step": 56440
+    },
+    {
+      "epoch": 2.803715108771233,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005757067646766663,
+      "loss": 0.5388,
+      "step": 56450
+    },
+    {
+      "epoch": 2.804211781066852,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005756670308930169,
+      "loss": 0.5468,
+      "step": 56460
+    },
+    {
+      "epoch": 2.8047084533624713,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005756272971093673,
+      "loss": 0.592,
+      "step": 56470
+    },
+    {
+      "epoch": 2.805205125658091,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005755875633257177,
+      "loss": 0.5579,
+      "step": 56480
+    },
+    {
+      "epoch": 2.80570179795371,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005755478295420682,
+      "loss": 0.556,
+      "step": 56490
+    },
+    {
+      "epoch": 2.8061984702493294,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005755080957584186,
+      "loss": 0.5605,
+      "step": 56500
+    },
+    {
+      "epoch": 2.8066951425449487,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.000575468361974769,
+      "loss": 0.5704,
+      "step": 56510
+    },
+    {
+      "epoch": 2.8071918148405683,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005754286281911196,
+      "loss": 0.538,
+      "step": 56520
+    },
+    {
+      "epoch": 2.8076884871361876,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00057538889440747,
+      "loss": 0.5666,
+      "step": 56530
+    },
+    {
+      "epoch": 2.808185159431807,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005753491606238204,
+      "loss": 0.579,
+      "step": 56540
+    },
+    {
+      "epoch": 2.8086818317274265,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005753094268401709,
+      "loss": 0.5407,
+      "step": 56550
+    },
+    {
+      "epoch": 2.8091785040230457,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005752696930565213,
+      "loss": 0.55,
+      "step": 56560
+    },
+    {
+      "epoch": 2.809675176318665,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005752299592728718,
+      "loss": 0.5505,
+      "step": 56570
+    },
+    {
+      "epoch": 2.810171848614284,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005751902254892223,
+      "loss": 0.5341,
+      "step": 56580
+    },
+    {
+      "epoch": 2.8106685209099034,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005751504917055726,
+      "loss": 0.5854,
+      "step": 56590
+    },
+    {
+      "epoch": 2.811165193205523,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005751107579219231,
+      "loss": 0.6014,
+      "step": 56600
+    },
+    {
+      "epoch": 2.8116618655011423,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005750710241382737,
+      "loss": 0.5923,
+      "step": 56610
+    },
+    {
+      "epoch": 2.812158537796762,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005750312903546241,
+      "loss": 0.5599,
+      "step": 56620
+    },
+    {
+      "epoch": 2.812655210092381,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005749915565709745,
+      "loss": 0.5852,
+      "step": 56630
+    },
+    {
+      "epoch": 2.8131518823880004,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005749518227873249,
+      "loss": 0.5262,
+      "step": 56640
+    },
+    {
+      "epoch": 2.8136485546836196,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005749120890036754,
+      "loss": 0.5398,
+      "step": 56650
+    },
+    {
+      "epoch": 2.814145226979239,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005748723552200258,
+      "loss": 0.5394,
+      "step": 56660
+    },
+    {
+      "epoch": 2.8146418992748585,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005748326214363763,
+      "loss": 0.5383,
+      "step": 56670
+    },
+    {
+      "epoch": 2.8151385715704778,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005747928876527268,
+      "loss": 0.5352,
+      "step": 56680
+    },
+    {
+      "epoch": 2.8156352438660974,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005747531538690772,
+      "loss": 0.5325,
+      "step": 56690
+    },
+    {
+      "epoch": 2.8161319161617167,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005747134200854276,
+      "loss": 0.5311,
+      "step": 56700
+    },
+    {
+      "epoch": 2.816628588457336,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005746736863017781,
+      "loss": 0.5797,
+      "step": 56710
+    },
+    {
+      "epoch": 2.817125260752955,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005746339525181286,
+      "loss": 0.5722,
+      "step": 56720
+    },
+    {
+      "epoch": 2.8176219330485743,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.000574594218734479,
+      "loss": 0.5606,
+      "step": 56730
+    },
+    {
+      "epoch": 2.818118605344194,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005745544849508295,
+      "loss": 0.5685,
+      "step": 56740
+    },
+    {
+      "epoch": 2.8186152776398132,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005745147511671799,
+      "loss": 0.5442,
+      "step": 56750
+    },
+    {
+      "epoch": 2.8191119499354325,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005744750173835303,
+      "loss": 0.5349,
+      "step": 56760
+    },
+    {
+      "epoch": 2.819608622231052,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005744352835998809,
+      "loss": 0.5479,
+      "step": 56770
+    },
+    {
+      "epoch": 2.8201052945266714,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005743955498162313,
+      "loss": 0.5517,
+      "step": 56780
+    },
+    {
+      "epoch": 2.8206019668222906,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005743558160325817,
+      "loss": 0.5598,
+      "step": 56790
+    },
+    {
+      "epoch": 2.82109863911791,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005743160822489322,
+      "loss": 0.546,
+      "step": 56800
+    },
+    {
+      "epoch": 2.8215953114135295,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005742763484652826,
+      "loss": 0.5627,
+      "step": 56810
+    },
+    {
+      "epoch": 2.8220919837091487,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005742366146816331,
+      "loss": 0.5589,
+      "step": 56820
+    },
+    {
+      "epoch": 2.822588656004768,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005741968808979836,
+      "loss": 0.5491,
+      "step": 56830
+    },
+    {
+      "epoch": 2.8230853283003876,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.000574157147114334,
+      "loss": 0.5563,
+      "step": 56840
+    },
+    {
+      "epoch": 2.823582000596007,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005741174133306844,
+      "loss": 0.5788,
+      "step": 56850
+    },
+    {
+      "epoch": 2.824078672891626,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005740776795470348,
+      "loss": 0.5798,
+      "step": 56860
+    },
+    {
+      "epoch": 2.8245753451872453,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005740379457633854,
+      "loss": 0.5597,
+      "step": 56870
+    },
+    {
+      "epoch": 2.825072017482865,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005739982119797358,
+      "loss": 0.5476,
+      "step": 56880
+    },
+    {
+      "epoch": 2.825568689778484,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005739584781960862,
+      "loss": 0.5343,
+      "step": 56890
+    },
+    {
+      "epoch": 2.8260653620741034,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005739187444124367,
+      "loss": 0.5601,
+      "step": 56900
+    },
+    {
+      "epoch": 2.826562034369723,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005738790106287871,
+      "loss": 0.5578,
+      "step": 56910
+    },
+    {
+      "epoch": 2.8270587066653423,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005738392768451376,
+      "loss": 0.5794,
+      "step": 56920
+    },
+    {
+      "epoch": 2.8275553789609615,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005737995430614881,
+      "loss": 0.5499,
+      "step": 56930
+    },
+    {
+      "epoch": 2.8280520512565808,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005737598092778385,
+      "loss": 0.5601,
+      "step": 56940
+    },
+    {
+      "epoch": 2.8285487235522,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005737200754941889,
+      "loss": 0.5375,
+      "step": 56950
+    },
+    {
+      "epoch": 2.8290453958478197,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005736803417105394,
+      "loss": 0.5793,
+      "step": 56960
+    },
+    {
+      "epoch": 2.829542068143439,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005736406079268899,
+      "loss": 0.5278,
+      "step": 56970
+    },
+    {
+      "epoch": 2.8300387404390586,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005736008741432403,
+      "loss": 0.5501,
+      "step": 56980
+    },
+    {
+      "epoch": 2.830535412734678,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005735611403595908,
+      "loss": 0.5422,
+      "step": 56990
+    },
+    {
+      "epoch": 2.831032085030297,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005735214065759412,
+      "loss": 0.5629,
+      "step": 57000
+    },
+    {
+      "epoch": 2.8315287573259162,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005734816727922916,
+      "loss": 0.5634,
+      "step": 57010
+    },
+    {
+      "epoch": 2.8320254296215355,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005734419390086422,
+      "loss": 0.5477,
+      "step": 57020
+    },
+    {
+      "epoch": 2.832522101917155,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005734022052249926,
+      "loss": 0.5744,
+      "step": 57030
+    },
+    {
+      "epoch": 2.8330187742127744,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.000573362471441343,
+      "loss": 0.5712,
+      "step": 57040
+    },
+    {
+      "epoch": 2.8335154465083936,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005733227376576934,
+      "loss": 0.513,
+      "step": 57050
+    },
+    {
+      "epoch": 2.8340121188040133,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005732830038740439,
+      "loss": 0.5702,
+      "step": 57060
+    },
+    {
+      "epoch": 2.8345087910996325,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005732432700903945,
+      "loss": 0.5636,
+      "step": 57070
+    },
+    {
+      "epoch": 2.8350054633952517,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005732035363067448,
+      "loss": 0.5351,
+      "step": 57080
+    },
+    {
+      "epoch": 2.835502135690871,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005731638025230953,
+      "loss": 0.5575,
+      "step": 57090
+    },
+    {
+      "epoch": 2.8359988079864906,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005731240687394458,
+      "loss": 0.5511,
+      "step": 57100
+    },
+    {
+      "epoch": 2.83649548028211,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005730843349557961,
+      "loss": 0.564,
+      "step": 57110
+    },
+    {
+      "epoch": 2.836992152577729,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005730446011721467,
+      "loss": 0.5408,
+      "step": 57120
+    },
+    {
+      "epoch": 2.8374888248733487,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005730048673884971,
+      "loss": 0.5589,
+      "step": 57130
+    },
+    {
+      "epoch": 2.837985497168968,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005729651336048475,
+      "loss": 0.5682,
+      "step": 57140
+    },
+    {
+      "epoch": 2.838482169464587,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.000572925399821198,
+      "loss": 0.5336,
+      "step": 57150
+    },
+    {
+      "epoch": 2.8389788417602064,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005728856660375484,
+      "loss": 0.5409,
+      "step": 57160
+    },
+    {
+      "epoch": 2.839475514055826,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005728459322538989,
+      "loss": 0.5335,
+      "step": 57170
+    },
+    {
+      "epoch": 2.8399721863514453,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005728061984702494,
+      "loss": 0.5436,
+      "step": 57180
+    },
+    {
+      "epoch": 2.8404688586470646,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005727664646865998,
+      "loss": 0.5297,
+      "step": 57190
+    },
+    {
+      "epoch": 2.8409655309426842,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005727267309029503,
+      "loss": 0.5597,
+      "step": 57200
+    },
+    {
+      "epoch": 2.8414622032383035,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005726869971193007,
+      "loss": 0.5425,
+      "step": 57210
+    },
+    {
+      "epoch": 2.8419588755339227,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005726472633356512,
+      "loss": 0.5408,
+      "step": 57220
+    },
+    {
+      "epoch": 2.842455547829542,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005726075295520017,
+      "loss": 0.5352,
+      "step": 57230
+    },
+    {
+      "epoch": 2.8429522201251616,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000572567795768352,
+      "loss": 0.574,
+      "step": 57240
+    },
+    {
+      "epoch": 2.843448892420781,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005725280619847025,
+      "loss": 0.5742,
+      "step": 57250
+    },
+    {
+      "epoch": 2.8439455647164,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.000572488328201053,
+      "loss": 0.5515,
+      "step": 57260
+    },
+    {
+      "epoch": 2.8444422370120197,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005724485944174034,
+      "loss": 0.593,
+      "step": 57270
+    },
+    {
+      "epoch": 2.844938909307639,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005724088606337539,
+      "loss": 0.528,
+      "step": 57280
+    },
+    {
+      "epoch": 2.845435581603258,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005723691268501044,
+      "loss": 0.551,
+      "step": 57290
+    },
+    {
+      "epoch": 2.8459322538988774,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005723293930664547,
+      "loss": 0.5678,
+      "step": 57300
+    },
+    {
+      "epoch": 2.8464289261944966,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005722896592828052,
+      "loss": 0.5736,
+      "step": 57310
+    },
+    {
+      "epoch": 2.8469255984901163,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005722499254991557,
+      "loss": 0.5452,
+      "step": 57320
+    },
+    {
+      "epoch": 2.8474222707857355,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005722101917155061,
+      "loss": 0.5354,
+      "step": 57330
+    },
+    {
+      "epoch": 2.847918943081355,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005721704579318566,
+      "loss": 0.5659,
+      "step": 57340
+    },
+    {
+      "epoch": 2.8484156153769744,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.000572130724148207,
+      "loss": 0.532,
+      "step": 57350
+    },
+    {
+      "epoch": 2.8489122876725936,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005720909903645575,
+      "loss": 0.5466,
+      "step": 57360
+    },
+    {
+      "epoch": 2.849408959968213,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.000572051256580908,
+      "loss": 0.5825,
+      "step": 57370
+    },
+    {
+      "epoch": 2.849905632263832,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005720115227972584,
+      "loss": 0.5741,
+      "step": 57380
+    },
+    {
+      "epoch": 2.8504023045594518,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005719717890136089,
+      "loss": 0.5571,
+      "step": 57390
+    },
+    {
+      "epoch": 2.850898976855071,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005719320552299593,
+      "loss": 0.5709,
+      "step": 57400
+    },
+    {
+      "epoch": 2.85139564915069,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005718923214463097,
+      "loss": 0.569,
+      "step": 57410
+    },
+    {
+      "epoch": 2.85189232144631,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005718525876626603,
+      "loss": 0.5289,
+      "step": 57420
+    },
+    {
+      "epoch": 2.852388993741929,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005718128538790107,
+      "loss": 0.5553,
+      "step": 57430
+    },
+    {
+      "epoch": 2.8528856660375483,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005717731200953611,
+      "loss": 0.539,
+      "step": 57440
+    },
+    {
+      "epoch": 2.8533823383331676,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005717333863117116,
+      "loss": 0.5497,
+      "step": 57450
+    },
+    {
+      "epoch": 2.8538790106287872,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005716936525280619,
+      "loss": 0.5593,
+      "step": 57460
+    },
+    {
+      "epoch": 2.8543756829244065,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005716539187444125,
+      "loss": 0.5529,
+      "step": 57470
+    },
+    {
+      "epoch": 2.8548723552200257,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.000571614184960763,
+      "loss": 0.557,
+      "step": 57480
+    },
+    {
+      "epoch": 2.8553690275156454,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005715744511771133,
+      "loss": 0.5768,
+      "step": 57490
+    },
+    {
+      "epoch": 2.8558656998112646,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005715347173934638,
+      "loss": 0.5636,
+      "step": 57500
+    },
+    {
+      "epoch": 2.856362372106884,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005714949836098142,
+      "loss": 0.5743,
+      "step": 57510
+    },
+    {
+      "epoch": 2.856859044402503,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005714552498261648,
+      "loss": 0.559,
+      "step": 57520
+    },
+    {
+      "epoch": 2.8573557166981227,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005714155160425152,
+      "loss": 0.5599,
+      "step": 57530
+    },
+    {
+      "epoch": 2.857852388993742,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005713757822588656,
+      "loss": 0.5598,
+      "step": 57540
+    },
+    {
+      "epoch": 2.858349061289361,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005713360484752161,
+      "loss": 0.5347,
+      "step": 57550
+    },
+    {
+      "epoch": 2.858845733584981,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0005712963146915665,
+      "loss": 0.5707,
+      "step": 57560
+    },
+    {
+      "epoch": 2.8593424058806,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.000571256580907917,
+      "loss": 0.528,
+      "step": 57570
+    },
+    {
+      "epoch": 2.8598390781762193,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005712168471242675,
+      "loss": 0.5671,
+      "step": 57580
+    },
+    {
+      "epoch": 2.8603357504718385,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005711771133406179,
+      "loss": 0.5585,
+      "step": 57590
+    },
+    {
+      "epoch": 2.860832422767458,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005711373795569683,
+      "loss": 0.5669,
+      "step": 57600
+    },
+    {
+      "epoch": 2.8613290950630774,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005710976457733188,
+      "loss": 0.579,
+      "step": 57610
+    },
+    {
+      "epoch": 2.8618257673586966,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005710579119896693,
+      "loss": 0.5508,
+      "step": 57620
+    },
+    {
+      "epoch": 2.8623224396543163,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005710181782060197,
+      "loss": 0.5277,
+      "step": 57630
+    },
+    {
+      "epoch": 2.8628191119499355,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005709784444223702,
+      "loss": 0.5405,
+      "step": 57640
+    },
+    {
+      "epoch": 2.8633157842455548,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005709387106387206,
+      "loss": 0.5483,
+      "step": 57650
+    },
+    {
+      "epoch": 2.863812456541174,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.000570898976855071,
+      "loss": 0.5332,
+      "step": 57660
+    },
+    {
+      "epoch": 2.8643091288367932,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005708592430714216,
+      "loss": 0.5571,
+      "step": 57670
+    },
+    {
+      "epoch": 2.864805801132413,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.000570819509287772,
+      "loss": 0.5255,
+      "step": 57680
+    },
+    {
+      "epoch": 2.865302473428032,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005707797755041224,
+      "loss": 0.5287,
+      "step": 57690
+    },
+    {
+      "epoch": 2.865799145723652,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005707400417204729,
+      "loss": 0.5412,
+      "step": 57700
+    },
+    {
+      "epoch": 2.866295818019271,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005707003079368233,
+      "loss": 0.5548,
+      "step": 57710
+    },
+    {
+      "epoch": 2.8667924903148903,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005706605741531738,
+      "loss": 0.5363,
+      "step": 57720
+    },
+    {
+      "epoch": 2.8672891626105095,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005706208403695242,
+      "loss": 0.5702,
+      "step": 57730
+    },
+    {
+      "epoch": 2.8677858349061287,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005705811065858747,
+      "loss": 0.5496,
+      "step": 57740
+    },
+    {
+      "epoch": 2.8682825072017484,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005705413728022251,
+      "loss": 0.5126,
+      "step": 57750
+    },
+    {
+      "epoch": 2.8687791794973676,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005705016390185755,
+      "loss": 0.5608,
+      "step": 57760
+    },
+    {
+      "epoch": 2.869275851792987,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.000570461905234926,
+      "loss": 0.5755,
+      "step": 57770
+    },
+    {
+      "epoch": 2.8697725240886065,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005704221714512765,
+      "loss": 0.5264,
+      "step": 57780
+    },
+    {
+      "epoch": 2.8702691963842257,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005703824376676269,
+      "loss": 0.5669,
+      "step": 57790
+    },
+    {
+      "epoch": 2.870765868679845,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005703427038839774,
+      "loss": 0.564,
+      "step": 57800
+    },
+    {
+      "epoch": 2.871262540975464,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005703029701003278,
+      "loss": 0.5497,
+      "step": 57810
+    },
+    {
+      "epoch": 2.871759213271084,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005702632363166782,
+      "loss": 0.5965,
+      "step": 57820
+    },
+    {
+      "epoch": 2.872255885566703,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005702235025330288,
+      "loss": 0.5384,
+      "step": 57830
+    },
+    {
+      "epoch": 2.8727525578623223,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005701837687493792,
+      "loss": 0.5137,
+      "step": 57840
+    },
+    {
+      "epoch": 2.873249230157942,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005701440349657296,
+      "loss": 0.562,
+      "step": 57850
+    },
+    {
+      "epoch": 2.873745902453561,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005701043011820801,
+      "loss": 0.5425,
+      "step": 57860
+    },
+    {
+      "epoch": 2.8742425747491804,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005700645673984305,
+      "loss": 0.5806,
+      "step": 57870
+    },
+    {
+      "epoch": 2.8747392470447997,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.000570024833614781,
+      "loss": 0.5594,
+      "step": 57880
+    },
+    {
+      "epoch": 2.8752359193404193,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005699850998311315,
+      "loss": 0.5513,
+      "step": 57890
+    },
+    {
+      "epoch": 2.8757325916360386,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005699453660474819,
+      "loss": 0.5623,
+      "step": 57900
+    },
+    {
+      "epoch": 2.876229263931658,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0005699056322638323,
+      "loss": 0.5909,
+      "step": 57910
+    },
+    {
+      "epoch": 2.8767259362272775,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005698658984801827,
+      "loss": 0.5394,
+      "step": 57920
+    },
+    {
+      "epoch": 2.8772226085228967,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005698261646965333,
+      "loss": 0.556,
+      "step": 57930
+    },
+    {
+      "epoch": 2.877719280818516,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005697864309128838,
+      "loss": 0.576,
+      "step": 57940
+    },
+    {
+      "epoch": 2.878215953114135,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005697466971292341,
+      "loss": 0.555,
+      "step": 57950
+    },
+    {
+      "epoch": 2.8787126254097544,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005697069633455846,
+      "loss": 0.5515,
+      "step": 57960
+    },
+    {
+      "epoch": 2.879209297705374,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005696672295619352,
+      "loss": 0.568,
+      "step": 57970
+    },
+    {
+      "epoch": 2.8797059700009933,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005696274957782855,
+      "loss": 0.5729,
+      "step": 57980
+    },
+    {
+      "epoch": 2.880202642296613,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.000569587761994636,
+      "loss": 0.5552,
+      "step": 57990
+    },
+    {
+      "epoch": 2.880699314592232,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005695480282109864,
+      "loss": 0.5485,
+      "step": 58000
+    },
+    {
+      "epoch": 2.8811959868878514,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0005695082944273368,
+      "loss": 0.545,
+      "step": 58010
+    },
+    {
+      "epoch": 2.8816926591834706,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005694685606436873,
+      "loss": 0.5598,
+      "step": 58020
+    },
+    {
+      "epoch": 2.88218933147909,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005694288268600378,
+      "loss": 0.5485,
+      "step": 58030
+    },
+    {
+      "epoch": 2.8826860037747095,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005693890930763882,
+      "loss": 0.5435,
+      "step": 58040
+    },
+    {
+      "epoch": 2.8831826760703287,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005693493592927387,
+      "loss": 0.5659,
+      "step": 58050
+    },
+    {
+      "epoch": 2.8836793483659484,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005693096255090891,
+      "loss": 0.5266,
+      "step": 58060
+    },
+    {
+      "epoch": 2.8841760206615676,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005692698917254395,
+      "loss": 0.5522,
+      "step": 58070
+    },
+    {
+      "epoch": 2.884672692957187,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005692301579417901,
+      "loss": 0.5434,
+      "step": 58080
+    },
+    {
+      "epoch": 2.885169365252806,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005691904241581405,
+      "loss": 0.5619,
+      "step": 58090
+    },
+    {
+      "epoch": 2.8856660375484253,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.000569150690374491,
+      "loss": 0.569,
+      "step": 58100
+    },
+    {
+      "epoch": 2.886162709844045,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005691109565908413,
+      "loss": 0.5799,
+      "step": 58110
+    },
+    {
+      "epoch": 2.886659382139664,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005690712228071918,
+      "loss": 0.572,
+      "step": 58120
+    },
+    {
+      "epoch": 2.8871560544352834,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005690314890235424,
+      "loss": 0.5379,
+      "step": 58130
+    },
+    {
+      "epoch": 2.887652726730903,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005689917552398927,
+      "loss": 0.5252,
+      "step": 58140
+    },
+    {
+      "epoch": 2.8881493990265223,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005689520214562432,
+      "loss": 0.5574,
+      "step": 58150
+    },
+    {
+      "epoch": 2.8886460713221416,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005689122876725937,
+      "loss": 0.551,
+      "step": 58160
+    },
+    {
+      "epoch": 2.889142743617761,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.000568872553888944,
+      "loss": 0.5691,
+      "step": 58170
+    },
+    {
+      "epoch": 2.8896394159133805,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005688328201052946,
+      "loss": 0.5442,
+      "step": 58180
+    },
+    {
+      "epoch": 2.8901360882089997,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.000568793086321645,
+      "loss": 0.5618,
+      "step": 58190
+    },
+    {
+      "epoch": 2.890632760504619,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005687533525379954,
+      "loss": 0.5685,
+      "step": 58200
+    },
+    {
+      "epoch": 2.8911294328002386,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005687136187543459,
+      "loss": 0.5804,
+      "step": 58210
+    },
+    {
+      "epoch": 2.891626105095858,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005686738849706963,
+      "loss": 0.5615,
+      "step": 58220
+    },
+    {
+      "epoch": 2.892122777391477,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005686341511870468,
+      "loss": 0.5725,
+      "step": 58230
+    },
+    {
+      "epoch": 2.8926194496870963,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005685944174033973,
+      "loss": 0.5545,
+      "step": 58240
+    },
+    {
+      "epoch": 2.893116121982716,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005685546836197477,
+      "loss": 0.5463,
+      "step": 58250
+    },
+    {
+      "epoch": 2.893612794278335,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005685149498360982,
+      "loss": 0.5872,
+      "step": 58260
+    },
+    {
+      "epoch": 2.8941094665739544,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005684752160524486,
+      "loss": 0.5727,
+      "step": 58270
+    },
+    {
+      "epoch": 2.894606138869574,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005684354822687991,
+      "loss": 0.5308,
+      "step": 58280
+    },
+    {
+      "epoch": 2.8951028111651933,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005683957484851496,
+      "loss": 0.5595,
+      "step": 58290
+    },
+    {
+      "epoch": 2.8955994834608125,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005683560147015,
+      "loss": 0.5706,
+      "step": 58300
+    },
+    {
+      "epoch": 2.8960961557564318,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005683162809178504,
+      "loss": 0.5375,
+      "step": 58310
+    },
+    {
+      "epoch": 2.896592828052051,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.000568276547134201,
+      "loss": 0.543,
+      "step": 58320
+    },
+    {
+      "epoch": 2.8970895003476707,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005682368133505513,
+      "loss": 0.5418,
+      "step": 58330
+    },
+    {
+      "epoch": 2.89758617264329,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005681970795669018,
+      "loss": 0.5714,
+      "step": 58340
+    },
+    {
+      "epoch": 2.8980828449389096,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005681573457832523,
+      "loss": 0.5453,
+      "step": 58350
+    },
+    {
+      "epoch": 2.898579517234529,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005681176119996026,
+      "loss": 0.5533,
+      "step": 58360
+    },
+    {
+      "epoch": 2.899076189530148,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005680778782159531,
+      "loss": 0.5524,
+      "step": 58370
+    },
+    {
+      "epoch": 2.8995728618257672,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005680381444323036,
+      "loss": 0.5252,
+      "step": 58380
+    },
+    {
+      "epoch": 2.9000695341213865,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005679984106486541,
+      "loss": 0.5223,
+      "step": 58390
+    },
+    {
+      "epoch": 2.900566206417006,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005679586768650045,
+      "loss": 0.553,
+      "step": 58400
+    },
+    {
+      "epoch": 2.9010628787126254,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0005679189430813549,
+      "loss": 0.5507,
+      "step": 58410
+    },
+    {
+      "epoch": 2.901559551008245,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005678792092977054,
+      "loss": 0.5367,
+      "step": 58420
+    },
+    {
+      "epoch": 2.9020562233038643,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005678394755140559,
+      "loss": 0.5669,
+      "step": 58430
+    },
+    {
+      "epoch": 2.9025528955994835,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005677997417304063,
+      "loss": 0.5491,
+      "step": 58440
+    },
+    {
+      "epoch": 2.9030495678951027,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005677600079467568,
+      "loss": 0.5746,
+      "step": 58450
+    },
+    {
+      "epoch": 2.903546240190722,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005677202741631072,
+      "loss": 0.5686,
+      "step": 58460
+    },
+    {
+      "epoch": 2.9040429124863416,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005676805403794576,
+      "loss": 0.5631,
+      "step": 58470
+    },
+    {
+      "epoch": 2.904539584781961,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005676408065958082,
+      "loss": 0.5275,
+      "step": 58480
+    },
+    {
+      "epoch": 2.90503625707758,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005676010728121586,
+      "loss": 0.5573,
+      "step": 58490
+    },
+    {
+      "epoch": 2.9055329293731997,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000567561339028509,
+      "loss": 0.5584,
+      "step": 58500
+    },
+    {
+      "epoch": 2.906029601668819,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005675216052448595,
+      "loss": 0.5452,
+      "step": 58510
+    },
+    {
+      "epoch": 2.906526273964438,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005674818714612098,
+      "loss": 0.5653,
+      "step": 58520
+    },
+    {
+      "epoch": 2.9070229462600574,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005674421376775604,
+      "loss": 0.5529,
+      "step": 58530
+    },
+    {
+      "epoch": 2.907519618555677,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005674024038939109,
+      "loss": 0.5543,
+      "step": 58540
+    },
+    {
+      "epoch": 2.9080162908512963,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005673626701102613,
+      "loss": 0.5662,
+      "step": 58550
+    },
+    {
+      "epoch": 2.9085129631469155,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005673229363266117,
+      "loss": 0.5571,
+      "step": 58560
+    },
+    {
+      "epoch": 2.909009635442535,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005672832025429621,
+      "loss": 0.5496,
+      "step": 58570
+    },
+    {
+      "epoch": 2.9095063077381544,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005672434687593127,
+      "loss": 0.5188,
+      "step": 58580
+    },
+    {
+      "epoch": 2.9100029800337737,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0005672037349756631,
+      "loss": 0.5523,
+      "step": 58590
+    },
+    {
+      "epoch": 2.910499652329393,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005671640011920135,
+      "loss": 0.5485,
+      "step": 58600
+    },
+    {
+      "epoch": 2.9109963246250126,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.000567124267408364,
+      "loss": 0.5503,
+      "step": 58610
+    },
+    {
+      "epoch": 2.911492996920632,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005670845336247144,
+      "loss": 0.562,
+      "step": 58620
+    },
+    {
+      "epoch": 2.911989669216251,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005670447998410649,
+      "loss": 0.5759,
+      "step": 58630
+    },
+    {
+      "epoch": 2.9124863415118707,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005670050660574154,
+      "loss": 0.5494,
+      "step": 58640
+    },
+    {
+      "epoch": 2.91298301380749,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005669653322737658,
+      "loss": 0.5299,
+      "step": 58650
+    },
+    {
+      "epoch": 2.913479686103109,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005669255984901162,
+      "loss": 0.5516,
+      "step": 58660
+    },
+    {
+      "epoch": 2.9139763583987284,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005668858647064667,
+      "loss": 0.5365,
+      "step": 58670
+    },
+    {
+      "epoch": 2.9144730306943476,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005668461309228172,
+      "loss": 0.5389,
+      "step": 58680
+    },
+    {
+      "epoch": 2.9149697029899673,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005668063971391676,
+      "loss": 0.5402,
+      "step": 58690
+    },
+    {
+      "epoch": 2.9154663752855865,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005667666633555181,
+      "loss": 0.5457,
+      "step": 58700
+    },
+    {
+      "epoch": 2.915963047581206,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005667269295718685,
+      "loss": 0.5668,
+      "step": 58710
+    },
+    {
+      "epoch": 2.9164597198768254,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005666871957882189,
+      "loss": 0.5632,
+      "step": 58720
+    },
+    {
+      "epoch": 2.9169563921724446,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005666474620045695,
+      "loss": 0.5504,
+      "step": 58730
+    },
+    {
+      "epoch": 2.917453064468064,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005666077282209199,
+      "loss": 0.5523,
+      "step": 58740
+    },
+    {
+      "epoch": 2.917949736763683,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005665679944372703,
+      "loss": 0.5607,
+      "step": 58750
+    },
+    {
+      "epoch": 2.9184464090593027,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005665282606536208,
+      "loss": 0.5626,
+      "step": 58760
+    },
+    {
+      "epoch": 2.918943081354922,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005664885268699712,
+      "loss": 0.5397,
+      "step": 58770
+    },
+    {
+      "epoch": 2.9194397536505416,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005664487930863217,
+      "loss": 0.5625,
+      "step": 58780
+    },
+    {
+      "epoch": 2.919936425946161,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005664090593026721,
+      "loss": 0.5383,
+      "step": 58790
+    },
+    {
+      "epoch": 2.92043309824178,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005663693255190226,
+      "loss": 0.5734,
+      "step": 58800
+    },
+    {
+      "epoch": 2.9209297705373993,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.000566329591735373,
+      "loss": 0.5363,
+      "step": 58810
+    },
+    {
+      "epoch": 2.9214264428330186,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005662898579517234,
+      "loss": 0.5437,
+      "step": 58820
+    },
+    {
+      "epoch": 2.9219231151286382,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.000566250124168074,
+      "loss": 0.5454,
+      "step": 58830
+    },
+    {
+      "epoch": 2.9224197874242575,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005662103903844244,
+      "loss": 0.5451,
+      "step": 58840
+    },
+    {
+      "epoch": 2.9229164597198767,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005661706566007748,
+      "loss": 0.5632,
+      "step": 58850
+    },
+    {
+      "epoch": 2.9234131320154964,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005661309228171253,
+      "loss": 0.5585,
+      "step": 58860
+    },
+    {
+      "epoch": 2.9239098043111156,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005660911890334757,
+      "loss": 0.557,
+      "step": 58870
+    },
+    {
+      "epoch": 2.924406476606735,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005660514552498262,
+      "loss": 0.5664,
+      "step": 58880
+    },
+    {
+      "epoch": 2.924903148902354,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005660117214661767,
+      "loss": 0.5604,
+      "step": 58890
+    },
+    {
+      "epoch": 2.9253998211979737,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005659719876825271,
+      "loss": 0.6002,
+      "step": 58900
+    },
+    {
+      "epoch": 2.925896493493593,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005659322538988775,
+      "loss": 0.5526,
+      "step": 58910
+    },
+    {
+      "epoch": 2.926393165789212,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.000565892520115228,
+      "loss": 0.5556,
+      "step": 58920
+    },
+    {
+      "epoch": 2.926889838084832,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005658527863315785,
+      "loss": 0.5418,
+      "step": 58930
+    },
+    {
+      "epoch": 2.927386510380451,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005658130525479289,
+      "loss": 0.5495,
+      "step": 58940
+    },
+    {
+      "epoch": 2.9278831826760703,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005657733187642794,
+      "loss": 0.5517,
+      "step": 58950
+    },
+    {
+      "epoch": 2.9283798549716895,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0005657335849806298,
+      "loss": 0.5698,
+      "step": 58960
+    },
+    {
+      "epoch": 2.928876527267309,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005656938511969802,
+      "loss": 0.5611,
+      "step": 58970
+    },
+    {
+      "epoch": 2.9293731995629284,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005656541174133306,
+      "loss": 0.5241,
+      "step": 58980
+    },
+    {
+      "epoch": 2.9298698718585476,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005656143836296812,
+      "loss": 0.5577,
+      "step": 58990
+    },
+    {
+      "epoch": 2.9303665441541673,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005655746498460317,
+      "loss": 0.5671,
+      "step": 59000
+    },
+    {
+      "epoch": 2.9308632164497865,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.000565534916062382,
+      "loss": 0.5119,
+      "step": 59010
+    },
+    {
+      "epoch": 2.9313598887454058,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005654951822787325,
+      "loss": 0.5553,
+      "step": 59020
+    },
+    {
+      "epoch": 2.931856561041025,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005654554484950831,
+      "loss": 0.5578,
+      "step": 59030
+    },
+    {
+      "epoch": 2.932353233336644,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005654157147114334,
+      "loss": 0.5429,
+      "step": 59040
+    },
+    {
+      "epoch": 2.932849905632264,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005653759809277839,
+      "loss": 0.5359,
+      "step": 59050
+    },
+    {
+      "epoch": 2.933346577927883,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005653362471441343,
+      "loss": 0.5667,
+      "step": 59060
+    },
+    {
+      "epoch": 2.933843250223503,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005652965133604847,
+      "loss": 0.5741,
+      "step": 59070
+    },
+    {
+      "epoch": 2.934339922519122,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005652567795768353,
+      "loss": 0.5537,
+      "step": 59080
+    },
+    {
+      "epoch": 2.9348365948147412,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005652170457931857,
+      "loss": 0.5458,
+      "step": 59090
+    },
+    {
+      "epoch": 2.9353332671103605,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005651773120095361,
+      "loss": 0.597,
+      "step": 59100
+    },
+    {
+      "epoch": 2.9358299394059797,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005651375782258866,
+      "loss": 0.5725,
+      "step": 59110
+    },
+    {
+      "epoch": 2.9363266117015994,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000565097844442237,
+      "loss": 0.5645,
+      "step": 59120
+    },
+    {
+      "epoch": 2.9368232839972186,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005650581106585874,
+      "loss": 0.5217,
+      "step": 59130
+    },
+    {
+      "epoch": 2.937319956292838,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.000565018376874938,
+      "loss": 0.5433,
+      "step": 59140
+    },
+    {
+      "epoch": 2.9378166285884575,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005649786430912884,
+      "loss": 0.5658,
+      "step": 59150
+    },
+    {
+      "epoch": 2.9383133008840767,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005649389093076389,
+      "loss": 0.5592,
+      "step": 59160
+    },
+    {
+      "epoch": 2.938809973179696,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005648991755239892,
+      "loss": 0.5329,
+      "step": 59170
+    },
+    {
+      "epoch": 2.939306645475315,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005648594417403397,
+      "loss": 0.5381,
+      "step": 59180
+    },
+    {
+      "epoch": 2.939803317770935,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005648197079566903,
+      "loss": 0.5439,
+      "step": 59190
+    },
+    {
+      "epoch": 2.940299990066554,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005647799741730406,
+      "loss": 0.5555,
+      "step": 59200
+    },
+    {
+      "epoch": 2.9407966623621733,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005647402403893911,
+      "loss": 0.5537,
+      "step": 59210
+    },
+    {
+      "epoch": 2.941293334657793,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005647005066057416,
+      "loss": 0.5486,
+      "step": 59220
+    },
+    {
+      "epoch": 2.941790006953412,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005646607728220919,
+      "loss": 0.5506,
+      "step": 59230
+    },
+    {
+      "epoch": 2.9422866792490314,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005646210390384425,
+      "loss": 0.5734,
+      "step": 59240
+    },
+    {
+      "epoch": 2.9427833515446506,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005645813052547929,
+      "loss": 0.5873,
+      "step": 59250
+    },
+    {
+      "epoch": 2.9432800238402703,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005645415714711433,
+      "loss": 0.5386,
+      "step": 59260
+    },
+    {
+      "epoch": 2.9437766961358895,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005645018376874938,
+      "loss": 0.542,
+      "step": 59270
+    },
+    {
+      "epoch": 2.9442733684315088,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005644621039038442,
+      "loss": 0.5314,
+      "step": 59280
+    },
+    {
+      "epoch": 2.9447700407271284,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005644223701201948,
+      "loss": 0.5418,
+      "step": 59290
+    },
+    {
+      "epoch": 2.9452667130227477,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005643826363365452,
+      "loss": 0.5572,
+      "step": 59300
+    },
+    {
+      "epoch": 2.945763385318367,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005643429025528956,
+      "loss": 0.5328,
+      "step": 59310
+    },
+    {
+      "epoch": 2.946260057613986,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005643031687692461,
+      "loss": 0.5504,
+      "step": 59320
+    },
+    {
+      "epoch": 2.946756729909606,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005642634349855965,
+      "loss": 0.5546,
+      "step": 59330
+    },
+    {
+      "epoch": 2.947253402205225,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.000564223701201947,
+      "loss": 0.5449,
+      "step": 59340
+    },
+    {
+      "epoch": 2.9477500745008443,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005641839674182975,
+      "loss": 0.5657,
+      "step": 59350
+    },
+    {
+      "epoch": 2.948246746796464,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005641442336346479,
+      "loss": 0.5305,
+      "step": 59360
+    },
+    {
+      "epoch": 2.948743419092083,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005641044998509983,
+      "loss": 0.5326,
+      "step": 59370
+    },
+    {
+      "epoch": 2.9492400913877024,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005640647660673489,
+      "loss": 0.5455,
+      "step": 59380
+    },
+    {
+      "epoch": 2.9497367636833216,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005640250322836992,
+      "loss": 0.5653,
+      "step": 59390
+    },
+    {
+      "epoch": 2.950233435978941,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005639852985000497,
+      "loss": 0.5481,
+      "step": 59400
+    },
+    {
+      "epoch": 2.9507301082745605,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005639455647164002,
+      "loss": 0.5329,
+      "step": 59410
+    },
+    {
+      "epoch": 2.9512267805701797,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005639058309327505,
+      "loss": 0.5738,
+      "step": 59420
+    },
+    {
+      "epoch": 2.9517234528657994,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.000563866097149101,
+      "loss": 0.5695,
+      "step": 59430
+    },
+    {
+      "epoch": 2.9522201251614186,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005638263633654515,
+      "loss": 0.5291,
+      "step": 59440
+    },
+    {
+      "epoch": 2.952716797457038,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.000563786629581802,
+      "loss": 0.5616,
+      "step": 59450
+    },
+    {
+      "epoch": 2.953213469752657,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005637468957981524,
+      "loss": 0.5414,
+      "step": 59460
+    },
+    {
+      "epoch": 2.9537101420482763,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005637071620145028,
+      "loss": 0.5406,
+      "step": 59470
+    },
+    {
+      "epoch": 2.954206814343896,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005636674282308533,
+      "loss": 0.5475,
+      "step": 59480
+    },
+    {
+      "epoch": 2.954703486639515,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005636276944472038,
+      "loss": 0.541,
+      "step": 59490
+    },
+    {
+      "epoch": 2.9552001589351344,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005635879606635542,
+      "loss": 0.5627,
+      "step": 59500
+    },
+    {
+      "epoch": 2.955696831230754,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005635482268799047,
+      "loss": 0.5566,
+      "step": 59510
+    },
+    {
+      "epoch": 2.9561935035263733,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005635084930962551,
+      "loss": 0.5491,
+      "step": 59520
+    },
+    {
+      "epoch": 2.9566901758219926,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005634687593126055,
+      "loss": 0.5696,
+      "step": 59530
+    },
+    {
+      "epoch": 2.957186848117612,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005634290255289561,
+      "loss": 0.5736,
+      "step": 59540
+    },
+    {
+      "epoch": 2.9576835204132315,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005633892917453065,
+      "loss": 0.5498,
+      "step": 59550
+    },
+    {
+      "epoch": 2.9581801927088507,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005633495579616569,
+      "loss": 0.571,
+      "step": 59560
+    },
+    {
+      "epoch": 2.95867686500447,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005633098241780074,
+      "loss": 0.5617,
+      "step": 59570
+    },
+    {
+      "epoch": 2.9591735373000896,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0005632700903943577,
+      "loss": 0.563,
+      "step": 59580
+    },
+    {
+      "epoch": 2.959670209595709,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005632303566107083,
+      "loss": 0.5737,
+      "step": 59590
+    },
+    {
+      "epoch": 2.960166881891328,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005631906228270588,
+      "loss": 0.5386,
+      "step": 59600
+    },
+    {
+      "epoch": 2.9606635541869473,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005631508890434092,
+      "loss": 0.561,
+      "step": 59610
+    },
+    {
+      "epoch": 2.961160226482567,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005631111552597596,
+      "loss": 0.5493,
+      "step": 59620
+    },
+    {
+      "epoch": 2.961656898778186,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005630714214761101,
+      "loss": 0.5367,
+      "step": 59630
+    },
+    {
+      "epoch": 2.9621535710738054,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005630316876924606,
+      "loss": 0.5491,
+      "step": 59640
+    },
+    {
+      "epoch": 2.962650243369425,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.000562991953908811,
+      "loss": 0.536,
+      "step": 59650
+    },
+    {
+      "epoch": 2.9631469156650443,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005629522201251614,
+      "loss": 0.5557,
+      "step": 59660
+    },
+    {
+      "epoch": 2.9636435879606635,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005629124863415119,
+      "loss": 0.57,
+      "step": 59670
+    },
+    {
+      "epoch": 2.9641402602562827,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005628727525578623,
+      "loss": 0.5497,
+      "step": 59680
+    },
+    {
+      "epoch": 2.9646369325519024,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005628330187742128,
+      "loss": 0.5465,
+      "step": 59690
+    },
+    {
+      "epoch": 2.9651336048475216,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005627932849905633,
+      "loss": 0.5469,
+      "step": 59700
+    },
+    {
+      "epoch": 2.965630277143141,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005627535512069137,
+      "loss": 0.5711,
+      "step": 59710
+    },
+    {
+      "epoch": 2.9661269494387605,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0005627138174232641,
+      "loss": 0.5855,
+      "step": 59720
+    },
+    {
+      "epoch": 2.9666236217343798,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005626740836396146,
+      "loss": 0.5443,
+      "step": 59730
+    },
+    {
+      "epoch": 2.967120294029999,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005626343498559651,
+      "loss": 0.593,
+      "step": 59740
+    },
+    {
+      "epoch": 2.967616966325618,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005625946160723155,
+      "loss": 0.5584,
+      "step": 59750
+    },
+    {
+      "epoch": 2.9681136386212374,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.000562554882288666,
+      "loss": 0.5535,
+      "step": 59760
+    },
+    {
+      "epoch": 2.968610310916857,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005625151485050164,
+      "loss": 0.5354,
+      "step": 59770
+    },
+    {
+      "epoch": 2.9691069832124763,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005624754147213668,
+      "loss": 0.5449,
+      "step": 59780
+    },
+    {
+      "epoch": 2.969603655508096,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005624356809377174,
+      "loss": 0.5572,
+      "step": 59790
+    },
+    {
+      "epoch": 2.9701003278037152,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005623959471540678,
+      "loss": 0.5555,
+      "step": 59800
+    },
+    {
+      "epoch": 2.9705970000993345,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005623562133704182,
+      "loss": 0.5341,
+      "step": 59810
+    },
+    {
+      "epoch": 2.9710936723949537,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005623164795867687,
+      "loss": 0.5561,
+      "step": 59820
+    },
+    {
+      "epoch": 2.971590344690573,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0005622767458031191,
+      "loss": 0.5475,
+      "step": 59830
+    },
+    {
+      "epoch": 2.9720870169861926,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005622370120194696,
+      "loss": 0.5521,
+      "step": 59840
+    },
+    {
+      "epoch": 2.972583689281812,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00056219727823582,
+      "loss": 0.5514,
+      "step": 59850
+    },
+    {
+      "epoch": 2.973080361577431,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005621575444521705,
+      "loss": 0.5566,
+      "step": 59860
+    },
+    {
+      "epoch": 2.9735770338730507,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005621178106685209,
+      "loss": 0.5545,
+      "step": 59870
+    },
+    {
+      "epoch": 2.97407370616867,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005620780768848713,
+      "loss": 0.5559,
+      "step": 59880
+    },
+    {
+      "epoch": 2.974570378464289,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005620383431012219,
+      "loss": 0.5307,
+      "step": 59890
+    },
+    {
+      "epoch": 2.9750670507599084,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005619986093175724,
+      "loss": 0.5339,
+      "step": 59900
+    },
+    {
+      "epoch": 2.975563723055528,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005619588755339227,
+      "loss": 0.5683,
+      "step": 59910
+    },
+    {
+      "epoch": 2.9760603953511473,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005619191417502732,
+      "loss": 0.5408,
+      "step": 59920
+    },
+    {
+      "epoch": 2.9765570676467665,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005618794079666236,
+      "loss": 0.5533,
+      "step": 59930
+    },
+    {
+      "epoch": 2.977053739942386,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.000561839674182974,
+      "loss": 0.5609,
+      "step": 59940
+    },
+    {
+      "epoch": 2.9775504122380054,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005617999403993246,
+      "loss": 0.5143,
+      "step": 59950
+    },
+    {
+      "epoch": 2.9780470845336247,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.000561760206615675,
+      "loss": 0.5777,
+      "step": 59960
+    },
+    {
+      "epoch": 2.978543756829244,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005617204728320254,
+      "loss": 0.5602,
+      "step": 59970
+    },
+    {
+      "epoch": 2.9790404291248636,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005616807390483759,
+      "loss": 0.5576,
+      "step": 59980
+    },
+    {
+      "epoch": 2.979537101420483,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005616410052647264,
+      "loss": 0.5405,
+      "step": 59990
+    },
+    {
+      "epoch": 2.980033773716102,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005616012714810768,
+      "loss": 0.5464,
+      "step": 60000
+    },
+    {
+      "epoch": 2.9805304460117217,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005615615376974273,
+      "loss": 0.557,
+      "step": 60010
+    },
+    {
+      "epoch": 2.981027118307341,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005615218039137777,
+      "loss": 0.5384,
+      "step": 60020
+    },
+    {
+      "epoch": 2.98152379060296,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005614820701301282,
+      "loss": 0.5597,
+      "step": 60030
+    },
+    {
+      "epoch": 2.9820204628985794,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005614423363464786,
+      "loss": 0.5859,
+      "step": 60040
+    },
+    {
+      "epoch": 2.982517135194199,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005614026025628291,
+      "loss": 0.5451,
+      "step": 60050
+    },
+    {
+      "epoch": 2.9830138074898183,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005613628687791796,
+      "loss": 0.5357,
+      "step": 60060
+    },
+    {
+      "epoch": 2.9835104797854375,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005613231349955299,
+      "loss": 0.5503,
+      "step": 60070
+    },
+    {
+      "epoch": 2.984007152081057,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005612834012118804,
+      "loss": 0.5272,
+      "step": 60080
+    },
+    {
+      "epoch": 2.9845038243766764,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.000561243667428231,
+      "loss": 0.5474,
+      "step": 60090
+    },
+    {
+      "epoch": 2.9850004966722956,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005612039336445813,
+      "loss": 0.5673,
+      "step": 60100
+    },
+    {
+      "epoch": 2.985497168967915,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005611641998609318,
+      "loss": 0.5401,
+      "step": 60110
+    },
+    {
+      "epoch": 2.985993841263534,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005611244660772822,
+      "loss": 0.5523,
+      "step": 60120
+    },
+    {
+      "epoch": 2.9864905135591537,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005610847322936326,
+      "loss": 0.5665,
+      "step": 60130
+    },
+    {
+      "epoch": 2.986987185854773,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005610449985099832,
+      "loss": 0.517,
+      "step": 60140
+    },
+    {
+      "epoch": 2.9874838581503926,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005610052647263336,
+      "loss": 0.552,
+      "step": 60150
+    },
+    {
+      "epoch": 2.987980530446012,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.000560965530942684,
+      "loss": 0.5065,
+      "step": 60160
+    },
+    {
+      "epoch": 2.988477202741631,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005609257971590345,
+      "loss": 0.542,
+      "step": 60170
+    },
+    {
+      "epoch": 2.9889738750372503,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005608860633753849,
+      "loss": 0.5447,
+      "step": 60180
+    },
+    {
+      "epoch": 2.9894705473328695,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005608463295917355,
+      "loss": 0.5466,
+      "step": 60190
+    },
+    {
+      "epoch": 2.989967219628489,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005608065958080859,
+      "loss": 0.5523,
+      "step": 60200
+    },
+    {
+      "epoch": 2.9904638919241084,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005607668620244363,
+      "loss": 0.5566,
+      "step": 60210
+    },
+    {
+      "epoch": 2.9909605642197277,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005607271282407868,
+      "loss": 0.5332,
+      "step": 60220
+    },
+    {
+      "epoch": 2.9914572365153473,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005606873944571372,
+      "loss": 0.5049,
+      "step": 60230
+    },
+    {
+      "epoch": 2.9919539088109666,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005606476606734877,
+      "loss": 0.5662,
+      "step": 60240
+    },
+    {
+      "epoch": 2.992450581106586,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005606079268898382,
+      "loss": 0.5306,
+      "step": 60250
+    },
+    {
+      "epoch": 2.992947253402205,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005605681931061885,
+      "loss": 0.5561,
+      "step": 60260
+    },
+    {
+      "epoch": 2.9934439256978247,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.000560528459322539,
+      "loss": 0.5283,
+      "step": 60270
+    },
+    {
+      "epoch": 2.993940597993444,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005604887255388895,
+      "loss": 0.5596,
+      "step": 60280
+    },
+    {
+      "epoch": 2.994437270289063,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005604489917552398,
+      "loss": 0.5754,
+      "step": 60290
+    },
+    {
+      "epoch": 2.994933942584683,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005604092579715904,
+      "loss": 0.5628,
+      "step": 60300
+    },
+    {
+      "epoch": 2.995430614880302,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005603695241879408,
+      "loss": 0.5484,
+      "step": 60310
+    },
+    {
+      "epoch": 2.9959272871759213,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005603297904042912,
+      "loss": 0.5422,
+      "step": 60320
+    },
+    {
+      "epoch": 2.9964239594715405,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005602900566206417,
+      "loss": 0.5361,
+      "step": 60330
+    },
+    {
+      "epoch": 2.99692063176716,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005602503228369921,
+      "loss": 0.5161,
+      "step": 60340
+    },
+    {
+      "epoch": 2.9974173040627794,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005602105890533427,
+      "loss": 0.5492,
+      "step": 60350
+    },
+    {
+      "epoch": 2.9979139763583986,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005601708552696931,
+      "loss": 0.5375,
+      "step": 60360
+    },
+    {
+      "epoch": 2.9984106486540183,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005601311214860435,
+      "loss": 0.5327,
+      "step": 60370
+    },
+    {
+      "epoch": 2.9989073209496375,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.000560091387702394,
+      "loss": 0.5367,
+      "step": 60380
+    },
+    {
+      "epoch": 2.9994039932452567,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005600516539187445,
+      "loss": 0.5471,
+      "step": 60390
+    },
+    {
+      "epoch": 2.999900665540876,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005600119201350949,
+      "loss": 0.5716,
+      "step": 60400
+    },
+    {
+      "epoch": 3.0003973378364956,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005599721863514454,
+      "loss": 0.5505,
+      "step": 60410
+    },
+    {
+      "epoch": 3.000894010132115,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005599324525677958,
+      "loss": 0.5274,
+      "step": 60420
+    },
+    {
+      "epoch": 3.001390682427734,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005598927187841462,
+      "loss": 0.5332,
+      "step": 60430
+    },
+    {
+      "epoch": 3.0018873547233533,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005598529850004968,
+      "loss": 0.5394,
+      "step": 60440
+    },
+    {
+      "epoch": 3.002384027018973,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005598132512168471,
+      "loss": 0.5643,
+      "step": 60450
+    },
+    {
+      "epoch": 3.0028806993145922,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005597735174331976,
+      "loss": 0.5487,
+      "step": 60460
+    },
+    {
+      "epoch": 3.0033773716102115,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005597337836495481,
+      "loss": 0.5516,
+      "step": 60470
+    },
+    {
+      "epoch": 3.003874043905831,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005596940498658985,
+      "loss": 0.5343,
+      "step": 60480
+    },
+    {
+      "epoch": 3.0043707162014504,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.000559654316082249,
+      "loss": 0.5349,
+      "step": 60490
+    },
+    {
+      "epoch": 3.0048673884970696,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005596145822985995,
+      "loss": 0.5361,
+      "step": 60500
+    },
+    {
+      "epoch": 3.005364060792689,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005595748485149499,
+      "loss": 0.5158,
+      "step": 60510
+    },
+    {
+      "epoch": 3.0058607330883085,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005595351147313003,
+      "loss": 0.5549,
+      "step": 60520
+    },
+    {
+      "epoch": 3.0063574053839277,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005594953809476507,
+      "loss": 0.5027,
+      "step": 60530
+    },
+    {
+      "epoch": 3.006854077679547,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005594556471640013,
+      "loss": 0.5512,
+      "step": 60540
+    },
+    {
+      "epoch": 3.0073507499751666,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005594159133803517,
+      "loss": 0.5563,
+      "step": 60550
+    },
+    {
+      "epoch": 3.007847422270786,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005593761795967021,
+      "loss": 0.551,
+      "step": 60560
+    },
+    {
+      "epoch": 3.008344094566405,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005593364458130526,
+      "loss": 0.5456,
+      "step": 60570
+    },
+    {
+      "epoch": 3.0088407668620243,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.000559296712029403,
+      "loss": 0.5606,
+      "step": 60580
+    },
+    {
+      "epoch": 3.009337439157644,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005592569782457534,
+      "loss": 0.5347,
+      "step": 60590
+    },
+    {
+      "epoch": 3.009834111453263,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.000559217244462104,
+      "loss": 0.5382,
+      "step": 60600
+    },
+    {
+      "epoch": 3.0103307837488824,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005591775106784544,
+      "loss": 0.5571,
+      "step": 60610
+    },
+    {
+      "epoch": 3.0108274560445016,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005591377768948048,
+      "loss": 0.5239,
+      "step": 60620
+    },
+    {
+      "epoch": 3.0113241283401213,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005590980431111553,
+      "loss": 0.5453,
+      "step": 60630
+    },
+    {
+      "epoch": 3.0118208006357405,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005590583093275057,
+      "loss": 0.5146,
+      "step": 60640
+    },
+    {
+      "epoch": 3.0123174729313598,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005590185755438562,
+      "loss": 0.506,
+      "step": 60650
+    },
+    {
+      "epoch": 3.0128141452269794,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0005589788417602067,
+      "loss": 0.5441,
+      "step": 60660
+    },
+    {
+      "epoch": 3.0133108175225987,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005589391079765571,
+      "loss": 0.5373,
+      "step": 60670
+    },
+    {
+      "epoch": 3.013807489818218,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005588993741929075,
+      "loss": 0.5262,
+      "step": 60680
+    },
+    {
+      "epoch": 3.014304162113837,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.000558859640409258,
+      "loss": 0.5653,
+      "step": 60690
+    },
+    {
+      "epoch": 3.014800834409457,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005588199066256085,
+      "loss": 0.5368,
+      "step": 60700
+    },
+    {
+      "epoch": 3.015297506705076,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005587801728419589,
+      "loss": 0.5373,
+      "step": 60710
+    },
+    {
+      "epoch": 3.0157941790006952,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005587404390583093,
+      "loss": 0.5432,
+      "step": 60720
+    },
+    {
+      "epoch": 3.016290851296315,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005587007052746598,
+      "loss": 0.5181,
+      "step": 60730
+    },
+    {
+      "epoch": 3.016787523591934,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0005586609714910102,
+      "loss": 0.5347,
+      "step": 60740
+    },
+    {
+      "epoch": 3.0172841958875534,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005586212377073607,
+      "loss": 0.5342,
+      "step": 60750
+    },
+    {
+      "epoch": 3.0177808681831726,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005585815039237112,
+      "loss": 0.5126,
+      "step": 60760
+    },
+    {
+      "epoch": 3.0182775404787923,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005585417701400616,
+      "loss": 0.5561,
+      "step": 60770
+    },
+    {
+      "epoch": 3.0187742127744115,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.000558502036356412,
+      "loss": 0.5455,
+      "step": 60780
+    },
+    {
+      "epoch": 3.0192708850700307,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005584623025727625,
+      "loss": 0.5802,
+      "step": 60790
+    },
+    {
+      "epoch": 3.01976755736565,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.000558422568789113,
+      "loss": 0.5601,
+      "step": 60800
+    },
+    {
+      "epoch": 3.0202642296612696,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005583828350054634,
+      "loss": 0.5845,
+      "step": 60810
+    },
+    {
+      "epoch": 3.020760901956889,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005583431012218139,
+      "loss": 0.5392,
+      "step": 60820
+    },
+    {
+      "epoch": 3.021257574252508,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005583033674381643,
+      "loss": 0.5334,
+      "step": 60830
+    },
+    {
+      "epoch": 3.0217542465481277,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005582636336545147,
+      "loss": 0.5474,
+      "step": 60840
+    },
+    {
+      "epoch": 3.022250918843747,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005582238998708653,
+      "loss": 0.5282,
+      "step": 60850
+    },
+    {
+      "epoch": 3.022747591139366,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005581841660872157,
+      "loss": 0.547,
+      "step": 60860
+    },
+    {
+      "epoch": 3.0232442634349854,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005581444323035661,
+      "loss": 0.5394,
+      "step": 60870
+    },
+    {
+      "epoch": 3.023740935730605,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005581046985199166,
+      "loss": 0.5465,
+      "step": 60880
+    },
+    {
+      "epoch": 3.0242376080262243,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.000558064964736267,
+      "loss": 0.5242,
+      "step": 60890
+    },
+    {
+      "epoch": 3.0247342803218435,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005580252309526175,
+      "loss": 0.5338,
+      "step": 60900
+    },
+    {
+      "epoch": 3.0252309526174628,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005579854971689679,
+      "loss": 0.5479,
+      "step": 60910
+    },
+    {
+      "epoch": 3.0257276249130824,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005579457633853184,
+      "loss": 0.5434,
+      "step": 60920
+    },
+    {
+      "epoch": 3.0262242972087017,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005579060296016689,
+      "loss": 0.5436,
+      "step": 60930
+    },
+    {
+      "epoch": 3.026720969504321,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005578662958180192,
+      "loss": 0.5053,
+      "step": 60940
+    },
+    {
+      "epoch": 3.0272176417999406,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005578265620343698,
+      "loss": 0.52,
+      "step": 60950
+    },
+    {
+      "epoch": 3.02771431409556,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005577868282507203,
+      "loss": 0.5536,
+      "step": 60960
+    },
+    {
+      "epoch": 3.028210986391179,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005577470944670706,
+      "loss": 0.54,
+      "step": 60970
+    },
+    {
+      "epoch": 3.0287076586867983,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005577073606834211,
+      "loss": 0.5429,
+      "step": 60980
+    },
+    {
+      "epoch": 3.029204330982418,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005576676268997715,
+      "loss": 0.4954,
+      "step": 60990
+    },
+    {
+      "epoch": 3.029701003278037,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.000557627893116122,
+      "loss": 0.5265,
+      "step": 61000
+    },
+    {
+      "epoch": 3.0301976755736564,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005575881593324725,
+      "loss": 0.554,
+      "step": 61010
+    },
+    {
+      "epoch": 3.030694347869276,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005575484255488229,
+      "loss": 0.595,
+      "step": 61020
+    },
+    {
+      "epoch": 3.0311910201648953,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005575086917651733,
+      "loss": 0.5713,
+      "step": 61030
+    },
+    {
+      "epoch": 3.0316876924605145,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005574689579815238,
+      "loss": 0.5409,
+      "step": 61040
+    },
+    {
+      "epoch": 3.0321843647561337,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005574292241978743,
+      "loss": 0.529,
+      "step": 61050
+    },
+    {
+      "epoch": 3.0326810370517534,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005573894904142247,
+      "loss": 0.5127,
+      "step": 61060
+    },
+    {
+      "epoch": 3.0331777093473726,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005573497566305752,
+      "loss": 0.5463,
+      "step": 61070
+    },
+    {
+      "epoch": 3.033674381642992,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005573100228469256,
+      "loss": 0.5276,
+      "step": 61080
+    },
+    {
+      "epoch": 3.0341710539386115,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005572702890632761,
+      "loss": 0.5473,
+      "step": 61090
+    },
+    {
+      "epoch": 3.0346677262342308,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005572305552796266,
+      "loss": 0.5491,
+      "step": 61100
+    },
+    {
+      "epoch": 3.03516439852985,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.000557190821495977,
+      "loss": 0.5376,
+      "step": 61110
+    },
+    {
+      "epoch": 3.035661070825469,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005571510877123275,
+      "loss": 0.5374,
+      "step": 61120
+    },
+    {
+      "epoch": 3.036157743121089,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005571113539286778,
+      "loss": 0.549,
+      "step": 61130
+    },
+    {
+      "epoch": 3.036654415416708,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005570716201450283,
+      "loss": 0.5694,
+      "step": 61140
+    },
+    {
+      "epoch": 3.0371510877123273,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005570318863613789,
+      "loss": 0.5125,
+      "step": 61150
+    },
+    {
+      "epoch": 3.0376477600079466,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005569921525777292,
+      "loss": 0.5519,
+      "step": 61160
+    },
+    {
+      "epoch": 3.0381444323035662,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005569524187940797,
+      "loss": 0.5451,
+      "step": 61170
+    },
+    {
+      "epoch": 3.0386411045991855,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005569126850104301,
+      "loss": 0.5599,
+      "step": 61180
+    },
+    {
+      "epoch": 3.0391377768948047,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005568729512267805,
+      "loss": 0.4976,
+      "step": 61190
+    },
+    {
+      "epoch": 3.0396344491904244,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005568332174431311,
+      "loss": 0.5451,
+      "step": 61200
+    },
+    {
+      "epoch": 3.0401311214860436,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005567934836594815,
+      "loss": 0.5469,
+      "step": 61210
+    },
+    {
+      "epoch": 3.040627793781663,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005567537498758319,
+      "loss": 0.5476,
+      "step": 61220
+    },
+    {
+      "epoch": 3.041124466077282,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005567140160921824,
+      "loss": 0.5501,
+      "step": 61230
+    },
+    {
+      "epoch": 3.0416211383729017,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005566742823085328,
+      "loss": 0.5679,
+      "step": 61240
+    },
+    {
+      "epoch": 3.042117810668521,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005566345485248834,
+      "loss": 0.5431,
+      "step": 61250
+    },
+    {
+      "epoch": 3.04261448296414,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005565948147412338,
+      "loss": 0.5192,
+      "step": 61260
+    },
+    {
+      "epoch": 3.0431111552597594,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005565550809575842,
+      "loss": 0.5309,
+      "step": 61270
+    },
+    {
+      "epoch": 3.043607827555379,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005565153471739347,
+      "loss": 0.5166,
+      "step": 61280
+    },
+    {
+      "epoch": 3.0441044998509983,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005564756133902851,
+      "loss": 0.5208,
+      "step": 61290
+    },
+    {
+      "epoch": 3.0446011721466175,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005564358796066356,
+      "loss": 0.5551,
+      "step": 61300
+    },
+    {
+      "epoch": 3.045097844442237,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005563961458229861,
+      "loss": 0.5407,
+      "step": 61310
+    },
+    {
+      "epoch": 3.0455945167378564,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005563564120393364,
+      "loss": 0.5215,
+      "step": 61320
+    },
+    {
+      "epoch": 3.0460911890334756,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005563166782556869,
+      "loss": 0.5664,
+      "step": 61330
+    },
+    {
+      "epoch": 3.046587861329095,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005562769444720374,
+      "loss": 0.5375,
+      "step": 61340
+    },
+    {
+      "epoch": 3.0470845336247145,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005562372106883878,
+      "loss": 0.5631,
+      "step": 61350
+    },
+    {
+      "epoch": 3.0475812059203338,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005561974769047383,
+      "loss": 0.5521,
+      "step": 61360
+    },
+    {
+      "epoch": 3.048077878215953,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005561577431210888,
+      "loss": 0.5416,
+      "step": 61370
+    },
+    {
+      "epoch": 3.0485745505115727,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005561180093374392,
+      "loss": 0.5347,
+      "step": 61380
+    },
+    {
+      "epoch": 3.049071222807192,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005560782755537896,
+      "loss": 0.5525,
+      "step": 61390
+    },
+    {
+      "epoch": 3.049567895102811,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00055603854177014,
+      "loss": 0.5232,
+      "step": 61400
+    },
+    {
+      "epoch": 3.0500645673984303,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005559988079864906,
+      "loss": 0.5682,
+      "step": 61410
+    },
+    {
+      "epoch": 3.05056123969405,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.000555959074202841,
+      "loss": 0.5561,
+      "step": 61420
+    },
+    {
+      "epoch": 3.0510579119896692,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005559193404191914,
+      "loss": 0.547,
+      "step": 61430
+    },
+    {
+      "epoch": 3.0515545842852885,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005558796066355419,
+      "loss": 0.5549,
+      "step": 61440
+    },
+    {
+      "epoch": 3.0520512565809077,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005558398728518924,
+      "loss": 0.534,
+      "step": 61450
+    },
+    {
+      "epoch": 3.0525479288765274,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005558001390682428,
+      "loss": 0.5269,
+      "step": 61460
+    },
+    {
+      "epoch": 3.0530446011721466,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005557604052845933,
+      "loss": 0.554,
+      "step": 61470
+    },
+    {
+      "epoch": 3.053541273467766,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005557206715009437,
+      "loss": 0.5262,
+      "step": 61480
+    },
+    {
+      "epoch": 3.0540379457633855,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005556809377172941,
+      "loss": 0.5439,
+      "step": 61490
+    },
+    {
+      "epoch": 3.0545346180590047,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005556412039336447,
+      "loss": 0.5188,
+      "step": 61500
+    },
+    {
+      "epoch": 3.055031290354624,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.000555601470149995,
+      "loss": 0.5862,
+      "step": 61510
+    },
+    {
+      "epoch": 3.055527962650243,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005555617363663455,
+      "loss": 0.5218,
+      "step": 61520
+    },
+    {
+      "epoch": 3.056024634945863,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.000555522002582696,
+      "loss": 0.5514,
+      "step": 61530
+    },
+    {
+      "epoch": 3.056521307241482,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005554822687990464,
+      "loss": 0.5412,
+      "step": 61540
+    },
+    {
+      "epoch": 3.0570179795371013,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005554425350153969,
+      "loss": 0.5377,
+      "step": 61550
+    },
+    {
+      "epoch": 3.057514651832721,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005554028012317474,
+      "loss": 0.519,
+      "step": 61560
+    },
+    {
+      "epoch": 3.05801132412834,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005553630674480978,
+      "loss": 0.5234,
+      "step": 61570
+    },
+    {
+      "epoch": 3.0585079964239594,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005553233336644482,
+      "loss": 0.5131,
+      "step": 61580
+    },
+    {
+      "epoch": 3.0590046687195787,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005552835998807986,
+      "loss": 0.5561,
+      "step": 61590
+    },
+    {
+      "epoch": 3.0595013410151983,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005552438660971492,
+      "loss": 0.5365,
+      "step": 61600
+    },
+    {
+      "epoch": 3.0599980133108176,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005552041323134996,
+      "loss": 0.5246,
+      "step": 61610
+    },
+    {
+      "epoch": 3.060494685606437,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00055516439852985,
+      "loss": 0.5475,
+      "step": 61620
+    },
+    {
+      "epoch": 3.060991357902056,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005551246647462005,
+      "loss": 0.5341,
+      "step": 61630
+    },
+    {
+      "epoch": 3.0614880301976757,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005550849309625509,
+      "loss": 0.529,
+      "step": 61640
+    },
+    {
+      "epoch": 3.061984702493295,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005550451971789013,
+      "loss": 0.5492,
+      "step": 61650
+    },
+    {
+      "epoch": 3.062481374788914,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005550054633952519,
+      "loss": 0.5384,
+      "step": 61660
+    },
+    {
+      "epoch": 3.062978047084534,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005549657296116023,
+      "loss": 0.5352,
+      "step": 61670
+    },
+    {
+      "epoch": 3.063474719380153,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005549259958279527,
+      "loss": 0.5321,
+      "step": 61680
+    },
+    {
+      "epoch": 3.0639713916757723,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005548862620443032,
+      "loss": 0.5378,
+      "step": 61690
+    },
+    {
+      "epoch": 3.0644680639713915,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005548465282606537,
+      "loss": 0.5709,
+      "step": 61700
+    },
+    {
+      "epoch": 3.064964736267011,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005548067944770041,
+      "loss": 0.5592,
+      "step": 61710
+    },
+    {
+      "epoch": 3.0654614085626304,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005547670606933546,
+      "loss": 0.5601,
+      "step": 61720
+    },
+    {
+      "epoch": 3.0659580808582496,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.000554727326909705,
+      "loss": 0.5452,
+      "step": 61730
+    },
+    {
+      "epoch": 3.0664547531538693,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005546875931260554,
+      "loss": 0.554,
+      "step": 61740
+    },
+    {
+      "epoch": 3.0669514254494885,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.000554647859342406,
+      "loss": 0.5515,
+      "step": 61750
+    },
+    {
+      "epoch": 3.0674480977451077,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005546081255587564,
+      "loss": 0.537,
+      "step": 61760
+    },
+    {
+      "epoch": 3.067944770040727,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005545683917751068,
+      "loss": 0.5719,
+      "step": 61770
+    },
+    {
+      "epoch": 3.0684414423363466,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005545286579914572,
+      "loss": 0.4947,
+      "step": 61780
+    },
+    {
+      "epoch": 3.068938114631966,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005544889242078077,
+      "loss": 0.5461,
+      "step": 61790
+    },
+    {
+      "epoch": 3.069434786927585,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005544491904241581,
+      "loss": 0.5729,
+      "step": 61800
+    },
+    {
+      "epoch": 3.0699314592232043,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005544094566405086,
+      "loss": 0.5313,
+      "step": 61810
+    },
+    {
+      "epoch": 3.070428131518824,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005543697228568591,
+      "loss": 0.5505,
+      "step": 61820
+    },
+    {
+      "epoch": 3.070924803814443,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005543299890732096,
+      "loss": 0.5322,
+      "step": 61830
+    },
+    {
+      "epoch": 3.0714214761100624,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005542902552895599,
+      "loss": 0.5548,
+      "step": 61840
+    },
+    {
+      "epoch": 3.071918148405682,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005542505215059105,
+      "loss": 0.5299,
+      "step": 61850
+    },
+    {
+      "epoch": 3.0724148207013013,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005542107877222609,
+      "loss": 0.5366,
+      "step": 61860
+    },
+    {
+      "epoch": 3.0729114929969206,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005541710539386113,
+      "loss": 0.5327,
+      "step": 61870
+    },
+    {
+      "epoch": 3.07340816529254,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005541313201549618,
+      "loss": 0.5327,
+      "step": 61880
+    },
+    {
+      "epoch": 3.0739048375881595,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005540915863713122,
+      "loss": 0.563,
+      "step": 61890
+    },
+    {
+      "epoch": 3.0744015098837787,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005540518525876626,
+      "loss": 0.5256,
+      "step": 61900
+    },
+    {
+      "epoch": 3.074898182179398,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005540121188040132,
+      "loss": 0.5415,
+      "step": 61910
+    },
+    {
+      "epoch": 3.0753948544750176,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005539723850203636,
+      "loss": 0.5438,
+      "step": 61920
+    },
+    {
+      "epoch": 3.075891526770637,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000553932651236714,
+      "loss": 0.5567,
+      "step": 61930
+    },
+    {
+      "epoch": 3.076388199066256,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005538929174530645,
+      "loss": 0.5373,
+      "step": 61940
+    },
+    {
+      "epoch": 3.0768848713618753,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.000553853183669415,
+      "loss": 0.5203,
+      "step": 61950
+    },
+    {
+      "epoch": 3.077381543657495,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005538134498857654,
+      "loss": 0.5388,
+      "step": 61960
+    },
+    {
+      "epoch": 3.077878215953114,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005537737161021158,
+      "loss": 0.5174,
+      "step": 61970
+    },
+    {
+      "epoch": 3.0783748882487334,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005537339823184663,
+      "loss": 0.5629,
+      "step": 61980
+    },
+    {
+      "epoch": 3.0788715605443526,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005536942485348168,
+      "loss": 0.5578,
+      "step": 61990
+    },
+    {
+      "epoch": 3.0793682328399723,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005536545147511671,
+      "loss": 0.5396,
+      "step": 62000
+    },
+    {
+      "epoch": 3.0798649051355915,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005536147809675177,
+      "loss": 0.5275,
+      "step": 62010
+    },
+    {
+      "epoch": 3.0803615774312108,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005535750471838682,
+      "loss": 0.5266,
+      "step": 62020
+    },
+    {
+      "epoch": 3.0808582497268304,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005535353134002185,
+      "loss": 0.5339,
+      "step": 62030
+    },
+    {
+      "epoch": 3.0813549220224496,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000553495579616569,
+      "loss": 0.5053,
+      "step": 62040
+    },
+    {
+      "epoch": 3.081851594318069,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005534558458329194,
+      "loss": 0.5093,
+      "step": 62050
+    },
+    {
+      "epoch": 3.082348266613688,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005534161120492699,
+      "loss": 0.5779,
+      "step": 62060
+    },
+    {
+      "epoch": 3.0828449389093078,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005533763782656204,
+      "loss": 0.5301,
+      "step": 62070
+    },
+    {
+      "epoch": 3.083341611204927,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005533366444819708,
+      "loss": 0.5422,
+      "step": 62080
+    },
+    {
+      "epoch": 3.0838382835005462,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005532969106983212,
+      "loss": 0.5463,
+      "step": 62090
+    },
+    {
+      "epoch": 3.084334955796166,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005532571769146717,
+      "loss": 0.5811,
+      "step": 62100
+    },
+    {
+      "epoch": 3.084831628091785,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005532174431310222,
+      "loss": 0.5423,
+      "step": 62110
+    },
+    {
+      "epoch": 3.0853283003874044,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005531777093473726,
+      "loss": 0.5375,
+      "step": 62120
+    },
+    {
+      "epoch": 3.0858249726830236,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005531379755637231,
+      "loss": 0.5421,
+      "step": 62130
+    },
+    {
+      "epoch": 3.0863216449786433,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005530982417800735,
+      "loss": 0.5549,
+      "step": 62140
+    },
+    {
+      "epoch": 3.0868183172742625,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.000553058507996424,
+      "loss": 0.5281,
+      "step": 62150
+    },
+    {
+      "epoch": 3.0873149895698817,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005530187742127745,
+      "loss": 0.5695,
+      "step": 62160
+    },
+    {
+      "epoch": 3.087811661865501,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005529790404291249,
+      "loss": 0.5456,
+      "step": 62170
+    },
+    {
+      "epoch": 3.0883083341611206,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005529393066454754,
+      "loss": 0.5652,
+      "step": 62180
+    },
+    {
+      "epoch": 3.08880500645674,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005528995728618257,
+      "loss": 0.5569,
+      "step": 62190
+    },
+    {
+      "epoch": 3.089301678752359,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005528598390781762,
+      "loss": 0.5453,
+      "step": 62200
+    },
+    {
+      "epoch": 3.0897983510479787,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005528201052945268,
+      "loss": 0.5413,
+      "step": 62210
+    },
+    {
+      "epoch": 3.090295023343598,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005527803715108771,
+      "loss": 0.545,
+      "step": 62220
+    },
+    {
+      "epoch": 3.090791695639217,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005527406377272276,
+      "loss": 0.5355,
+      "step": 62230
+    },
+    {
+      "epoch": 3.0912883679348364,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.000552700903943578,
+      "loss": 0.5564,
+      "step": 62240
+    },
+    {
+      "epoch": 3.091785040230456,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005526611701599284,
+      "loss": 0.5451,
+      "step": 62250
+    },
+    {
+      "epoch": 3.0922817125260753,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.000552621436376279,
+      "loss": 0.5192,
+      "step": 62260
+    },
+    {
+      "epoch": 3.0927783848216945,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005525817025926294,
+      "loss": 0.521,
+      "step": 62270
+    },
+    {
+      "epoch": 3.093275057117314,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005525419688089799,
+      "loss": 0.5656,
+      "step": 62280
+    },
+    {
+      "epoch": 3.0937717294129334,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005525022350253303,
+      "loss": 0.5135,
+      "step": 62290
+    },
+    {
+      "epoch": 3.0942684017085527,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005524625012416807,
+      "loss": 0.5452,
+      "step": 62300
+    },
+    {
+      "epoch": 3.094765074004172,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005524227674580313,
+      "loss": 0.5428,
+      "step": 62310
+    },
+    {
+      "epoch": 3.0952617462997916,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005523830336743817,
+      "loss": 0.5487,
+      "step": 62320
+    },
+    {
+      "epoch": 3.095758418595411,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005523432998907321,
+      "loss": 0.5499,
+      "step": 62330
+    },
+    {
+      "epoch": 3.09625509089103,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005523035661070826,
+      "loss": 0.55,
+      "step": 62340
+    },
+    {
+      "epoch": 3.0967517631866492,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.000552263832323433,
+      "loss": 0.5263,
+      "step": 62350
+    },
+    {
+      "epoch": 3.097248435482269,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005522240985397835,
+      "loss": 0.5364,
+      "step": 62360
+    },
+    {
+      "epoch": 3.097745107777888,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000552184364756134,
+      "loss": 0.5405,
+      "step": 62370
+    },
+    {
+      "epoch": 3.0982417800735074,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005521446309724843,
+      "loss": 0.5298,
+      "step": 62380
+    },
+    {
+      "epoch": 3.098738452369127,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005521048971888348,
+      "loss": 0.5596,
+      "step": 62390
+    },
+    {
+      "epoch": 3.0992351246647463,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005520651634051853,
+      "loss": 0.5401,
+      "step": 62400
+    },
+    {
+      "epoch": 3.0997317969603655,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005520254296215357,
+      "loss": 0.5401,
+      "step": 62410
+    },
+    {
+      "epoch": 3.1002284692559847,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005519856958378862,
+      "loss": 0.5443,
+      "step": 62420
+    },
+    {
+      "epoch": 3.1007251415516044,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005519459620542367,
+      "loss": 0.538,
+      "step": 62430
+    },
+    {
+      "epoch": 3.1012218138472236,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005519062282705871,
+      "loss": 0.524,
+      "step": 62440
+    },
+    {
+      "epoch": 3.101718486142843,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005518664944869375,
+      "loss": 0.5423,
+      "step": 62450
+    },
+    {
+      "epoch": 3.1022151584384625,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.000551826760703288,
+      "loss": 0.5312,
+      "step": 62460
+    },
+    {
+      "epoch": 3.1027118307340817,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005517870269196385,
+      "loss": 0.5245,
+      "step": 62470
+    },
+    {
+      "epoch": 3.103208503029701,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005517472931359889,
+      "loss": 0.5601,
+      "step": 62480
+    },
+    {
+      "epoch": 3.10370517532532,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005517075593523393,
+      "loss": 0.5267,
+      "step": 62490
+    },
+    {
+      "epoch": 3.10420184762094,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005516678255686898,
+      "loss": 0.569,
+      "step": 62500
+    },
+    {
+      "epoch": 3.104698519916559,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005516280917850403,
+      "loss": 0.5465,
+      "step": 62510
+    },
+    {
+      "epoch": 3.1051951922121783,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0005515883580013907,
+      "loss": 0.5337,
+      "step": 62520
+    },
+    {
+      "epoch": 3.1056918645077976,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005515486242177412,
+      "loss": 0.5241,
+      "step": 62530
+    },
+    {
+      "epoch": 3.106188536803417,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0005515088904340916,
+      "loss": 0.5715,
+      "step": 62540
+    },
+    {
+      "epoch": 3.1066852090990364,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.000551469156650442,
+      "loss": 0.5474,
+      "step": 62550
+    },
+    {
+      "epoch": 3.1071818813946557,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005514294228667926,
+      "loss": 0.5265,
+      "step": 62560
+    },
+    {
+      "epoch": 3.1076785536902753,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.000551389689083143,
+      "loss": 0.5357,
+      "step": 62570
+    },
+    {
+      "epoch": 3.1081752259858946,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0005513499552994934,
+      "loss": 0.5471,
+      "step": 62580
+    },
+    {
+      "epoch": 3.108671898281514,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005513102215158439,
+      "loss": 0.5478,
+      "step": 62590
+    },
+    {
+      "epoch": 3.109168570577133,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005512704877321943,
+      "loss": 0.5785,
+      "step": 62600
+    },
+    {
+      "epoch": 3.1096652428727527,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005512307539485448,
+      "loss": 0.5711,
+      "step": 62610
+    },
+    {
+      "epoch": 3.110161915168372,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005511910201648953,
+      "loss": 0.533,
+      "step": 62620
+    },
+    {
+      "epoch": 3.110658587463991,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005511512863812457,
+      "loss": 0.5451,
+      "step": 62630
+    },
+    {
+      "epoch": 3.1111552597596104,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005511115525975961,
+      "loss": 0.544,
+      "step": 62640
+    },
+    {
+      "epoch": 3.11165193205523,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005510718188139465,
+      "loss": 0.529,
+      "step": 62650
+    },
+    {
+      "epoch": 3.1121486043508493,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005510320850302971,
+      "loss": 0.5538,
+      "step": 62660
+    },
+    {
+      "epoch": 3.1126452766464685,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005509923512466475,
+      "loss": 0.5443,
+      "step": 62670
+    },
+    {
+      "epoch": 3.113141948942088,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005509526174629979,
+      "loss": 0.5412,
+      "step": 62680
+    },
+    {
+      "epoch": 3.1136386212377074,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005509128836793484,
+      "loss": 0.5474,
+      "step": 62690
+    },
+    {
+      "epoch": 3.1141352935333266,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005508731498956988,
+      "loss": 0.5529,
+      "step": 62700
+    },
+    {
+      "epoch": 3.114631965828946,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005508334161120493,
+      "loss": 0.5555,
+      "step": 62710
+    },
+    {
+      "epoch": 3.1151286381245655,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005507936823283998,
+      "loss": 0.5342,
+      "step": 62720
+    },
+    {
+      "epoch": 3.1156253104201848,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005507539485447502,
+      "loss": 0.5442,
+      "step": 62730
+    },
+    {
+      "epoch": 3.116121982715804,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005507142147611006,
+      "loss": 0.5198,
+      "step": 62740
+    },
+    {
+      "epoch": 3.1166186550114237,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005506744809774511,
+      "loss": 0.5384,
+      "step": 62750
+    },
+    {
+      "epoch": 3.117115327307043,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005506347471938016,
+      "loss": 0.5622,
+      "step": 62760
+    },
+    {
+      "epoch": 3.117611999602662,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.000550595013410152,
+      "loss": 0.5312,
+      "step": 62770
+    },
+    {
+      "epoch": 3.1181086718982813,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005505552796265025,
+      "loss": 0.5747,
+      "step": 62780
+    },
+    {
+      "epoch": 3.118605344193901,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005505155458428529,
+      "loss": 0.5556,
+      "step": 62790
+    },
+    {
+      "epoch": 3.1191020164895202,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005504758120592033,
+      "loss": 0.5464,
+      "step": 62800
+    },
+    {
+      "epoch": 3.1195986887851395,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005504360782755539,
+      "loss": 0.5591,
+      "step": 62810
+    },
+    {
+      "epoch": 3.120095361080759,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005503963444919043,
+      "loss": 0.5534,
+      "step": 62820
+    },
+    {
+      "epoch": 3.1205920333763784,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005503566107082547,
+      "loss": 0.5294,
+      "step": 62830
+    },
+    {
+      "epoch": 3.1210887056719976,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005503168769246051,
+      "loss": 0.5288,
+      "step": 62840
+    },
+    {
+      "epoch": 3.121585377967617,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005502771431409556,
+      "loss": 0.5457,
+      "step": 62850
+    },
+    {
+      "epoch": 3.1220820502632365,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.000550237409357306,
+      "loss": 0.5185,
+      "step": 62860
+    },
+    {
+      "epoch": 3.1225787225588557,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005501976755736565,
+      "loss": 0.5331,
+      "step": 62870
+    },
+    {
+      "epoch": 3.123075394854475,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.000550157941790007,
+      "loss": 0.5471,
+      "step": 62880
+    },
+    {
+      "epoch": 3.123572067150094,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005501182080063575,
+      "loss": 0.5435,
+      "step": 62890
+    },
+    {
+      "epoch": 3.124068739445714,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005500784742227078,
+      "loss": 0.5571,
+      "step": 62900
+    },
+    {
+      "epoch": 3.124565411741333,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005500387404390584,
+      "loss": 0.5406,
+      "step": 62910
+    },
+    {
+      "epoch": 3.1250620840369523,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005499990066554088,
+      "loss": 0.5539,
+      "step": 62920
+    },
+    {
+      "epoch": 3.125558756332572,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005499592728717592,
+      "loss": 0.5627,
+      "step": 62930
+    },
+    {
+      "epoch": 3.126055428628191,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005499195390881097,
+      "loss": 0.5457,
+      "step": 62940
+    },
+    {
+      "epoch": 3.1265521009238104,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005498798053044601,
+      "loss": 0.5293,
+      "step": 62950
+    },
+    {
+      "epoch": 3.1270487732194296,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005498400715208105,
+      "loss": 0.5657,
+      "step": 62960
+    },
+    {
+      "epoch": 3.1275454455150493,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005498003377371611,
+      "loss": 0.5352,
+      "step": 62970
+    },
+    {
+      "epoch": 3.1280421178106685,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005497606039535115,
+      "loss": 0.5144,
+      "step": 62980
+    },
+    {
+      "epoch": 3.1285387901062878,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005497208701698619,
+      "loss": 0.5005,
+      "step": 62990
+    },
+    {
+      "epoch": 3.129035462401907,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005496811363862124,
+      "loss": 0.5394,
+      "step": 63000
+    },
+    {
+      "epoch": 3.1295321346975267,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005496414026025629,
+      "loss": 0.5577,
+      "step": 63010
+    },
+    {
+      "epoch": 3.130028806993146,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005496016688189134,
+      "loss": 0.5583,
+      "step": 63020
+    },
+    {
+      "epoch": 3.130525479288765,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005495619350352638,
+      "loss": 0.5588,
+      "step": 63030
+    },
+    {
+      "epoch": 3.131022151584385,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005495222012516142,
+      "loss": 0.5226,
+      "step": 63040
+    },
+    {
+      "epoch": 3.131518823880004,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005494824674679647,
+      "loss": 0.5239,
+      "step": 63050
+    },
+    {
+      "epoch": 3.1320154961756232,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.000549442733684315,
+      "loss": 0.5049,
+      "step": 63060
+    },
+    {
+      "epoch": 3.1325121684712425,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005494029999006656,
+      "loss": 0.5572,
+      "step": 63070
+    },
+    {
+      "epoch": 3.133008840766862,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005493632661170161,
+      "loss": 0.5536,
+      "step": 63080
+    },
+    {
+      "epoch": 3.1335055130624814,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005493235323333664,
+      "loss": 0.5535,
+      "step": 63090
+    },
+    {
+      "epoch": 3.1340021853581006,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0005492837985497169,
+      "loss": 0.5422,
+      "step": 63100
+    },
+    {
+      "epoch": 3.1344988576537203,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005492440647660673,
+      "loss": 0.569,
+      "step": 63110
+    },
+    {
+      "epoch": 3.1349955299493395,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005492043309824178,
+      "loss": 0.5438,
+      "step": 63120
+    },
+    {
+      "epoch": 3.1354922022449587,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005491645971987683,
+      "loss": 0.5527,
+      "step": 63130
+    },
+    {
+      "epoch": 3.135988874540578,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005491248634151187,
+      "loss": 0.5441,
+      "step": 63140
+    },
+    {
+      "epoch": 3.1364855468361976,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005490851296314691,
+      "loss": 0.5532,
+      "step": 63150
+    },
+    {
+      "epoch": 3.136982219131817,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005490453958478197,
+      "loss": 0.5299,
+      "step": 63160
+    },
+    {
+      "epoch": 3.137478891427436,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005490056620641701,
+      "loss": 0.5217,
+      "step": 63170
+    },
+    {
+      "epoch": 3.1379755637230557,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0005489659282805206,
+      "loss": 0.5396,
+      "step": 63180
+    },
+    {
+      "epoch": 3.138472236018675,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.000548926194496871,
+      "loss": 0.5253,
+      "step": 63190
+    },
+    {
+      "epoch": 3.138968908314294,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005488864607132214,
+      "loss": 0.5396,
+      "step": 63200
+    },
+    {
+      "epoch": 3.1394655806099134,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.000548846726929572,
+      "loss": 0.5891,
+      "step": 63210
+    },
+    {
+      "epoch": 3.139962252905533,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005488069931459224,
+      "loss": 0.5388,
+      "step": 63220
+    },
+    {
+      "epoch": 3.1404589252011523,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005487672593622728,
+      "loss": 0.5178,
+      "step": 63230
+    },
+    {
+      "epoch": 3.1409555974967716,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005487275255786233,
+      "loss": 0.5149,
+      "step": 63240
+    },
+    {
+      "epoch": 3.141452269792391,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005486877917949736,
+      "loss": 0.5216,
+      "step": 63250
+    },
+    {
+      "epoch": 3.1419489420880105,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005486480580113241,
+      "loss": 0.5156,
+      "step": 63260
+    },
+    {
+      "epoch": 3.1424456143836297,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005486083242276747,
+      "loss": 0.5361,
+      "step": 63270
+    },
+    {
+      "epoch": 3.142942286679249,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.000548568590444025,
+      "loss": 0.5553,
+      "step": 63280
+    },
+    {
+      "epoch": 3.1434389589748686,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005485288566603755,
+      "loss": 0.5313,
+      "step": 63290
+    },
+    {
+      "epoch": 3.143935631270488,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.000548489122876726,
+      "loss": 0.5264,
+      "step": 63300
+    },
+    {
+      "epoch": 3.144432303566107,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005484493890930763,
+      "loss": 0.5595,
+      "step": 63310
+    },
+    {
+      "epoch": 3.1449289758617263,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005484096553094269,
+      "loss": 0.526,
+      "step": 63320
+    },
+    {
+      "epoch": 3.145425648157346,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005483699215257773,
+      "loss": 0.5358,
+      "step": 63330
+    },
+    {
+      "epoch": 3.145922320452965,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005483301877421278,
+      "loss": 0.5369,
+      "step": 63340
+    },
+    {
+      "epoch": 3.1464189927485844,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005482904539584782,
+      "loss": 0.537,
+      "step": 63350
+    },
+    {
+      "epoch": 3.1469156650442036,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005482507201748286,
+      "loss": 0.5195,
+      "step": 63360
+    },
+    {
+      "epoch": 3.1474123373398233,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005482109863911792,
+      "loss": 0.5355,
+      "step": 63370
+    },
+    {
+      "epoch": 3.1479090096354425,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005481712526075296,
+      "loss": 0.5421,
+      "step": 63380
+    },
+    {
+      "epoch": 3.1484056819310617,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00054813151882388,
+      "loss": 0.5275,
+      "step": 63390
+    },
+    {
+      "epoch": 3.1489023542266814,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005480917850402305,
+      "loss": 0.5508,
+      "step": 63400
+    },
+    {
+      "epoch": 3.1493990265223006,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.000548052051256581,
+      "loss": 0.5531,
+      "step": 63410
+    },
+    {
+      "epoch": 3.14989569881792,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005480123174729314,
+      "loss": 0.5266,
+      "step": 63420
+    },
+    {
+      "epoch": 3.150392371113539,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005479725836892819,
+      "loss": 0.5375,
+      "step": 63430
+    },
+    {
+      "epoch": 3.1508890434091588,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005479328499056322,
+      "loss": 0.5275,
+      "step": 63440
+    },
+    {
+      "epoch": 3.151385715704778,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005478931161219827,
+      "loss": 0.5394,
+      "step": 63450
+    },
+    {
+      "epoch": 3.151882388000397,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005478533823383332,
+      "loss": 0.5159,
+      "step": 63460
+    },
+    {
+      "epoch": 3.152379060296017,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005478136485546837,
+      "loss": 0.5586,
+      "step": 63470
+    },
+    {
+      "epoch": 3.152875732591636,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005477739147710341,
+      "loss": 0.5405,
+      "step": 63480
+    },
+    {
+      "epoch": 3.1533724048872553,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005477341809873846,
+      "loss": 0.5514,
+      "step": 63490
+    },
+    {
+      "epoch": 3.1538690771828746,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.000547694447203735,
+      "loss": 0.5254,
+      "step": 63500
+    },
+    {
+      "epoch": 3.1543657494784942,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005476547134200854,
+      "loss": 0.5192,
+      "step": 63510
+    },
+    {
+      "epoch": 3.1548624217741135,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005476149796364359,
+      "loss": 0.5334,
+      "step": 63520
+    },
+    {
+      "epoch": 3.1553590940697327,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005475752458527864,
+      "loss": 0.5305,
+      "step": 63530
+    },
+    {
+      "epoch": 3.1558557663653524,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005475355120691368,
+      "loss": 0.5516,
+      "step": 63540
+    },
+    {
+      "epoch": 3.1563524386609716,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005474957782854872,
+      "loss": 0.5457,
+      "step": 63550
+    },
+    {
+      "epoch": 3.156849110956591,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005474560445018377,
+      "loss": 0.5303,
+      "step": 63560
+    },
+    {
+      "epoch": 3.15734578325221,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005474163107181882,
+      "loss": 0.5292,
+      "step": 63570
+    },
+    {
+      "epoch": 3.1578424555478297,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005473765769345386,
+      "loss": 0.5164,
+      "step": 63580
+    },
+    {
+      "epoch": 3.158339127843449,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005473368431508891,
+      "loss": 0.5831,
+      "step": 63590
+    },
+    {
+      "epoch": 3.158835800139068,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005472971093672395,
+      "loss": 0.5374,
+      "step": 63600
+    },
+    {
+      "epoch": 3.1593324724346874,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005472573755835899,
+      "loss": 0.5068,
+      "step": 63610
+    },
+    {
+      "epoch": 3.159829144730307,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005472176417999405,
+      "loss": 0.5486,
+      "step": 63620
+    },
+    {
+      "epoch": 3.1603258170259263,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005471779080162909,
+      "loss": 0.5547,
+      "step": 63630
+    },
+    {
+      "epoch": 3.1608224893215455,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005471381742326413,
+      "loss": 0.5387,
+      "step": 63640
+    },
+    {
+      "epoch": 3.161319161617165,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005470984404489918,
+      "loss": 0.5184,
+      "step": 63650
+    },
+    {
+      "epoch": 3.1618158339127844,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005470587066653422,
+      "loss": 0.5584,
+      "step": 63660
+    },
+    {
+      "epoch": 3.1623125062084037,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005470189728816927,
+      "loss": 0.5666,
+      "step": 63670
+    },
+    {
+      "epoch": 3.162809178504023,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005469792390980432,
+      "loss": 0.552,
+      "step": 63680
+    },
+    {
+      "epoch": 3.1633058507996425,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005469395053143936,
+      "loss": 0.564,
+      "step": 63690
+    },
+    {
+      "epoch": 3.1638025230952618,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.000546899771530744,
+      "loss": 0.5487,
+      "step": 63700
+    },
+    {
+      "epoch": 3.164299195390881,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005468600377470944,
+      "loss": 0.5383,
+      "step": 63710
+    },
+    {
+      "epoch": 3.1647958676865002,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.000546820303963445,
+      "loss": 0.5467,
+      "step": 63720
+    },
+    {
+      "epoch": 3.16529253998212,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005467805701797954,
+      "loss": 0.573,
+      "step": 63730
+    },
+    {
+      "epoch": 3.165789212277739,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005467408363961458,
+      "loss": 0.5421,
+      "step": 63740
+    },
+    {
+      "epoch": 3.1662858845733584,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005467011026124963,
+      "loss": 0.5125,
+      "step": 63750
+    },
+    {
+      "epoch": 3.166782556868978,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005466613688288467,
+      "loss": 0.5599,
+      "step": 63760
+    },
+    {
+      "epoch": 3.1672792291645973,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005466216350451972,
+      "loss": 0.5077,
+      "step": 63770
+    },
+    {
+      "epoch": 3.1677759014602165,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005465819012615477,
+      "loss": 0.5331,
+      "step": 63780
+    },
+    {
+      "epoch": 3.1682725737558357,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005465421674778981,
+      "loss": 0.5454,
+      "step": 63790
+    },
+    {
+      "epoch": 3.1687692460514554,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005465024336942485,
+      "loss": 0.5786,
+      "step": 63800
+    },
+    {
+      "epoch": 3.1692659183470746,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.000546462699910599,
+      "loss": 0.5214,
+      "step": 63810
+    },
+    {
+      "epoch": 3.169762590642694,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005464229661269495,
+      "loss": 0.5239,
+      "step": 63820
+    },
+    {
+      "epoch": 3.1702592629383135,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005463832323432999,
+      "loss": 0.5442,
+      "step": 63830
+    },
+    {
+      "epoch": 3.1707559352339327,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005463434985596504,
+      "loss": 0.5585,
+      "step": 63840
+    },
+    {
+      "epoch": 3.171252607529552,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005463037647760008,
+      "loss": 0.5462,
+      "step": 63850
+    },
+    {
+      "epoch": 3.171749279825171,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005462640309923512,
+      "loss": 0.5352,
+      "step": 63860
+    },
+    {
+      "epoch": 3.172245952120791,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005462242972087018,
+      "loss": 0.5169,
+      "step": 63870
+    },
+    {
+      "epoch": 3.17274262441641,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005461845634250522,
+      "loss": 0.5455,
+      "step": 63880
+    },
+    {
+      "epoch": 3.1732392967120293,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005461448296414026,
+      "loss": 0.5168,
+      "step": 63890
+    },
+    {
+      "epoch": 3.173735969007649,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005461050958577531,
+      "loss": 0.5427,
+      "step": 63900
+    },
+    {
+      "epoch": 3.174232641303268,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005460653620741035,
+      "loss": 0.5546,
+      "step": 63910
+    },
+    {
+      "epoch": 3.1747293135988874,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005460256282904541,
+      "loss": 0.5372,
+      "step": 63920
+    },
+    {
+      "epoch": 3.1752259858945067,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005459858945068044,
+      "loss": 0.5347,
+      "step": 63930
+    },
+    {
+      "epoch": 3.1757226581901263,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005459461607231549,
+      "loss": 0.5307,
+      "step": 63940
+    },
+    {
+      "epoch": 3.1762193304857456,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005459064269395054,
+      "loss": 0.5004,
+      "step": 63950
+    },
+    {
+      "epoch": 3.176716002781365,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005458666931558557,
+      "loss": 0.5764,
+      "step": 63960
+    },
+    {
+      "epoch": 3.177212675076984,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005458269593722063,
+      "loss": 0.5565,
+      "step": 63970
+    },
+    {
+      "epoch": 3.1777093473726037,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005457872255885567,
+      "loss": 0.5638,
+      "step": 63980
+    },
+    {
+      "epoch": 3.178206019668223,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005457474918049071,
+      "loss": 0.5408,
+      "step": 63990
+    },
+    {
+      "epoch": 3.178702691963842,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005457077580212576,
+      "loss": 0.5744,
+      "step": 64000
+    },
+    {
+      "epoch": 3.179199364259462,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.000545668024237608,
+      "loss": 0.5473,
+      "step": 64010
+    },
+    {
+      "epoch": 3.179696036555081,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005456282904539585,
+      "loss": 0.5372,
+      "step": 64020
+    },
+    {
+      "epoch": 3.1801927088507003,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.000545588556670309,
+      "loss": 0.5091,
+      "step": 64030
+    },
+    {
+      "epoch": 3.1806893811463195,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005455488228866594,
+      "loss": 0.5632,
+      "step": 64040
+    },
+    {
+      "epoch": 3.181186053441939,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005455090891030098,
+      "loss": 0.5374,
+      "step": 64050
+    },
+    {
+      "epoch": 3.1816827257375584,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005454693553193603,
+      "loss": 0.5519,
+      "step": 64060
+    },
+    {
+      "epoch": 3.1821793980331776,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005454296215357108,
+      "loss": 0.5184,
+      "step": 64070
+    },
+    {
+      "epoch": 3.182676070328797,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0005453898877520613,
+      "loss": 0.5471,
+      "step": 64080
+    },
+    {
+      "epoch": 3.1831727426244165,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0005453501539684117,
+      "loss": 0.5274,
+      "step": 64090
+    },
+    {
+      "epoch": 3.1836694149200357,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005453104201847621,
+      "loss": 0.5229,
+      "step": 64100
+    },
+    {
+      "epoch": 3.184166087215655,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005452706864011126,
+      "loss": 0.5344,
+      "step": 64110
+    },
+    {
+      "epoch": 3.1846627595112746,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.000545230952617463,
+      "loss": 0.5281,
+      "step": 64120
+    },
+    {
+      "epoch": 3.185159431806894,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005451912188338135,
+      "loss": 0.5332,
+      "step": 64130
+    },
+    {
+      "epoch": 3.185656104102513,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.000545151485050164,
+      "loss": 0.5792,
+      "step": 64140
+    },
+    {
+      "epoch": 3.1861527763981323,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005451117512665143,
+      "loss": 0.5684,
+      "step": 64150
+    },
+    {
+      "epoch": 3.186649448693752,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005450720174828648,
+      "loss": 0.5583,
+      "step": 64160
+    },
+    {
+      "epoch": 3.1871461209893712,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005450322836992154,
+      "loss": 0.5471,
+      "step": 64170
+    },
+    {
+      "epoch": 3.1876427932849905,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005449925499155657,
+      "loss": 0.5191,
+      "step": 64180
+    },
+    {
+      "epoch": 3.18813946558061,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005449528161319162,
+      "loss": 0.5254,
+      "step": 64190
+    },
+    {
+      "epoch": 3.1886361378762293,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005449130823482666,
+      "loss": 0.5393,
+      "step": 64200
+    },
+    {
+      "epoch": 3.1891328101718486,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.000544873348564617,
+      "loss": 0.5409,
+      "step": 64210
+    },
+    {
+      "epoch": 3.189629482467468,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005448336147809676,
+      "loss": 0.5351,
+      "step": 64220
+    },
+    {
+      "epoch": 3.1901261547630875,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.000544793880997318,
+      "loss": 0.5417,
+      "step": 64230
+    },
+    {
+      "epoch": 3.1906228270587067,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005447541472136685,
+      "loss": 0.5297,
+      "step": 64240
+    },
+    {
+      "epoch": 3.191119499354326,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005447144134300189,
+      "loss": 0.5272,
+      "step": 64250
+    },
+    {
+      "epoch": 3.1916161716499456,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005446746796463693,
+      "loss": 0.5497,
+      "step": 64260
+    },
+    {
+      "epoch": 3.192112843945565,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005446349458627199,
+      "loss": 0.5246,
+      "step": 64270
+    },
+    {
+      "epoch": 3.192609516241184,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005445952120790703,
+      "loss": 0.5257,
+      "step": 64280
+    },
+    {
+      "epoch": 3.1931061885368033,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005445554782954207,
+      "loss": 0.5019,
+      "step": 64290
+    },
+    {
+      "epoch": 3.193602860832423,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005445157445117712,
+      "loss": 0.5745,
+      "step": 64300
+    },
+    {
+      "epoch": 3.194099533128042,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005444760107281215,
+      "loss": 0.5407,
+      "step": 64310
+    },
+    {
+      "epoch": 3.1945962054236614,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.000544436276944472,
+      "loss": 0.5295,
+      "step": 64320
+    },
+    {
+      "epoch": 3.1950928777192806,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005443965431608226,
+      "loss": 0.5248,
+      "step": 64330
+    },
+    {
+      "epoch": 3.1955895500149003,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005443568093771729,
+      "loss": 0.5247,
+      "step": 64340
+    },
+    {
+      "epoch": 3.1960862223105195,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005443170755935234,
+      "loss": 0.5307,
+      "step": 64350
+    },
+    {
+      "epoch": 3.1965828946061388,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005442773418098739,
+      "loss": 0.5546,
+      "step": 64360
+    },
+    {
+      "epoch": 3.197079566901758,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005442376080262244,
+      "loss": 0.5415,
+      "step": 64370
+    },
+    {
+      "epoch": 3.1975762391973777,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005441978742425748,
+      "loss": 0.5384,
+      "step": 64380
+    },
+    {
+      "epoch": 3.198072911492997,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005441581404589252,
+      "loss": 0.5277,
+      "step": 64390
+    },
+    {
+      "epoch": 3.198569583788616,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005441184066752757,
+      "loss": 0.5105,
+      "step": 64400
+    },
+    {
+      "epoch": 3.199066256084236,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005440786728916261,
+      "loss": 0.526,
+      "step": 64410
+    },
+    {
+      "epoch": 3.199562928379855,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005440389391079765,
+      "loss": 0.5225,
+      "step": 64420
+    },
+    {
+      "epoch": 3.2000596006754742,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005439992053243271,
+      "loss": 0.5477,
+      "step": 64430
+    },
+    {
+      "epoch": 3.2005562729710935,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005439594715406775,
+      "loss": 0.5536,
+      "step": 64440
+    },
+    {
+      "epoch": 3.201052945266713,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005439197377570279,
+      "loss": 0.5606,
+      "step": 64450
+    },
+    {
+      "epoch": 3.2015496175623324,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005438800039733784,
+      "loss": 0.5213,
+      "step": 64460
+    },
+    {
+      "epoch": 3.2020462898579516,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005438402701897289,
+      "loss": 0.5197,
+      "step": 64470
+    },
+    {
+      "epoch": 3.2025429621535713,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005438005364060793,
+      "loss": 0.5368,
+      "step": 64480
+    },
+    {
+      "epoch": 3.2030396344491905,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005437608026224298,
+      "loss": 0.5127,
+      "step": 64490
+    },
+    {
+      "epoch": 3.2035363067448097,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005437210688387802,
+      "loss": 0.5114,
+      "step": 64500
+    },
+    {
+      "epoch": 3.204032979040429,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005436813350551306,
+      "loss": 0.5162,
+      "step": 64510
+    },
+    {
+      "epoch": 3.2045296513360486,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005436416012714812,
+      "loss": 0.5653,
+      "step": 64520
+    },
+    {
+      "epoch": 3.205026323631668,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005436018674878316,
+      "loss": 0.56,
+      "step": 64530
+    },
+    {
+      "epoch": 3.205522995927287,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.000543562133704182,
+      "loss": 0.5731,
+      "step": 64540
+    },
+    {
+      "epoch": 3.2060196682229067,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005435223999205325,
+      "loss": 0.5533,
+      "step": 64550
+    },
+    {
+      "epoch": 3.206516340518526,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005434826661368829,
+      "loss": 0.5644,
+      "step": 64560
+    },
+    {
+      "epoch": 3.207013012814145,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005434429323532333,
+      "loss": 0.5294,
+      "step": 64570
+    },
+    {
+      "epoch": 3.2075096851097644,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005434031985695838,
+      "loss": 0.5623,
+      "step": 64580
+    },
+    {
+      "epoch": 3.208006357405384,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005433634647859343,
+      "loss": 0.5674,
+      "step": 64590
+    },
+    {
+      "epoch": 3.2085030297010033,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005433237310022847,
+      "loss": 0.5413,
+      "step": 64600
+    },
+    {
+      "epoch": 3.2089997019966225,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005432839972186351,
+      "loss": 0.517,
+      "step": 64610
+    },
+    {
+      "epoch": 3.2094963742922418,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005432442634349856,
+      "loss": 0.5401,
+      "step": 64620
+    },
+    {
+      "epoch": 3.2099930465878614,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005432045296513361,
+      "loss": 0.5611,
+      "step": 64630
+    },
+    {
+      "epoch": 3.2104897188834807,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005431647958676865,
+      "loss": 0.5207,
+      "step": 64640
+    },
+    {
+      "epoch": 3.2109863911791,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.000543125062084037,
+      "loss": 0.5277,
+      "step": 64650
+    },
+    {
+      "epoch": 3.2114830634747196,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005430853283003874,
+      "loss": 0.5313,
+      "step": 64660
+    },
+    {
+      "epoch": 3.211979735770339,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0005430455945167378,
+      "loss": 0.5264,
+      "step": 64670
+    },
+    {
+      "epoch": 3.212476408065958,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005430058607330884,
+      "loss": 0.5481,
+      "step": 64680
+    },
+    {
+      "epoch": 3.2129730803615772,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005429661269494388,
+      "loss": 0.5533,
+      "step": 64690
+    },
+    {
+      "epoch": 3.213469752657197,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005429263931657892,
+      "loss": 0.5321,
+      "step": 64700
+    },
+    {
+      "epoch": 3.213966424952816,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005428866593821397,
+      "loss": 0.5544,
+      "step": 64710
+    },
+    {
+      "epoch": 3.2144630972484354,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005428469255984901,
+      "loss": 0.5066,
+      "step": 64720
+    },
+    {
+      "epoch": 3.2149597695440546,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005428071918148406,
+      "loss": 0.5334,
+      "step": 64730
+    },
+    {
+      "epoch": 3.2154564418396743,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005427674580311911,
+      "loss": 0.5496,
+      "step": 64740
+    },
+    {
+      "epoch": 3.2159531141352935,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005427277242475415,
+      "loss": 0.5501,
+      "step": 64750
+    },
+    {
+      "epoch": 3.2164497864309127,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005426879904638919,
+      "loss": 0.5431,
+      "step": 64760
+    },
+    {
+      "epoch": 3.2169464587265324,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005426482566802424,
+      "loss": 0.5598,
+      "step": 64770
+    },
+    {
+      "epoch": 3.2174431310221516,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005426085228965929,
+      "loss": 0.5536,
+      "step": 64780
+    },
+    {
+      "epoch": 3.217939803317771,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005425687891129433,
+      "loss": 0.5343,
+      "step": 64790
+    },
+    {
+      "epoch": 3.21843647561339,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005425290553292937,
+      "loss": 0.5431,
+      "step": 64800
+    },
+    {
+      "epoch": 3.2189331479090098,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005424893215456442,
+      "loss": 0.5393,
+      "step": 64810
+    },
+    {
+      "epoch": 3.219429820204629,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005424495877619948,
+      "loss": 0.5208,
+      "step": 64820
+    },
+    {
+      "epoch": 3.219926492500248,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0005424098539783451,
+      "loss": 0.5303,
+      "step": 64830
+    },
+    {
+      "epoch": 3.220423164795868,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005423701201946956,
+      "loss": 0.5394,
+      "step": 64840
+    },
+    {
+      "epoch": 3.220919837091487,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.000542330386411046,
+      "loss": 0.5353,
+      "step": 64850
+    },
+    {
+      "epoch": 3.2214165093871063,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005422906526273964,
+      "loss": 0.5584,
+      "step": 64860
+    },
+    {
+      "epoch": 3.2219131816827256,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005422509188437469,
+      "loss": 0.5353,
+      "step": 64870
+    },
+    {
+      "epoch": 3.2224098539783452,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005422111850600974,
+      "loss": 0.524,
+      "step": 64880
+    },
+    {
+      "epoch": 3.2229065262739645,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005421714512764478,
+      "loss": 0.5158,
+      "step": 64890
+    },
+    {
+      "epoch": 3.2234031985695837,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005421317174927983,
+      "loss": 0.537,
+      "step": 64900
+    },
+    {
+      "epoch": 3.2238998708652034,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005420919837091487,
+      "loss": 0.511,
+      "step": 64910
+    },
+    {
+      "epoch": 3.2243965431608226,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005420522499254991,
+      "loss": 0.5867,
+      "step": 64920
+    },
+    {
+      "epoch": 3.224893215456442,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005420125161418497,
+      "loss": 0.5472,
+      "step": 64930
+    },
+    {
+      "epoch": 3.225389887752061,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005419727823582001,
+      "loss": 0.5355,
+      "step": 64940
+    },
+    {
+      "epoch": 3.2258865600476807,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005419330485745505,
+      "loss": 0.5414,
+      "step": 64950
+    },
+    {
+      "epoch": 3.2263832323433,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.000541893314790901,
+      "loss": 0.5407,
+      "step": 64960
+    },
+    {
+      "epoch": 3.226879904638919,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005418535810072514,
+      "loss": 0.5396,
+      "step": 64970
+    },
+    {
+      "epoch": 3.2273765769345384,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.000541813847223602,
+      "loss": 0.5581,
+      "step": 64980
+    },
+    {
+      "epoch": 3.227873249230158,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005417741134399523,
+      "loss": 0.5349,
+      "step": 64990
+    },
+    {
+      "epoch": 3.2283699215257773,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0005417343796563028,
+      "loss": 0.5391,
+      "step": 65000
+    },
+    {
+      "epoch": 3.2288665938213965,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005416946458726533,
+      "loss": 0.5329,
+      "step": 65010
+    },
+    {
+      "epoch": 3.229363266117016,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005416549120890036,
+      "loss": 0.5257,
+      "step": 65020
+    },
+    {
+      "epoch": 3.2298599384126354,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005416151783053542,
+      "loss": 0.5322,
+      "step": 65030
+    },
+    {
+      "epoch": 3.2303566107082546,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005415754445217047,
+      "loss": 0.5467,
+      "step": 65040
+    },
+    {
+      "epoch": 3.230853283003874,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.000541535710738055,
+      "loss": 0.5571,
+      "step": 65050
+    },
+    {
+      "epoch": 3.2313499552994935,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005414959769544055,
+      "loss": 0.5597,
+      "step": 65060
+    },
+    {
+      "epoch": 3.2318466275951128,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005414562431707559,
+      "loss": 0.5388,
+      "step": 65070
+    },
+    {
+      "epoch": 3.232343299890732,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005414165093871064,
+      "loss": 0.5341,
+      "step": 65080
+    },
+    {
+      "epoch": 3.232839972186351,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005413767756034569,
+      "loss": 0.5413,
+      "step": 65090
+    },
+    {
+      "epoch": 3.233336644481971,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005413370418198073,
+      "loss": 0.5701,
+      "step": 65100
+    },
+    {
+      "epoch": 3.23383331677759,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005412973080361578,
+      "loss": 0.5333,
+      "step": 65110
+    },
+    {
+      "epoch": 3.2343299890732093,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005412575742525082,
+      "loss": 0.5367,
+      "step": 65120
+    },
+    {
+      "epoch": 3.234826661368829,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005412178404688587,
+      "loss": 0.5415,
+      "step": 65130
+    },
+    {
+      "epoch": 3.2353233336644482,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005411781066852092,
+      "loss": 0.5212,
+      "step": 65140
+    },
+    {
+      "epoch": 3.2358200059600675,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005411383729015596,
+      "loss": 0.529,
+      "step": 65150
+    },
+    {
+      "epoch": 3.2363166782556867,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.00054109863911791,
+      "loss": 0.5667,
+      "step": 65160
+    },
+    {
+      "epoch": 3.2368133505513064,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005410589053342605,
+      "loss": 0.5229,
+      "step": 65170
+    },
+    {
+      "epoch": 3.2373100228469256,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005410191715506109,
+      "loss": 0.5514,
+      "step": 65180
+    },
+    {
+      "epoch": 3.237806695142545,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005409794377669614,
+      "loss": 0.5677,
+      "step": 65190
+    },
+    {
+      "epoch": 3.2383033674381645,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005409397039833119,
+      "loss": 0.5259,
+      "step": 65200
+    },
+    {
+      "epoch": 3.2388000397337837,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005408999701996622,
+      "loss": 0.5209,
+      "step": 65210
+    },
+    {
+      "epoch": 3.239296712029403,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005408602364160127,
+      "loss": 0.532,
+      "step": 65220
+    },
+    {
+      "epoch": 3.239793384325022,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005408205026323633,
+      "loss": 0.5373,
+      "step": 65230
+    },
+    {
+      "epoch": 3.240290056620642,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005407807688487136,
+      "loss": 0.5476,
+      "step": 65240
+    },
+    {
+      "epoch": 3.240786728916261,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005407410350650641,
+      "loss": 0.5462,
+      "step": 65250
+    },
+    {
+      "epoch": 3.2412834012118803,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005407013012814145,
+      "loss": 0.5377,
+      "step": 65260
+    },
+    {
+      "epoch": 3.2417800735075,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.000540661567497765,
+      "loss": 0.557,
+      "step": 65270
+    },
+    {
+      "epoch": 3.242276745803119,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005406218337141155,
+      "loss": 0.5376,
+      "step": 65280
+    },
+    {
+      "epoch": 3.2427734180987384,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005405820999304659,
+      "loss": 0.5319,
+      "step": 65290
+    },
+    {
+      "epoch": 3.2432700903943577,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005405423661468164,
+      "loss": 0.536,
+      "step": 65300
+    },
+    {
+      "epoch": 3.2437667626899773,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005405026323631668,
+      "loss": 0.5568,
+      "step": 65310
+    },
+    {
+      "epoch": 3.2442634349855966,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005404628985795172,
+      "loss": 0.5463,
+      "step": 65320
+    },
+    {
+      "epoch": 3.2447601072812158,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005404231647958678,
+      "loss": 0.5409,
+      "step": 65330
+    },
+    {
+      "epoch": 3.245256779576835,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005403834310122182,
+      "loss": 0.5473,
+      "step": 65340
+    },
+    {
+      "epoch": 3.2457534518724547,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005403436972285686,
+      "loss": 0.5728,
+      "step": 65350
+    },
+    {
+      "epoch": 3.246250124168074,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005403039634449191,
+      "loss": 0.5339,
+      "step": 65360
+    },
+    {
+      "epoch": 3.246746796463693,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005402642296612694,
+      "loss": 0.57,
+      "step": 65370
+    },
+    {
+      "epoch": 3.247243468759313,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00054022449587762,
+      "loss": 0.5259,
+      "step": 65380
+    },
+    {
+      "epoch": 3.247740141054932,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005401847620939705,
+      "loss": 0.5278,
+      "step": 65390
+    },
+    {
+      "epoch": 3.2482368133505513,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005401450283103208,
+      "loss": 0.5341,
+      "step": 65400
+    },
+    {
+      "epoch": 3.2487334856461705,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005401052945266713,
+      "loss": 0.5482,
+      "step": 65410
+    },
+    {
+      "epoch": 3.24923015794179,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005400655607430218,
+      "loss": 0.5626,
+      "step": 65420
+    },
+    {
+      "epoch": 3.2497268302374094,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005400258269593723,
+      "loss": 0.5739,
+      "step": 65430
+    },
+    {
+      "epoch": 3.2502235025330286,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005399860931757227,
+      "loss": 0.5266,
+      "step": 65440
+    },
+    {
+      "epoch": 3.250720174828648,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005399463593920731,
+      "loss": 0.541,
+      "step": 65450
+    },
+    {
+      "epoch": 3.2512168471242675,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005399066256084236,
+      "loss": 0.5476,
+      "step": 65460
+    },
+    {
+      "epoch": 3.2517135194198867,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.000539866891824774,
+      "loss": 0.5393,
+      "step": 65470
+    },
+    {
+      "epoch": 3.252210191715506,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005398271580411245,
+      "loss": 0.5299,
+      "step": 65480
+    },
+    {
+      "epoch": 3.2527068640111256,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.000539787424257475,
+      "loss": 0.5222,
+      "step": 65490
+    },
+    {
+      "epoch": 3.253203536306745,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005397476904738254,
+      "loss": 0.544,
+      "step": 65500
+    },
+    {
+      "epoch": 3.253700208602364,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005397079566901758,
+      "loss": 0.547,
+      "step": 65510
+    },
+    {
+      "epoch": 3.2541968808979833,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005396682229065263,
+      "loss": 0.5586,
+      "step": 65520
+    },
+    {
+      "epoch": 3.254693553193603,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005396284891228768,
+      "loss": 0.55,
+      "step": 65530
+    },
+    {
+      "epoch": 3.255190225489222,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005395887553392272,
+      "loss": 0.5558,
+      "step": 65540
+    },
+    {
+      "epoch": 3.2556868977848414,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0005395490215555777,
+      "loss": 0.5421,
+      "step": 65550
+    },
+    {
+      "epoch": 3.256183570080461,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005395092877719281,
+      "loss": 0.5712,
+      "step": 65560
+    },
+    {
+      "epoch": 3.2566802423760803,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005394695539882785,
+      "loss": 0.5159,
+      "step": 65570
+    },
+    {
+      "epoch": 3.2571769146716996,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005394298202046291,
+      "loss": 0.545,
+      "step": 65580
+    },
+    {
+      "epoch": 3.257673586967319,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005393900864209795,
+      "loss": 0.535,
+      "step": 65590
+    },
+    {
+      "epoch": 3.2581702592629385,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005393503526373299,
+      "loss": 0.5352,
+      "step": 65600
+    },
+    {
+      "epoch": 3.2586669315585577,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005393106188536804,
+      "loss": 0.5546,
+      "step": 65610
+    },
+    {
+      "epoch": 3.259163603854177,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005392708850700308,
+      "loss": 0.5228,
+      "step": 65620
+    },
+    {
+      "epoch": 3.2596602761497966,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005392311512863813,
+      "loss": 0.5123,
+      "step": 65630
+    },
+    {
+      "epoch": 3.260156948445416,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005391914175027317,
+      "loss": 0.5049,
+      "step": 65640
+    },
+    {
+      "epoch": 3.260653620741035,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005391516837190822,
+      "loss": 0.5455,
+      "step": 65650
+    },
+    {
+      "epoch": 3.2611502930366543,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005391119499354326,
+      "loss": 0.5342,
+      "step": 65660
+    },
+    {
+      "epoch": 3.261646965332274,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.000539072216151783,
+      "loss": 0.5565,
+      "step": 65670
+    },
+    {
+      "epoch": 3.262143637627893,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005390324823681336,
+      "loss": 0.5447,
+      "step": 65680
+    },
+    {
+      "epoch": 3.2626403099235124,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.000538992748584484,
+      "loss": 0.5521,
+      "step": 65690
+    },
+    {
+      "epoch": 3.263136982219132,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005389530148008344,
+      "loss": 0.5454,
+      "step": 65700
+    },
+    {
+      "epoch": 3.2636336545147513,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0005389132810171849,
+      "loss": 0.5164,
+      "step": 65710
+    },
+    {
+      "epoch": 3.2641303268103705,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005388735472335353,
+      "loss": 0.5159,
+      "step": 65720
+    },
+    {
+      "epoch": 3.2646269991059897,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005388338134498857,
+      "loss": 0.5367,
+      "step": 65730
+    },
+    {
+      "epoch": 3.265123671401609,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005387940796662363,
+      "loss": 0.5297,
+      "step": 65740
+    },
+    {
+      "epoch": 3.2656203436972286,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005387543458825867,
+      "loss": 0.5748,
+      "step": 65750
+    },
+    {
+      "epoch": 3.266117015992848,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005387146120989371,
+      "loss": 0.5504,
+      "step": 65760
+    },
+    {
+      "epoch": 3.266613688288467,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005386748783152876,
+      "loss": 0.4994,
+      "step": 65770
+    },
+    {
+      "epoch": 3.2671103605840868,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.000538635144531638,
+      "loss": 0.5118,
+      "step": 65780
+    },
+    {
+      "epoch": 3.267607032879706,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005385954107479885,
+      "loss": 0.529,
+      "step": 65790
+    },
+    {
+      "epoch": 3.2681037051753252,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.000538555676964339,
+      "loss": 0.5371,
+      "step": 65800
+    },
+    {
+      "epoch": 3.2686003774709445,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005385159431806894,
+      "loss": 0.5454,
+      "step": 65810
+    },
+    {
+      "epoch": 3.269097049766564,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005384762093970398,
+      "loss": 0.5322,
+      "step": 65820
+    },
+    {
+      "epoch": 3.2695937220621833,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005384364756133904,
+      "loss": 0.561,
+      "step": 65830
+    },
+    {
+      "epoch": 3.2700903943578026,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005383967418297408,
+      "loss": 0.5195,
+      "step": 65840
+    },
+    {
+      "epoch": 3.2705870666534222,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005383570080460912,
+      "loss": 0.5318,
+      "step": 65850
+    },
+    {
+      "epoch": 3.2710837389490415,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005383172742624416,
+      "loss": 0.5518,
+      "step": 65860
+    },
+    {
+      "epoch": 3.2715804112446607,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005382775404787921,
+      "loss": 0.5417,
+      "step": 65870
+    },
+    {
+      "epoch": 3.27207708354028,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005382378066951427,
+      "loss": 0.5254,
+      "step": 65880
+    },
+    {
+      "epoch": 3.2725737558358996,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.000538198072911493,
+      "loss": 0.5294,
+      "step": 65890
+    },
+    {
+      "epoch": 3.273070428131519,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005381583391278435,
+      "loss": 0.5571,
+      "step": 65900
+    },
+    {
+      "epoch": 3.273567100427138,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005381186053441939,
+      "loss": 0.5311,
+      "step": 65910
+    },
+    {
+      "epoch": 3.2740637727227577,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005380788715605443,
+      "loss": 0.5349,
+      "step": 65920
+    },
+    {
+      "epoch": 3.274560445018377,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005380391377768948,
+      "loss": 0.5339,
+      "step": 65930
+    },
+    {
+      "epoch": 3.275057117313996,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005379994039932453,
+      "loss": 0.5454,
+      "step": 65940
+    },
+    {
+      "epoch": 3.2755537896096154,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005379596702095957,
+      "loss": 0.5472,
+      "step": 65950
+    },
+    {
+      "epoch": 3.276050461905235,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0005379199364259462,
+      "loss": 0.5065,
+      "step": 65960
+    },
+    {
+      "epoch": 3.2765471342008543,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005378802026422966,
+      "loss": 0.5536,
+      "step": 65970
+    },
+    {
+      "epoch": 3.2770438064964735,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.000537840468858647,
+      "loss": 0.5362,
+      "step": 65980
+    },
+    {
+      "epoch": 3.277540478792093,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005378007350749976,
+      "loss": 0.5506,
+      "step": 65990
+    },
+    {
+      "epoch": 3.2780371510877124,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.000537761001291348,
+      "loss": 0.5296,
+      "step": 66000
+    },
+    {
+      "epoch": 3.2785338233833317,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005377212675076985,
+      "loss": 0.5581,
+      "step": 66010
+    },
+    {
+      "epoch": 3.279030495678951,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005376815337240489,
+      "loss": 0.5496,
+      "step": 66020
+    },
+    {
+      "epoch": 3.2795271679745706,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005376417999403993,
+      "loss": 0.5502,
+      "step": 66030
+    },
+    {
+      "epoch": 3.28002384027019,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0005376020661567499,
+      "loss": 0.5442,
+      "step": 66040
+    },
+    {
+      "epoch": 3.280520512565809,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005375623323731002,
+      "loss": 0.5387,
+      "step": 66050
+    },
+    {
+      "epoch": 3.2810171848614282,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005375225985894507,
+      "loss": 0.5739,
+      "step": 66060
+    },
+    {
+      "epoch": 3.281513857157048,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005374828648058012,
+      "loss": 0.5438,
+      "step": 66070
+    },
+    {
+      "epoch": 3.282010529452667,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005374431310221515,
+      "loss": 0.5437,
+      "step": 66080
+    },
+    {
+      "epoch": 3.2825072017482864,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005374033972385021,
+      "loss": 0.5223,
+      "step": 66090
+    },
+    {
+      "epoch": 3.2830038740439056,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005373636634548526,
+      "loss": 0.5179,
+      "step": 66100
+    },
+    {
+      "epoch": 3.2835005463395253,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005373239296712029,
+      "loss": 0.5414,
+      "step": 66110
+    },
+    {
+      "epoch": 3.2839972186351445,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005372841958875534,
+      "loss": 0.5291,
+      "step": 66120
+    },
+    {
+      "epoch": 3.2844938909307637,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005372444621039038,
+      "loss": 0.5523,
+      "step": 66130
+    },
+    {
+      "epoch": 3.2849905632263834,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005372047283202543,
+      "loss": 0.5257,
+      "step": 66140
+    },
+    {
+      "epoch": 3.2854872355220026,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0005371649945366048,
+      "loss": 0.5845,
+      "step": 66150
+    },
+    {
+      "epoch": 3.285983907817622,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005371252607529552,
+      "loss": 0.5372,
+      "step": 66160
+    },
+    {
+      "epoch": 3.286480580113241,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005370855269693057,
+      "loss": 0.5626,
+      "step": 66170
+    },
+    {
+      "epoch": 3.2869772524088607,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005370457931856561,
+      "loss": 0.5303,
+      "step": 66180
+    },
+    {
+      "epoch": 3.28747392470448,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005370060594020066,
+      "loss": 0.5337,
+      "step": 66190
+    },
+    {
+      "epoch": 3.287970597000099,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005369663256183571,
+      "loss": 0.5547,
+      "step": 66200
+    },
+    {
+      "epoch": 3.288467269295719,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005369265918347075,
+      "loss": 0.5274,
+      "step": 66210
+    },
+    {
+      "epoch": 3.288963941591338,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005368868580510579,
+      "loss": 0.5411,
+      "step": 66220
+    },
+    {
+      "epoch": 3.2894606138869573,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005368471242674084,
+      "loss": 0.5588,
+      "step": 66230
+    },
+    {
+      "epoch": 3.2899572861825765,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005368073904837588,
+      "loss": 0.5181,
+      "step": 66240
+    },
+    {
+      "epoch": 3.290453958478196,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005367676567001093,
+      "loss": 0.5378,
+      "step": 66250
+    },
+    {
+      "epoch": 3.2909506307738154,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005367279229164598,
+      "loss": 0.5639,
+      "step": 66260
+    },
+    {
+      "epoch": 3.2914473030694347,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005366881891328101,
+      "loss": 0.5596,
+      "step": 66270
+    },
+    {
+      "epoch": 3.2919439753650543,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005366484553491606,
+      "loss": 0.557,
+      "step": 66280
+    },
+    {
+      "epoch": 3.2924406476606736,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005366087215655112,
+      "loss": 0.55,
+      "step": 66290
+    },
+    {
+      "epoch": 3.292937319956293,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005365689877818615,
+      "loss": 0.5637,
+      "step": 66300
+    },
+    {
+      "epoch": 3.293433992251912,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.000536529253998212,
+      "loss": 0.5482,
+      "step": 66310
+    },
+    {
+      "epoch": 3.2939306645475317,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005364895202145624,
+      "loss": 0.5404,
+      "step": 66320
+    },
+    {
+      "epoch": 3.294427336843151,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0005364497864309129,
+      "loss": 0.5382,
+      "step": 66330
+    },
+    {
+      "epoch": 3.29492400913877,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005364100526472634,
+      "loss": 0.5647,
+      "step": 66340
+    },
+    {
+      "epoch": 3.29542068143439,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005363703188636138,
+      "loss": 0.5365,
+      "step": 66350
+    },
+    {
+      "epoch": 3.295917353730009,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005363305850799643,
+      "loss": 0.5564,
+      "step": 66360
+    },
+    {
+      "epoch": 3.2964140260256283,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005362908512963147,
+      "loss": 0.531,
+      "step": 66370
+    },
+    {
+      "epoch": 3.2969106983212475,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005362511175126651,
+      "loss": 0.5241,
+      "step": 66380
+    },
+    {
+      "epoch": 3.297407370616867,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005362113837290157,
+      "loss": 0.528,
+      "step": 66390
+    },
+    {
+      "epoch": 3.2979040429124864,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005361716499453661,
+      "loss": 0.5528,
+      "step": 66400
+    },
+    {
+      "epoch": 3.2984007152081056,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005361319161617165,
+      "loss": 0.5735,
+      "step": 66410
+    },
+    {
+      "epoch": 3.298897387503725,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.000536092182378067,
+      "loss": 0.5501,
+      "step": 66420
+    },
+    {
+      "epoch": 3.2993940597993445,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005360524485944174,
+      "loss": 0.5369,
+      "step": 66430
+    },
+    {
+      "epoch": 3.2998907320949638,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005360127148107679,
+      "loss": 0.5435,
+      "step": 66440
+    },
+    {
+      "epoch": 3.300387404390583,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005359729810271184,
+      "loss": 0.5274,
+      "step": 66450
+    },
+    {
+      "epoch": 3.300884076686202,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005359332472434688,
+      "loss": 0.5446,
+      "step": 66460
+    },
+    {
+      "epoch": 3.301380748981822,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005358935134598192,
+      "loss": 0.5579,
+      "step": 66470
+    },
+    {
+      "epoch": 3.301877421277441,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005358537796761697,
+      "loss": 0.5127,
+      "step": 66480
+    },
+    {
+      "epoch": 3.3023740935730603,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005358140458925202,
+      "loss": 0.5156,
+      "step": 66490
+    },
+    {
+      "epoch": 3.30287076586868,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005357743121088706,
+      "loss": 0.5647,
+      "step": 66500
+    },
+    {
+      "epoch": 3.3033674381642992,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.000535734578325221,
+      "loss": 0.5426,
+      "step": 66510
+    },
+    {
+      "epoch": 3.3038641104599185,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005356948445415715,
+      "loss": 0.5492,
+      "step": 66520
+    },
+    {
+      "epoch": 3.3043607827555377,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005356551107579219,
+      "loss": 0.5311,
+      "step": 66530
+    },
+    {
+      "epoch": 3.3048574550511574,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005356153769742724,
+      "loss": 0.5401,
+      "step": 66540
+    },
+    {
+      "epoch": 3.3053541273467766,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005355756431906229,
+      "loss": 0.5331,
+      "step": 66550
+    },
+    {
+      "epoch": 3.305850799642396,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005355359094069733,
+      "loss": 0.5468,
+      "step": 66560
+    },
+    {
+      "epoch": 3.3063474719380155,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005354961756233237,
+      "loss": 0.5447,
+      "step": 66570
+    },
+    {
+      "epoch": 3.3068441442336347,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005354564418396742,
+      "loss": 0.5458,
+      "step": 66580
+    },
+    {
+      "epoch": 3.307340816529254,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005354167080560247,
+      "loss": 0.5301,
+      "step": 66590
+    },
+    {
+      "epoch": 3.307837488824873,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005353769742723751,
+      "loss": 0.5473,
+      "step": 66600
+    },
+    {
+      "epoch": 3.308334161120493,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0005353372404887256,
+      "loss": 0.5619,
+      "step": 66610
+    },
+    {
+      "epoch": 3.308830833416112,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.000535297506705076,
+      "loss": 0.5796,
+      "step": 66620
+    },
+    {
+      "epoch": 3.3093275057117313,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005352577729214264,
+      "loss": 0.5296,
+      "step": 66630
+    },
+    {
+      "epoch": 3.309824178007351,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000535218039137777,
+      "loss": 0.5319,
+      "step": 66640
+    },
+    {
+      "epoch": 3.31032085030297,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005351783053541274,
+      "loss": 0.5339,
+      "step": 66650
+    },
+    {
+      "epoch": 3.3108175225985894,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005351385715704778,
+      "loss": 0.535,
+      "step": 66660
+    },
+    {
+      "epoch": 3.3113141948942086,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005350988377868283,
+      "loss": 0.5615,
+      "step": 66670
+    },
+    {
+      "epoch": 3.3118108671898283,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005350591040031787,
+      "loss": 0.5534,
+      "step": 66680
+    },
+    {
+      "epoch": 3.3123075394854475,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005350193702195292,
+      "loss": 0.5532,
+      "step": 66690
+    },
+    {
+      "epoch": 3.3128042117810668,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005349796364358797,
+      "loss": 0.5475,
+      "step": 66700
+    },
+    {
+      "epoch": 3.3133008840766864,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005349399026522301,
+      "loss": 0.5217,
+      "step": 66710
+    },
+    {
+      "epoch": 3.3137975563723057,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005349001688685805,
+      "loss": 0.5623,
+      "step": 66720
+    },
+    {
+      "epoch": 3.314294228667925,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005348604350849309,
+      "loss": 0.5609,
+      "step": 66730
+    },
+    {
+      "epoch": 3.314790900963544,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005348207013012815,
+      "loss": 0.5325,
+      "step": 66740
+    },
+    {
+      "epoch": 3.315287573259164,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005347809675176319,
+      "loss": 0.516,
+      "step": 66750
+    },
+    {
+      "epoch": 3.315784245554783,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005347412337339823,
+      "loss": 0.5524,
+      "step": 66760
+    },
+    {
+      "epoch": 3.3162809178504022,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005347014999503328,
+      "loss": 0.549,
+      "step": 66770
+    },
+    {
+      "epoch": 3.3167775901460215,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005346617661666832,
+      "loss": 0.5395,
+      "step": 66780
+    },
+    {
+      "epoch": 3.317274262441641,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005346220323830337,
+      "loss": 0.5433,
+      "step": 66790
+    },
+    {
+      "epoch": 3.3177709347372604,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005345822985993842,
+      "loss": 0.526,
+      "step": 66800
+    },
+    {
+      "epoch": 3.3182676070328796,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005345425648157346,
+      "loss": 0.5638,
+      "step": 66810
+    },
+    {
+      "epoch": 3.318764279328499,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.000534502831032085,
+      "loss": 0.5345,
+      "step": 66820
+    },
+    {
+      "epoch": 3.3192609516241185,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005344630972484355,
+      "loss": 0.5446,
+      "step": 66830
+    },
+    {
+      "epoch": 3.3197576239197377,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.000534423363464786,
+      "loss": 0.5389,
+      "step": 66840
+    },
+    {
+      "epoch": 3.320254296215357,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005343836296811364,
+      "loss": 0.5606,
+      "step": 66850
+    },
+    {
+      "epoch": 3.3207509685109766,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005343438958974869,
+      "loss": 0.528,
+      "step": 66860
+    },
+    {
+      "epoch": 3.321247640806596,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005343041621138373,
+      "loss": 0.556,
+      "step": 66870
+    },
+    {
+      "epoch": 3.321744313102215,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005342644283301877,
+      "loss": 0.5319,
+      "step": 66880
+    },
+    {
+      "epoch": 3.3222409853978343,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005342246945465383,
+      "loss": 0.523,
+      "step": 66890
+    },
+    {
+      "epoch": 3.322737657693454,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005341849607628887,
+      "loss": 0.5343,
+      "step": 66900
+    },
+    {
+      "epoch": 3.323234329989073,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005341452269792392,
+      "loss": 0.5325,
+      "step": 66910
+    },
+    {
+      "epoch": 3.3237310022846924,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005341054931955895,
+      "loss": 0.5168,
+      "step": 66920
+    },
+    {
+      "epoch": 3.324227674580312,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00053406575941194,
+      "loss": 0.5139,
+      "step": 66930
+    },
+    {
+      "epoch": 3.3247243468759313,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005340260256282906,
+      "loss": 0.5396,
+      "step": 66940
+    },
+    {
+      "epoch": 3.3252210191715506,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005339862918446409,
+      "loss": 0.5139,
+      "step": 66950
+    },
+    {
+      "epoch": 3.32571769146717,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005339465580609914,
+      "loss": 0.5107,
+      "step": 66960
+    },
+    {
+      "epoch": 3.3262143637627894,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005339068242773419,
+      "loss": 0.5205,
+      "step": 66970
+    },
+    {
+      "epoch": 3.3267110360584087,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005338670904936922,
+      "loss": 0.5306,
+      "step": 66980
+    },
+    {
+      "epoch": 3.327207708354028,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005338273567100428,
+      "loss": 0.5516,
+      "step": 66990
+    },
+    {
+      "epoch": 3.3277043806496476,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005337876229263932,
+      "loss": 0.5457,
+      "step": 67000
+    },
+    {
+      "epoch": 3.328201052945267,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005337478891427436,
+      "loss": 0.5139,
+      "step": 67010
+    },
+    {
+      "epoch": 3.328697725240886,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005337081553590941,
+      "loss": 0.5391,
+      "step": 67020
+    },
+    {
+      "epoch": 3.3291943975365053,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005336684215754445,
+      "loss": 0.5683,
+      "step": 67030
+    },
+    {
+      "epoch": 3.329691069832125,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.000533628687791795,
+      "loss": 0.5214,
+      "step": 67040
+    },
+    {
+      "epoch": 3.330187742127744,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005335889540081455,
+      "loss": 0.5251,
+      "step": 67050
+    },
+    {
+      "epoch": 3.3306844144233634,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005335492202244959,
+      "loss": 0.5435,
+      "step": 67060
+    },
+    {
+      "epoch": 3.331181086718983,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005335094864408464,
+      "loss": 0.5237,
+      "step": 67070
+    },
+    {
+      "epoch": 3.3316777590146023,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005334697526571968,
+      "loss": 0.5006,
+      "step": 67080
+    },
+    {
+      "epoch": 3.3321744313102215,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005334300188735472,
+      "loss": 0.5426,
+      "step": 67090
+    },
+    {
+      "epoch": 3.3326711036058407,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005333902850898978,
+      "loss": 0.5197,
+      "step": 67100
+    },
+    {
+      "epoch": 3.3331677759014604,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005333505513062481,
+      "loss": 0.5393,
+      "step": 67110
+    },
+    {
+      "epoch": 3.3336644481970796,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005333108175225986,
+      "loss": 0.5338,
+      "step": 67120
+    },
+    {
+      "epoch": 3.334161120492699,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005332710837389491,
+      "loss": 0.554,
+      "step": 67130
+    },
+    {
+      "epoch": 3.334657792788318,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005332313499552994,
+      "loss": 0.536,
+      "step": 67140
+    },
+    {
+      "epoch": 3.3351544650839378,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00053319161617165,
+      "loss": 0.5317,
+      "step": 67150
+    },
+    {
+      "epoch": 3.335651137379557,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005331518823880005,
+      "loss": 0.5181,
+      "step": 67160
+    },
+    {
+      "epoch": 3.336147809675176,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005331121486043508,
+      "loss": 0.5273,
+      "step": 67170
+    },
+    {
+      "epoch": 3.3366444819707954,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005330724148207013,
+      "loss": 0.5372,
+      "step": 67180
+    },
+    {
+      "epoch": 3.337141154266415,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005330326810370517,
+      "loss": 0.5223,
+      "step": 67190
+    },
+    {
+      "epoch": 3.3376378265620343,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0005329929472534023,
+      "loss": 0.5198,
+      "step": 67200
+    },
+    {
+      "epoch": 3.3381344988576536,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005329532134697527,
+      "loss": 0.5214,
+      "step": 67210
+    },
+    {
+      "epoch": 3.3386311711532732,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005329134796861031,
+      "loss": 0.5321,
+      "step": 67220
+    },
+    {
+      "epoch": 3.3391278434488925,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005328737459024536,
+      "loss": 0.553,
+      "step": 67230
+    },
+    {
+      "epoch": 3.3396245157445117,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.000532834012118804,
+      "loss": 0.5799,
+      "step": 67240
+    },
+    {
+      "epoch": 3.340121188040131,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005327942783351545,
+      "loss": 0.5631,
+      "step": 67250
+    },
+    {
+      "epoch": 3.3406178603357506,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.000532754544551505,
+      "loss": 0.5336,
+      "step": 67260
+    },
+    {
+      "epoch": 3.34111453263137,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005327148107678554,
+      "loss": 0.527,
+      "step": 67270
+    },
+    {
+      "epoch": 3.341611204926989,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005326750769842058,
+      "loss": 0.5445,
+      "step": 67280
+    },
+    {
+      "epoch": 3.3421078772226087,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005326353432005564,
+      "loss": 0.5429,
+      "step": 67290
+    },
+    {
+      "epoch": 3.342604549518228,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005325956094169068,
+      "loss": 0.5717,
+      "step": 67300
+    },
+    {
+      "epoch": 3.343101221813847,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005325558756332572,
+      "loss": 0.5346,
+      "step": 67310
+    },
+    {
+      "epoch": 3.3435978941094664,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005325161418496077,
+      "loss": 0.518,
+      "step": 67320
+    },
+    {
+      "epoch": 3.344094566405086,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.000532476408065958,
+      "loss": 0.5257,
+      "step": 67330
+    },
+    {
+      "epoch": 3.3445912387007053,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005324366742823085,
+      "loss": 0.5612,
+      "step": 67340
+    },
+    {
+      "epoch": 3.3450879109963245,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005323969404986591,
+      "loss": 0.5445,
+      "step": 67350
+    },
+    {
+      "epoch": 3.345584583291944,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005323572067150095,
+      "loss": 0.5257,
+      "step": 67360
+    },
+    {
+      "epoch": 3.3460812555875634,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005323174729313599,
+      "loss": 0.5268,
+      "step": 67370
+    },
+    {
+      "epoch": 3.3465779278831826,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005322777391477103,
+      "loss": 0.5465,
+      "step": 67380
+    },
+    {
+      "epoch": 3.347074600178802,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005322380053640608,
+      "loss": 0.5473,
+      "step": 67390
+    },
+    {
+      "epoch": 3.3475712724744215,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005321982715804113,
+      "loss": 0.5397,
+      "step": 67400
+    },
+    {
+      "epoch": 3.3480679447700408,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005321585377967617,
+      "loss": 0.5467,
+      "step": 67410
+    },
+    {
+      "epoch": 3.34856461706566,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005321188040131122,
+      "loss": 0.5278,
+      "step": 67420
+    },
+    {
+      "epoch": 3.3490612893612797,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005320790702294626,
+      "loss": 0.5759,
+      "step": 67430
+    },
+    {
+      "epoch": 3.349557961656899,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.000532039336445813,
+      "loss": 0.5251,
+      "step": 67440
+    },
+    {
+      "epoch": 3.350054633952518,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005319996026621636,
+      "loss": 0.5322,
+      "step": 67450
+    },
+    {
+      "epoch": 3.3505513062481374,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.000531959868878514,
+      "loss": 0.5063,
+      "step": 67460
+    },
+    {
+      "epoch": 3.351047978543757,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005319201350948644,
+      "loss": 0.5257,
+      "step": 67470
+    },
+    {
+      "epoch": 3.3515446508393762,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0005318804013112149,
+      "loss": 0.5474,
+      "step": 67480
+    },
+    {
+      "epoch": 3.3520413231349955,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005318406675275653,
+      "loss": 0.5393,
+      "step": 67490
+    },
+    {
+      "epoch": 3.3525379954306147,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005318009337439158,
+      "loss": 0.5335,
+      "step": 67500
+    },
+    {
+      "epoch": 3.3530346677262344,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005317611999602663,
+      "loss": 0.5153,
+      "step": 67510
+    },
+    {
+      "epoch": 3.3535313400218536,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005317214661766167,
+      "loss": 0.528,
+      "step": 67520
+    },
+    {
+      "epoch": 3.354028012317473,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005316817323929671,
+      "loss": 0.5356,
+      "step": 67530
+    },
+    {
+      "epoch": 3.354524684613092,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005316419986093176,
+      "loss": 0.5734,
+      "step": 67540
+    },
+    {
+      "epoch": 3.3550213569087117,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005316022648256681,
+      "loss": 0.5431,
+      "step": 67550
+    },
+    {
+      "epoch": 3.355518029204331,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005315625310420185,
+      "loss": 0.5309,
+      "step": 67560
+    },
+    {
+      "epoch": 3.35601470149995,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.000531522797258369,
+      "loss": 0.5435,
+      "step": 67570
+    },
+    {
+      "epoch": 3.35651137379557,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005314830634747194,
+      "loss": 0.552,
+      "step": 67580
+    },
+    {
+      "epoch": 3.357008046091189,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005314433296910698,
+      "loss": 0.5352,
+      "step": 67590
+    },
+    {
+      "epoch": 3.3575047183868083,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005314035959074203,
+      "loss": 0.5418,
+      "step": 67600
+    },
+    {
+      "epoch": 3.3580013906824275,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0005313638621237708,
+      "loss": 0.5286,
+      "step": 67610
+    },
+    {
+      "epoch": 3.358498062978047,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005313241283401212,
+      "loss": 0.5356,
+      "step": 67620
+    },
+    {
+      "epoch": 3.3589947352736664,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005312843945564716,
+      "loss": 0.5358,
+      "step": 67630
+    },
+    {
+      "epoch": 3.3594914075692857,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005312446607728221,
+      "loss": 0.5482,
+      "step": 67640
+    },
+    {
+      "epoch": 3.3599880798649053,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005312049269891726,
+      "loss": 0.5531,
+      "step": 67650
+    },
+    {
+      "epoch": 3.3604847521605246,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.000531165193205523,
+      "loss": 0.5543,
+      "step": 67660
+    },
+    {
+      "epoch": 3.360981424456144,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005311254594218735,
+      "loss": 0.5504,
+      "step": 67670
+    },
+    {
+      "epoch": 3.361478096751763,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005310857256382239,
+      "loss": 0.5481,
+      "step": 67680
+    },
+    {
+      "epoch": 3.3619747690473827,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005310459918545743,
+      "loss": 0.5184,
+      "step": 67690
+    },
+    {
+      "epoch": 3.362471441343002,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005310062580709249,
+      "loss": 0.5619,
+      "step": 67700
+    },
+    {
+      "epoch": 3.362968113638621,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005309665242872753,
+      "loss": 0.5395,
+      "step": 67710
+    },
+    {
+      "epoch": 3.363464785934241,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005309267905036257,
+      "loss": 0.5428,
+      "step": 67720
+    },
+    {
+      "epoch": 3.36396145822986,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005308870567199762,
+      "loss": 0.538,
+      "step": 67730
+    },
+    {
+      "epoch": 3.3644581305254793,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005308473229363266,
+      "loss": 0.5436,
+      "step": 67740
+    },
+    {
+      "epoch": 3.3649548028210985,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005308075891526771,
+      "loss": 0.5553,
+      "step": 67750
+    },
+    {
+      "epoch": 3.365451475116718,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005307678553690276,
+      "loss": 0.5249,
+      "step": 67760
+    },
+    {
+      "epoch": 3.3659481474123374,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.000530728121585378,
+      "loss": 0.5369,
+      "step": 67770
+    },
+    {
+      "epoch": 3.3664448197079566,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005306883878017284,
+      "loss": 0.5338,
+      "step": 67780
+    },
+    {
+      "epoch": 3.3669414920035763,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005306486540180788,
+      "loss": 0.5266,
+      "step": 67790
+    },
+    {
+      "epoch": 3.3674381642991955,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005306089202344294,
+      "loss": 0.5549,
+      "step": 67800
+    },
+    {
+      "epoch": 3.3679348365948147,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005305691864507799,
+      "loss": 0.5414,
+      "step": 67810
+    },
+    {
+      "epoch": 3.368431508890434,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0005305294526671302,
+      "loss": 0.5242,
+      "step": 67820
+    },
+    {
+      "epoch": 3.368928181186053,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005304897188834807,
+      "loss": 0.5392,
+      "step": 67830
+    },
+    {
+      "epoch": 3.369424853481673,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005304499850998312,
+      "loss": 0.5174,
+      "step": 67840
+    },
+    {
+      "epoch": 3.369921525777292,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005304102513161816,
+      "loss": 0.5748,
+      "step": 67850
+    },
+    {
+      "epoch": 3.3704181980729113,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005303705175325321,
+      "loss": 0.5387,
+      "step": 67860
+    },
+    {
+      "epoch": 3.370914870368531,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005303307837488825,
+      "loss": 0.5072,
+      "step": 67870
+    },
+    {
+      "epoch": 3.37141154266415,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005302910499652329,
+      "loss": 0.5303,
+      "step": 67880
+    },
+    {
+      "epoch": 3.3719082149597694,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005302513161815834,
+      "loss": 0.5631,
+      "step": 67890
+    },
+    {
+      "epoch": 3.3724048872553887,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005302115823979339,
+      "loss": 0.5304,
+      "step": 67900
+    },
+    {
+      "epoch": 3.3729015595510083,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005301718486142843,
+      "loss": 0.5107,
+      "step": 67910
+    },
+    {
+      "epoch": 3.3733982318466276,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005301321148306348,
+      "loss": 0.5199,
+      "step": 67920
+    },
+    {
+      "epoch": 3.373894904142247,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005300923810469852,
+      "loss": 0.5309,
+      "step": 67930
+    },
+    {
+      "epoch": 3.3743915764378665,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005300526472633356,
+      "loss": 0.5479,
+      "step": 67940
+    },
+    {
+      "epoch": 3.3748882487334857,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005300129134796862,
+      "loss": 0.513,
+      "step": 67950
+    },
+    {
+      "epoch": 3.375384921029105,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005299731796960366,
+      "loss": 0.5582,
+      "step": 67960
+    },
+    {
+      "epoch": 3.375881593324724,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005299334459123871,
+      "loss": 0.5594,
+      "step": 67970
+    },
+    {
+      "epoch": 3.376378265620344,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005298937121287374,
+      "loss": 0.5227,
+      "step": 67980
+    },
+    {
+      "epoch": 3.376874937915963,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005298539783450879,
+      "loss": 0.5474,
+      "step": 67990
+    },
+    {
+      "epoch": 3.3773716102115823,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005298142445614385,
+      "loss": 0.5357,
+      "step": 68000
+    },
+    {
+      "epoch": 3.377868282507202,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005297745107777888,
+      "loss": 0.5372,
+      "step": 68010
+    },
+    {
+      "epoch": 3.378364954802821,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005297347769941393,
+      "loss": 0.5519,
+      "step": 68020
+    },
+    {
+      "epoch": 3.3788616270984404,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005296950432104898,
+      "loss": 0.5776,
+      "step": 68030
+    },
+    {
+      "epoch": 3.3793582993940596,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005296553094268401,
+      "loss": 0.537,
+      "step": 68040
+    },
+    {
+      "epoch": 3.3798549716896793,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005296155756431907,
+      "loss": 0.5529,
+      "step": 68050
+    },
+    {
+      "epoch": 3.3803516439852985,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005295758418595411,
+      "loss": 0.5543,
+      "step": 68060
+    },
+    {
+      "epoch": 3.3808483162809178,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005295361080758915,
+      "loss": 0.5728,
+      "step": 68070
+    },
+    {
+      "epoch": 3.3813449885765374,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.000529496374292242,
+      "loss": 0.525,
+      "step": 68080
+    },
+    {
+      "epoch": 3.3818416608721567,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005294566405085924,
+      "loss": 0.5424,
+      "step": 68090
+    },
+    {
+      "epoch": 3.382338333167776,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.000529416906724943,
+      "loss": 0.5249,
+      "step": 68100
+    },
+    {
+      "epoch": 3.382835005463395,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005293771729412934,
+      "loss": 0.559,
+      "step": 68110
+    },
+    {
+      "epoch": 3.3833316777590148,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005293374391576438,
+      "loss": 0.52,
+      "step": 68120
+    },
+    {
+      "epoch": 3.383828350054634,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005292977053739943,
+      "loss": 0.5467,
+      "step": 68130
+    },
+    {
+      "epoch": 3.3843250223502532,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005292579715903447,
+      "loss": 0.5558,
+      "step": 68140
+    },
+    {
+      "epoch": 3.384821694645873,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005292182378066952,
+      "loss": 0.554,
+      "step": 68150
+    },
+    {
+      "epoch": 3.385318366941492,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005291785040230457,
+      "loss": 0.5432,
+      "step": 68160
+    },
+    {
+      "epoch": 3.3858150392371114,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005291387702393961,
+      "loss": 0.5391,
+      "step": 68170
+    },
+    {
+      "epoch": 3.3863117115327306,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005290990364557465,
+      "loss": 0.5326,
+      "step": 68180
+    },
+    {
+      "epoch": 3.38680838382835,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.000529059302672097,
+      "loss": 0.5463,
+      "step": 68190
+    },
+    {
+      "epoch": 3.3873050561239695,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005290195688884473,
+      "loss": 0.5734,
+      "step": 68200
+    },
+    {
+      "epoch": 3.3878017284195887,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005289798351047979,
+      "loss": 0.521,
+      "step": 68210
+    },
+    {
+      "epoch": 3.388298400715208,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005289401013211484,
+      "loss": 0.5451,
+      "step": 68220
+    },
+    {
+      "epoch": 3.3887950730108276,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005289003675374987,
+      "loss": 0.5402,
+      "step": 68230
+    },
+    {
+      "epoch": 3.389291745306447,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005288606337538492,
+      "loss": 0.5204,
+      "step": 68240
+    },
+    {
+      "epoch": 3.389788417602066,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005288208999701997,
+      "loss": 0.5358,
+      "step": 68250
+    },
+    {
+      "epoch": 3.3902850898976853,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005287811661865502,
+      "loss": 0.5163,
+      "step": 68260
+    },
+    {
+      "epoch": 3.390781762193305,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005287414324029006,
+      "loss": 0.5371,
+      "step": 68270
+    },
+    {
+      "epoch": 3.391278434488924,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.000528701698619251,
+      "loss": 0.5256,
+      "step": 68280
+    },
+    {
+      "epoch": 3.3917751067845434,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005286619648356015,
+      "loss": 0.5205,
+      "step": 68290
+    },
+    {
+      "epoch": 3.392271779080163,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.000528622231051952,
+      "loss": 0.496,
+      "step": 68300
+    },
+    {
+      "epoch": 3.3927684513757823,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005285824972683024,
+      "loss": 0.5303,
+      "step": 68310
+    },
+    {
+      "epoch": 3.3932651236714015,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005285427634846529,
+      "loss": 0.5417,
+      "step": 68320
+    },
+    {
+      "epoch": 3.3937617959670208,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005285030297010033,
+      "loss": 0.5569,
+      "step": 68330
+    },
+    {
+      "epoch": 3.3942584682626404,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005284632959173537,
+      "loss": 0.5545,
+      "step": 68340
+    },
+    {
+      "epoch": 3.3947551405582597,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005284235621337043,
+      "loss": 0.5759,
+      "step": 68350
+    },
+    {
+      "epoch": 3.395251812853879,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005283838283500547,
+      "loss": 0.5187,
+      "step": 68360
+    },
+    {
+      "epoch": 3.3957484851494986,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005283440945664051,
+      "loss": 0.5299,
+      "step": 68370
+    },
+    {
+      "epoch": 3.396245157445118,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005283043607827556,
+      "loss": 0.5582,
+      "step": 68380
+    },
+    {
+      "epoch": 3.396741829740737,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005282646269991059,
+      "loss": 0.5287,
+      "step": 68390
+    },
+    {
+      "epoch": 3.3972385020363562,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005282248932154564,
+      "loss": 0.5188,
+      "step": 68400
+    },
+    {
+      "epoch": 3.397735174331976,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.000528185159431807,
+      "loss": 0.5522,
+      "step": 68410
+    },
+    {
+      "epoch": 3.398231846627595,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005281454256481574,
+      "loss": 0.5268,
+      "step": 68420
+    },
+    {
+      "epoch": 3.3987285189232144,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005281056918645078,
+      "loss": 0.5343,
+      "step": 68430
+    },
+    {
+      "epoch": 3.399225191218834,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005280659580808582,
+      "loss": 0.5571,
+      "step": 68440
+    },
+    {
+      "epoch": 3.3997218635144533,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005280262242972088,
+      "loss": 0.5443,
+      "step": 68450
+    },
+    {
+      "epoch": 3.4002185358100725,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005279864905135592,
+      "loss": 0.5339,
+      "step": 68460
+    },
+    {
+      "epoch": 3.4007152081056917,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005279467567299096,
+      "loss": 0.5359,
+      "step": 68470
+    },
+    {
+      "epoch": 3.4012118804013114,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005279070229462601,
+      "loss": 0.5478,
+      "step": 68480
+    },
+    {
+      "epoch": 3.4017085526969306,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005278672891626105,
+      "loss": 0.5385,
+      "step": 68490
+    },
+    {
+      "epoch": 3.40220522499255,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.000527827555378961,
+      "loss": 0.5467,
+      "step": 68500
+    },
+    {
+      "epoch": 3.402701897288169,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005277878215953115,
+      "loss": 0.5592,
+      "step": 68510
+    },
+    {
+      "epoch": 3.4031985695837887,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005277480878116619,
+      "loss": 0.5589,
+      "step": 68520
+    },
+    {
+      "epoch": 3.403695241879408,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005277083540280123,
+      "loss": 0.5371,
+      "step": 68530
+    },
+    {
+      "epoch": 3.404191914175027,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005276686202443628,
+      "loss": 0.5365,
+      "step": 68540
+    },
+    {
+      "epoch": 3.4046885864706464,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005276288864607132,
+      "loss": 0.5139,
+      "step": 68550
+    },
+    {
+      "epoch": 3.405185258766266,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005275891526770637,
+      "loss": 0.546,
+      "step": 68560
+    },
+    {
+      "epoch": 3.4056819310618853,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005275494188934142,
+      "loss": 0.4999,
+      "step": 68570
+    },
+    {
+      "epoch": 3.4061786033575046,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005275096851097646,
+      "loss": 0.5094,
+      "step": 68580
+    },
+    {
+      "epoch": 3.4066752756531242,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000527469951326115,
+      "loss": 0.5602,
+      "step": 68590
+    },
+    {
+      "epoch": 3.4071719479487435,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005274302175424656,
+      "loss": 0.5446,
+      "step": 68600
+    },
+    {
+      "epoch": 3.4076686202443627,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.000527390483758816,
+      "loss": 0.5177,
+      "step": 68610
+    },
+    {
+      "epoch": 3.408165292539982,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0005273507499751664,
+      "loss": 0.5689,
+      "step": 68620
+    },
+    {
+      "epoch": 3.4086619648356016,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005273110161915169,
+      "loss": 0.5237,
+      "step": 68630
+    },
+    {
+      "epoch": 3.409158637131221,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005272712824078673,
+      "loss": 0.5226,
+      "step": 68640
+    },
+    {
+      "epoch": 3.40965530942684,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0005272315486242177,
+      "loss": 0.5478,
+      "step": 68650
+    },
+    {
+      "epoch": 3.4101519817224597,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005271918148405682,
+      "loss": 0.5422,
+      "step": 68660
+    },
+    {
+      "epoch": 3.410648654018079,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005271520810569187,
+      "loss": 0.5278,
+      "step": 68670
+    },
+    {
+      "epoch": 3.411145326313698,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005271123472732691,
+      "loss": 0.5128,
+      "step": 68680
+    },
+    {
+      "epoch": 3.4116419986093174,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005270726134896195,
+      "loss": 0.5396,
+      "step": 68690
+    },
+    {
+      "epoch": 3.412138670904937,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00052703287970597,
+      "loss": 0.523,
+      "step": 68700
+    },
+    {
+      "epoch": 3.4126353432005563,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005269931459223205,
+      "loss": 0.5256,
+      "step": 68710
+    },
+    {
+      "epoch": 3.4131320154961755,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005269534121386709,
+      "loss": 0.5363,
+      "step": 68720
+    },
+    {
+      "epoch": 3.413628687791795,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005269136783550214,
+      "loss": 0.5413,
+      "step": 68730
+    },
+    {
+      "epoch": 3.4141253600874144,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005268739445713718,
+      "loss": 0.544,
+      "step": 68740
+    },
+    {
+      "epoch": 3.4146220323830336,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005268342107877222,
+      "loss": 0.541,
+      "step": 68750
+    },
+    {
+      "epoch": 3.415118704678653,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005267944770040728,
+      "loss": 0.5474,
+      "step": 68760
+    },
+    {
+      "epoch": 3.4156153769742725,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005267547432204232,
+      "loss": 0.5393,
+      "step": 68770
+    },
+    {
+      "epoch": 3.4161120492698918,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005267150094367736,
+      "loss": 0.5744,
+      "step": 68780
+    },
+    {
+      "epoch": 3.416608721565511,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005266752756531241,
+      "loss": 0.5633,
+      "step": 68790
+    },
+    {
+      "epoch": 3.4171053938611307,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005266355418694745,
+      "loss": 0.5428,
+      "step": 68800
+    },
+    {
+      "epoch": 3.41760206615675,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.000526595808085825,
+      "loss": 0.5468,
+      "step": 68810
+    },
+    {
+      "epoch": 3.418098738452369,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005265560743021755,
+      "loss": 0.5306,
+      "step": 68820
+    },
+    {
+      "epoch": 3.4185954107479883,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005265163405185259,
+      "loss": 0.5672,
+      "step": 68830
+    },
+    {
+      "epoch": 3.419092083043608,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005264766067348763,
+      "loss": 0.5292,
+      "step": 68840
+    },
+    {
+      "epoch": 3.4195887553392272,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005264368729512267,
+      "loss": 0.5451,
+      "step": 68850
+    },
+    {
+      "epoch": 3.4200854276348465,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005263971391675773,
+      "loss": 0.5447,
+      "step": 68860
+    },
+    {
+      "epoch": 3.4205820999304657,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005263574053839278,
+      "loss": 0.5387,
+      "step": 68870
+    },
+    {
+      "epoch": 3.4210787722260854,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005263176716002781,
+      "loss": 0.5183,
+      "step": 68880
+    },
+    {
+      "epoch": 3.4215754445217046,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005262779378166286,
+      "loss": 0.5336,
+      "step": 68890
+    },
+    {
+      "epoch": 3.422072116817324,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005262382040329792,
+      "loss": 0.5533,
+      "step": 68900
+    },
+    {
+      "epoch": 3.422568789112943,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005261984702493295,
+      "loss": 0.5835,
+      "step": 68910
+    },
+    {
+      "epoch": 3.4230654614085627,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00052615873646568,
+      "loss": 0.5446,
+      "step": 68920
+    },
+    {
+      "epoch": 3.423562133704182,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005261190026820304,
+      "loss": 0.548,
+      "step": 68930
+    },
+    {
+      "epoch": 3.424058805999801,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005260792688983808,
+      "loss": 0.5148,
+      "step": 68940
+    },
+    {
+      "epoch": 3.424555478295421,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005260395351147313,
+      "loss": 0.5478,
+      "step": 68950
+    },
+    {
+      "epoch": 3.42505215059104,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005259998013310818,
+      "loss": 0.541,
+      "step": 68960
+    },
+    {
+      "epoch": 3.4255488228866593,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005259600675474322,
+      "loss": 0.5277,
+      "step": 68970
+    },
+    {
+      "epoch": 3.4260454951822785,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005259203337637827,
+      "loss": 0.537,
+      "step": 68980
+    },
+    {
+      "epoch": 3.426542167477898,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005258805999801331,
+      "loss": 0.551,
+      "step": 68990
+    },
+    {
+      "epoch": 3.4270388397735174,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005258408661964836,
+      "loss": 0.5412,
+      "step": 69000
+    },
+    {
+      "epoch": 3.4275355120691366,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005258011324128341,
+      "loss": 0.5622,
+      "step": 69010
+    },
+    {
+      "epoch": 3.4280321843647563,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005257613986291845,
+      "loss": 0.5339,
+      "step": 69020
+    },
+    {
+      "epoch": 3.4285288566603755,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.000525721664845535,
+      "loss": 0.5559,
+      "step": 69030
+    },
+    {
+      "epoch": 3.4290255289559948,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005256819310618853,
+      "loss": 0.5526,
+      "step": 69040
+    },
+    {
+      "epoch": 3.429522201251614,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005256421972782358,
+      "loss": 0.5512,
+      "step": 69050
+    },
+    {
+      "epoch": 3.4300188735472337,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005256024634945864,
+      "loss": 0.5645,
+      "step": 69060
+    },
+    {
+      "epoch": 3.430515545842853,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0005255627297109367,
+      "loss": 0.538,
+      "step": 69070
+    },
+    {
+      "epoch": 3.431012218138472,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005255229959272872,
+      "loss": 0.5065,
+      "step": 69080
+    },
+    {
+      "epoch": 3.431508890434092,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005254832621436377,
+      "loss": 0.5269,
+      "step": 69090
+    },
+    {
+      "epoch": 3.432005562729711,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.000525443528359988,
+      "loss": 0.5125,
+      "step": 69100
+    },
+    {
+      "epoch": 3.4325022350253303,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005254037945763386,
+      "loss": 0.5389,
+      "step": 69110
+    },
+    {
+      "epoch": 3.4329989073209495,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.000525364060792689,
+      "loss": 0.5265,
+      "step": 69120
+    },
+    {
+      "epoch": 3.433495579616569,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005253243270090394,
+      "loss": 0.5164,
+      "step": 69130
+    },
+    {
+      "epoch": 3.4339922519121884,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005252845932253899,
+      "loss": 0.5255,
+      "step": 69140
+    },
+    {
+      "epoch": 3.4344889242078076,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005252448594417403,
+      "loss": 0.5521,
+      "step": 69150
+    },
+    {
+      "epoch": 3.4349855965034273,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005252051256580909,
+      "loss": 0.5428,
+      "step": 69160
+    },
+    {
+      "epoch": 3.4354822687990465,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0005251653918744413,
+      "loss": 0.5408,
+      "step": 69170
+    },
+    {
+      "epoch": 3.4359789410946657,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005251256580907917,
+      "loss": 0.547,
+      "step": 69180
+    },
+    {
+      "epoch": 3.436475613390285,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005250859243071422,
+      "loss": 0.5285,
+      "step": 69190
+    },
+    {
+      "epoch": 3.4369722856859046,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005250461905234926,
+      "loss": 0.5264,
+      "step": 69200
+    },
+    {
+      "epoch": 3.437468957981524,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005250064567398431,
+      "loss": 0.512,
+      "step": 69210
+    },
+    {
+      "epoch": 3.437965630277143,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005249667229561936,
+      "loss": 0.535,
+      "step": 69220
+    },
+    {
+      "epoch": 3.4384623025727623,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.000524926989172544,
+      "loss": 0.5482,
+      "step": 69230
+    },
+    {
+      "epoch": 3.438958974868382,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005248872553888944,
+      "loss": 0.5437,
+      "step": 69240
+    },
+    {
+      "epoch": 3.439455647164001,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005248475216052449,
+      "loss": 0.543,
+      "step": 69250
+    },
+    {
+      "epoch": 3.4399523194596204,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005248077878215953,
+      "loss": 0.5419,
+      "step": 69260
+    },
+    {
+      "epoch": 3.4404489917552397,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005247680540379458,
+      "loss": 0.5374,
+      "step": 69270
+    },
+    {
+      "epoch": 3.4409456640508593,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005247283202542963,
+      "loss": 0.5166,
+      "step": 69280
+    },
+    {
+      "epoch": 3.4414423363464786,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0005246885864706466,
+      "loss": 0.5189,
+      "step": 69290
+    },
+    {
+      "epoch": 3.441939008642098,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005246488526869971,
+      "loss": 0.5521,
+      "step": 69300
+    },
+    {
+      "epoch": 3.4424356809377175,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005246091189033476,
+      "loss": 0.5179,
+      "step": 69310
+    },
+    {
+      "epoch": 3.4429323532333367,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005245693851196981,
+      "loss": 0.5458,
+      "step": 69320
+    },
+    {
+      "epoch": 3.443429025528956,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005245296513360485,
+      "loss": 0.5593,
+      "step": 69330
+    },
+    {
+      "epoch": 3.443925697824575,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0005244899175523989,
+      "loss": 0.5407,
+      "step": 69340
+    },
+    {
+      "epoch": 3.444422370120195,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005244501837687494,
+      "loss": 0.5337,
+      "step": 69350
+    },
+    {
+      "epoch": 3.444919042415814,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005244104499850999,
+      "loss": 0.5273,
+      "step": 69360
+    },
+    {
+      "epoch": 3.4454157147114333,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005243707162014503,
+      "loss": 0.5145,
+      "step": 69370
+    },
+    {
+      "epoch": 3.445912387007053,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005243309824178008,
+      "loss": 0.5227,
+      "step": 69380
+    },
+    {
+      "epoch": 3.446409059302672,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005242912486341512,
+      "loss": 0.5793,
+      "step": 69390
+    },
+    {
+      "epoch": 3.4469057315982914,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005242515148505016,
+      "loss": 0.5365,
+      "step": 69400
+    },
+    {
+      "epoch": 3.4474024038939106,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005242117810668522,
+      "loss": 0.5376,
+      "step": 69410
+    },
+    {
+      "epoch": 3.4478990761895303,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005241720472832026,
+      "loss": 0.5581,
+      "step": 69420
+    },
+    {
+      "epoch": 3.4483957484851495,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.000524132313499553,
+      "loss": 0.5289,
+      "step": 69430
+    },
+    {
+      "epoch": 3.4488924207807687,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005240925797159035,
+      "loss": 0.5134,
+      "step": 69440
+    },
+    {
+      "epoch": 3.4493890930763884,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005240528459322539,
+      "loss": 0.5416,
+      "step": 69450
+    },
+    {
+      "epoch": 3.4498857653720076,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005240131121486044,
+      "loss": 0.5462,
+      "step": 69460
+    },
+    {
+      "epoch": 3.450382437667627,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005239733783649549,
+      "loss": 0.5286,
+      "step": 69470
+    },
+    {
+      "epoch": 3.450879109963246,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005239336445813053,
+      "loss": 0.5301,
+      "step": 69480
+    },
+    {
+      "epoch": 3.4513757822588658,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005238939107976557,
+      "loss": 0.5386,
+      "step": 69490
+    },
+    {
+      "epoch": 3.451872454554485,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005238541770140062,
+      "loss": 0.5514,
+      "step": 69500
+    },
+    {
+      "epoch": 3.452369126850104,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005238144432303567,
+      "loss": 0.5283,
+      "step": 69510
+    },
+    {
+      "epoch": 3.452865799145724,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005237747094467071,
+      "loss": 0.5174,
+      "step": 69520
+    },
+    {
+      "epoch": 3.453362471441343,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005237349756630575,
+      "loss": 0.53,
+      "step": 69530
+    },
+    {
+      "epoch": 3.4538591437369623,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.000523695241879408,
+      "loss": 0.5549,
+      "step": 69540
+    },
+    {
+      "epoch": 3.4543558160325816,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005236555080957584,
+      "loss": 0.5244,
+      "step": 69550
+    },
+    {
+      "epoch": 3.4548524883282012,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005236157743121089,
+      "loss": 0.5421,
+      "step": 69560
+    },
+    {
+      "epoch": 3.4553491606238205,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0005235760405284594,
+      "loss": 0.5856,
+      "step": 69570
+    },
+    {
+      "epoch": 3.4558458329194397,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005235363067448098,
+      "loss": 0.5223,
+      "step": 69580
+    },
+    {
+      "epoch": 3.456342505215059,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005234965729611602,
+      "loss": 0.5458,
+      "step": 69590
+    },
+    {
+      "epoch": 3.4568391775106786,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005234568391775107,
+      "loss": 0.5358,
+      "step": 69600
+    },
+    {
+      "epoch": 3.457335849806298,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005234171053938612,
+      "loss": 0.5253,
+      "step": 69610
+    },
+    {
+      "epoch": 3.457832522101917,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005233773716102116,
+      "loss": 0.5263,
+      "step": 69620
+    },
+    {
+      "epoch": 3.4583291943975363,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005233376378265621,
+      "loss": 0.5653,
+      "step": 69630
+    },
+    {
+      "epoch": 3.458825866693156,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005232979040429125,
+      "loss": 0.5114,
+      "step": 69640
+    },
+    {
+      "epoch": 3.459322538988775,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005232581702592629,
+      "loss": 0.5446,
+      "step": 69650
+    },
+    {
+      "epoch": 3.4598192112843944,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005232184364756135,
+      "loss": 0.5211,
+      "step": 69660
+    },
+    {
+      "epoch": 3.460315883580014,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005231787026919639,
+      "loss": 0.5337,
+      "step": 69670
+    },
+    {
+      "epoch": 3.4608125558756333,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0005231389689083143,
+      "loss": 0.5658,
+      "step": 69680
+    },
+    {
+      "epoch": 3.4613092281712525,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005230992351246648,
+      "loss": 0.5337,
+      "step": 69690
+    },
+    {
+      "epoch": 3.4618059004668718,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005230595013410152,
+      "loss": 0.5631,
+      "step": 69700
+    },
+    {
+      "epoch": 3.4623025727624914,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005230197675573656,
+      "loss": 0.5428,
+      "step": 69710
+    },
+    {
+      "epoch": 3.4627992450581107,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005229800337737161,
+      "loss": 0.5287,
+      "step": 69720
+    },
+    {
+      "epoch": 3.46329591735373,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005229402999900666,
+      "loss": 0.5648,
+      "step": 69730
+    },
+    {
+      "epoch": 3.4637925896493496,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005229005662064171,
+      "loss": 0.5414,
+      "step": 69740
+    },
+    {
+      "epoch": 3.464289261944969,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005228608324227674,
+      "loss": 0.5192,
+      "step": 69750
+    },
+    {
+      "epoch": 3.464785934240588,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.000522821098639118,
+      "loss": 0.538,
+      "step": 69760
+    },
+    {
+      "epoch": 3.4652826065362072,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005227813648554685,
+      "loss": 0.5175,
+      "step": 69770
+    },
+    {
+      "epoch": 3.465779278831827,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005227416310718188,
+      "loss": 0.548,
+      "step": 69780
+    },
+    {
+      "epoch": 3.466275951127446,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005227018972881693,
+      "loss": 0.5178,
+      "step": 69790
+    },
+    {
+      "epoch": 3.4667726234230654,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005226621635045197,
+      "loss": 0.5473,
+      "step": 69800
+    },
+    {
+      "epoch": 3.467269295718685,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005226224297208701,
+      "loss": 0.5624,
+      "step": 69810
+    },
+    {
+      "epoch": 3.4677659680143043,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005225826959372207,
+      "loss": 0.5316,
+      "step": 69820
+    },
+    {
+      "epoch": 3.4682626403099235,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005225429621535711,
+      "loss": 0.5235,
+      "step": 69830
+    },
+    {
+      "epoch": 3.4687593126055427,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005225032283699215,
+      "loss": 0.5579,
+      "step": 69840
+    },
+    {
+      "epoch": 3.4692559849011624,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.000522463494586272,
+      "loss": 0.525,
+      "step": 69850
+    },
+    {
+      "epoch": 3.4697526571967816,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005224237608026224,
+      "loss": 0.5302,
+      "step": 69860
+    },
+    {
+      "epoch": 3.470249329492401,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005223840270189729,
+      "loss": 0.5348,
+      "step": 69870
+    },
+    {
+      "epoch": 3.4707460017880205,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005223442932353234,
+      "loss": 0.5421,
+      "step": 69880
+    },
+    {
+      "epoch": 3.4712426740836397,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005223045594516738,
+      "loss": 0.5356,
+      "step": 69890
+    },
+    {
+      "epoch": 3.471739346379259,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005222648256680243,
+      "loss": 0.5201,
+      "step": 69900
+    },
+    {
+      "epoch": 3.472236018674878,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005222250918843746,
+      "loss": 0.5484,
+      "step": 69910
+    },
+    {
+      "epoch": 3.472732690970498,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005221853581007252,
+      "loss": 0.5577,
+      "step": 69920
+    },
+    {
+      "epoch": 3.473229363266117,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005221456243170757,
+      "loss": 0.5364,
+      "step": 69930
+    },
+    {
+      "epoch": 3.4737260355617363,
+      "grad_norm": 0.125,
+      "learning_rate": 0.000522105890533426,
+      "loss": 0.5182,
+      "step": 69940
+    },
+    {
+      "epoch": 3.4742227078573555,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005220661567497765,
+      "loss": 0.5141,
+      "step": 69950
+    },
+    {
+      "epoch": 3.474719380152975,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.000522026422966127,
+      "loss": 0.5548,
+      "step": 69960
+    },
+    {
+      "epoch": 3.4752160524485944,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005219866891824774,
+      "loss": 0.5866,
+      "step": 69970
+    },
+    {
+      "epoch": 3.4757127247442137,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0005219469553988279,
+      "loss": 0.503,
+      "step": 69980
+    },
+    {
+      "epoch": 3.476209397039833,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005219072216151783,
+      "loss": 0.5192,
+      "step": 69990
+    },
+    {
+      "epoch": 3.4767060693354526,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005218674878315287,
+      "loss": 0.5409,
+      "step": 70000
+    },
+    {
+      "epoch": 3.477202741631072,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005218277540478792,
+      "loss": 0.5133,
+      "step": 70010
+    },
+    {
+      "epoch": 3.477699413926691,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005217880202642297,
+      "loss": 0.5512,
+      "step": 70020
+    },
+    {
+      "epoch": 3.4781960862223107,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005217482864805801,
+      "loss": 0.5461,
+      "step": 70030
+    },
+    {
+      "epoch": 3.47869275851793,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005217085526969306,
+      "loss": 0.5825,
+      "step": 70040
+    },
+    {
+      "epoch": 3.479189430813549,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.000521668818913281,
+      "loss": 0.5304,
+      "step": 70050
+    },
+    {
+      "epoch": 3.4796861031091684,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005216290851296316,
+      "loss": 0.5333,
+      "step": 70060
+    },
+    {
+      "epoch": 3.480182775404788,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.000521589351345982,
+      "loss": 0.5195,
+      "step": 70070
+    },
+    {
+      "epoch": 3.4806794477004073,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005215496175623324,
+      "loss": 0.5215,
+      "step": 70080
+    },
+    {
+      "epoch": 3.4811761199960265,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005215098837786829,
+      "loss": 0.5377,
+      "step": 70090
+    },
+    {
+      "epoch": 3.481672792291646,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005214701499950333,
+      "loss": 0.5759,
+      "step": 70100
+    },
+    {
+      "epoch": 3.4821694645872654,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.0005214304162113837,
+      "loss": 0.5493,
+      "step": 70110
+    },
+    {
+      "epoch": 3.4826661368828846,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005213906824277343,
+      "loss": 0.5331,
+      "step": 70120
+    },
+    {
+      "epoch": 3.483162809178504,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005213509486440846,
+      "loss": 0.5143,
+      "step": 70130
+    },
+    {
+      "epoch": 3.4836594814741235,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005213112148604351,
+      "loss": 0.5208,
+      "step": 70140
+    },
+    {
+      "epoch": 3.4841561537697427,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005212714810767856,
+      "loss": 0.5189,
+      "step": 70150
+    },
+    {
+      "epoch": 3.484652826065362,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005212317472931359,
+      "loss": 0.5287,
+      "step": 70160
+    },
+    {
+      "epoch": 3.4851494983609816,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005211920135094865,
+      "loss": 0.5186,
+      "step": 70170
+    },
+    {
+      "epoch": 3.485646170656601,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005211522797258369,
+      "loss": 0.5472,
+      "step": 70180
+    },
+    {
+      "epoch": 3.48614284295222,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005211125459421874,
+      "loss": 0.5225,
+      "step": 70190
+    },
+    {
+      "epoch": 3.4866395152478393,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005210728121585378,
+      "loss": 0.5379,
+      "step": 70200
+    },
+    {
+      "epoch": 3.487136187543459,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005210330783748882,
+      "loss": 0.5436,
+      "step": 70210
+    },
+    {
+      "epoch": 3.4876328598390782,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005209933445912388,
+      "loss": 0.5352,
+      "step": 70220
+    },
+    {
+      "epoch": 3.4881295321346975,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005209536108075892,
+      "loss": 0.5112,
+      "step": 70230
+    },
+    {
+      "epoch": 3.488626204430317,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005209138770239396,
+      "loss": 0.5121,
+      "step": 70240
+    },
+    {
+      "epoch": 3.4891228767259364,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005208741432402901,
+      "loss": 0.5324,
+      "step": 70250
+    },
+    {
+      "epoch": 3.4896195490215556,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005208344094566405,
+      "loss": 0.5386,
+      "step": 70260
+    },
+    {
+      "epoch": 3.490116221317175,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.000520794675672991,
+      "loss": 0.5745,
+      "step": 70270
+    },
+    {
+      "epoch": 3.490612893612794,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005207549418893415,
+      "loss": 0.5429,
+      "step": 70280
+    },
+    {
+      "epoch": 3.4911095659084137,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005207152081056919,
+      "loss": 0.567,
+      "step": 70290
+    },
+    {
+      "epoch": 3.491606238204033,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005206754743220423,
+      "loss": 0.5386,
+      "step": 70300
+    },
+    {
+      "epoch": 3.492102910499652,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005206357405383928,
+      "loss": 0.5078,
+      "step": 70310
+    },
+    {
+      "epoch": 3.492599582795272,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005205960067547432,
+      "loss": 0.5088,
+      "step": 70320
+    },
+    {
+      "epoch": 3.493096255090891,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005205562729710937,
+      "loss": 0.5491,
+      "step": 70330
+    },
+    {
+      "epoch": 3.4935929273865103,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005205165391874442,
+      "loss": 0.5201,
+      "step": 70340
+    },
+    {
+      "epoch": 3.4940895996821295,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005204768054037946,
+      "loss": 0.5218,
+      "step": 70350
+    },
+    {
+      "epoch": 3.494586271977749,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.000520437071620145,
+      "loss": 0.5051,
+      "step": 70360
+    },
+    {
+      "epoch": 3.4950829442733684,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005203973378364956,
+      "loss": 0.5346,
+      "step": 70370
+    },
+    {
+      "epoch": 3.4955796165689876,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.000520357604052846,
+      "loss": 0.5233,
+      "step": 70380
+    },
+    {
+      "epoch": 3.4960762888646073,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005203178702691964,
+      "loss": 0.526,
+      "step": 70390
+    },
+    {
+      "epoch": 3.4965729611602265,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005202781364855468,
+      "loss": 0.5111,
+      "step": 70400
+    },
+    {
+      "epoch": 3.4970696334558458,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005202384027018973,
+      "loss": 0.5377,
+      "step": 70410
+    },
+    {
+      "epoch": 3.497566305751465,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0005201986689182478,
+      "loss": 0.5488,
+      "step": 70420
+    },
+    {
+      "epoch": 3.4980629780470847,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005201589351345982,
+      "loss": 0.5356,
+      "step": 70430
+    },
+    {
+      "epoch": 3.498559650342704,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005201192013509487,
+      "loss": 0.5158,
+      "step": 70440
+    },
+    {
+      "epoch": 3.499056322638323,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005200794675672991,
+      "loss": 0.545,
+      "step": 70450
+    },
+    {
+      "epoch": 3.499552994933943,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005200397337836495,
+      "loss": 0.5401,
+      "step": 70460
+    },
+    {
+      "epoch": 3.500049667229562,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005200000000000001,
+      "loss": 0.519,
+      "step": 70470
+    },
+    {
+      "epoch": 3.5005463395251812,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005199602662163505,
+      "loss": 0.5401,
+      "step": 70480
+    },
+    {
+      "epoch": 3.5010430118208005,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005199205324327009,
+      "loss": 0.5255,
+      "step": 70490
+    },
+    {
+      "epoch": 3.50153968411642,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005198807986490514,
+      "loss": 0.5162,
+      "step": 70500
+    },
+    {
+      "epoch": 3.5020363564120394,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005198410648654018,
+      "loss": 0.5138,
+      "step": 70510
+    },
+    {
+      "epoch": 3.5025330287076586,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005198013310817523,
+      "loss": 0.5377,
+      "step": 70520
+    },
+    {
+      "epoch": 3.5030297010032783,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005197615972981028,
+      "loss": 0.534,
+      "step": 70530
+    },
+    {
+      "epoch": 3.5035263732988975,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005197218635144532,
+      "loss": 0.557,
+      "step": 70540
+    },
+    {
+      "epoch": 3.5040230455945167,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005196821297308036,
+      "loss": 0.5508,
+      "step": 70550
+    },
+    {
+      "epoch": 3.504519717890136,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005196423959471541,
+      "loss": 0.5263,
+      "step": 70560
+    },
+    {
+      "epoch": 3.505016390185755,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005196026621635046,
+      "loss": 0.5132,
+      "step": 70570
+    },
+    {
+      "epoch": 3.505513062481375,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.000519562928379855,
+      "loss": 0.5644,
+      "step": 70580
+    },
+    {
+      "epoch": 3.506009734776994,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005195231945962054,
+      "loss": 0.501,
+      "step": 70590
+    },
+    {
+      "epoch": 3.5065064070726137,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005194834608125559,
+      "loss": 0.5329,
+      "step": 70600
+    },
+    {
+      "epoch": 3.507003079368233,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005194437270289063,
+      "loss": 0.5443,
+      "step": 70610
+    },
+    {
+      "epoch": 3.507499751663852,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005194039932452568,
+      "loss": 0.5276,
+      "step": 70620
+    },
+    {
+      "epoch": 3.5079964239594714,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005193642594616073,
+      "loss": 0.5405,
+      "step": 70630
+    },
+    {
+      "epoch": 3.5084930962550906,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005193245256779578,
+      "loss": 0.5529,
+      "step": 70640
+    },
+    {
+      "epoch": 3.5089897685507103,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005192847918943081,
+      "loss": 0.5094,
+      "step": 70650
+    },
+    {
+      "epoch": 3.5094864408463295,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005192450581106586,
+      "loss": 0.5102,
+      "step": 70660
+    },
+    {
+      "epoch": 3.509983113141949,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005192053243270091,
+      "loss": 0.5402,
+      "step": 70670
+    },
+    {
+      "epoch": 3.5104797854375684,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005191655905433595,
+      "loss": 0.5365,
+      "step": 70680
+    },
+    {
+      "epoch": 3.5109764577331877,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00051912585675971,
+      "loss": 0.554,
+      "step": 70690
+    },
+    {
+      "epoch": 3.511473130028807,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005190861229760604,
+      "loss": 0.5333,
+      "step": 70700
+    },
+    {
+      "epoch": 3.511969802324426,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005190463891924108,
+      "loss": 0.5351,
+      "step": 70710
+    },
+    {
+      "epoch": 3.512466474620046,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005190066554087614,
+      "loss": 0.5524,
+      "step": 70720
+    },
+    {
+      "epoch": 3.512963146915665,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005189669216251118,
+      "loss": 0.5221,
+      "step": 70730
+    },
+    {
+      "epoch": 3.5134598192112843,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005189271878414622,
+      "loss": 0.5439,
+      "step": 70740
+    },
+    {
+      "epoch": 3.513956491506904,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005188874540578127,
+      "loss": 0.534,
+      "step": 70750
+    },
+    {
+      "epoch": 3.514453163802523,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005188477202741631,
+      "loss": 0.5449,
+      "step": 70760
+    },
+    {
+      "epoch": 3.5149498360981424,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005188079864905136,
+      "loss": 0.5213,
+      "step": 70770
+    },
+    {
+      "epoch": 3.5154465083937616,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.000518768252706864,
+      "loss": 0.5238,
+      "step": 70780
+    },
+    {
+      "epoch": 3.5159431806893813,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005187285189232145,
+      "loss": 0.5503,
+      "step": 70790
+    },
+    {
+      "epoch": 3.5164398529850005,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.000518688785139565,
+      "loss": 0.5157,
+      "step": 70800
+    },
+    {
+      "epoch": 3.5169365252806197,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005186490513559153,
+      "loss": 0.5357,
+      "step": 70810
+    },
+    {
+      "epoch": 3.5174331975762394,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005186093175722659,
+      "loss": 0.5286,
+      "step": 70820
+    },
+    {
+      "epoch": 3.5179298698718586,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005185695837886164,
+      "loss": 0.556,
+      "step": 70830
+    },
+    {
+      "epoch": 3.518426542167478,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005185298500049667,
+      "loss": 0.5277,
+      "step": 70840
+    },
+    {
+      "epoch": 3.518923214463097,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005184901162213172,
+      "loss": 0.5423,
+      "step": 70850
+    },
+    {
+      "epoch": 3.5194198867587168,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005184503824376676,
+      "loss": 0.5051,
+      "step": 70860
+    },
+    {
+      "epoch": 3.519916559054336,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.000518410648654018,
+      "loss": 0.5264,
+      "step": 70870
+    },
+    {
+      "epoch": 3.520413231349955,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005183709148703686,
+      "loss": 0.5475,
+      "step": 70880
+    },
+    {
+      "epoch": 3.520909903645575,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.000518331181086719,
+      "loss": 0.5207,
+      "step": 70890
+    },
+    {
+      "epoch": 3.521406575941194,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0005182914473030694,
+      "loss": 0.532,
+      "step": 70900
+    },
+    {
+      "epoch": 3.5219032482368133,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005182517135194199,
+      "loss": 0.5313,
+      "step": 70910
+    },
+    {
+      "epoch": 3.5223999205324326,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005182119797357704,
+      "loss": 0.5218,
+      "step": 70920
+    },
+    {
+      "epoch": 3.522896592828052,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0005181722459521208,
+      "loss": 0.5331,
+      "step": 70930
+    },
+    {
+      "epoch": 3.5233932651236715,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005181325121684713,
+      "loss": 0.5286,
+      "step": 70940
+    },
+    {
+      "epoch": 3.5238899374192907,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005180927783848217,
+      "loss": 0.534,
+      "step": 70950
+    },
+    {
+      "epoch": 3.5243866097149104,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005180530446011722,
+      "loss": 0.5566,
+      "step": 70960
+    },
+    {
+      "epoch": 3.5248832820105296,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005180133108175227,
+      "loss": 0.5477,
+      "step": 70970
+    },
+    {
+      "epoch": 3.525379954306149,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005179735770338731,
+      "loss": 0.512,
+      "step": 70980
+    },
+    {
+      "epoch": 3.525876626601768,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005179338432502236,
+      "loss": 0.5373,
+      "step": 70990
+    },
+    {
+      "epoch": 3.5263732988973873,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005178941094665739,
+      "loss": 0.5523,
+      "step": 71000
+    },
+    {
+      "epoch": 3.526869971193007,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005178543756829244,
+      "loss": 0.5356,
+      "step": 71010
+    },
+    {
+      "epoch": 3.527366643488626,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.000517814641899275,
+      "loss": 0.4895,
+      "step": 71020
+    },
+    {
+      "epoch": 3.5278633157842454,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005177749081156253,
+      "loss": 0.5625,
+      "step": 71030
+    },
+    {
+      "epoch": 3.528359988079865,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005177351743319758,
+      "loss": 0.5389,
+      "step": 71040
+    },
+    {
+      "epoch": 3.5288566603754843,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005176954405483262,
+      "loss": 0.5461,
+      "step": 71050
+    },
+    {
+      "epoch": 3.5293533326711035,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005176557067646766,
+      "loss": 0.5489,
+      "step": 71060
+    },
+    {
+      "epoch": 3.5298500049667227,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005176159729810272,
+      "loss": 0.5162,
+      "step": 71070
+    },
+    {
+      "epoch": 3.5303466772623424,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005175762391973776,
+      "loss": 0.5644,
+      "step": 71080
+    },
+    {
+      "epoch": 3.5308433495579616,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005175365054137281,
+      "loss": 0.5453,
+      "step": 71090
+    },
+    {
+      "epoch": 3.531340021853581,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005174967716300785,
+      "loss": 0.5447,
+      "step": 71100
+    },
+    {
+      "epoch": 3.5318366941492005,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005174570378464289,
+      "loss": 0.5287,
+      "step": 71110
+    },
+    {
+      "epoch": 3.5323333664448198,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0005174173040627795,
+      "loss": 0.5363,
+      "step": 71120
+    },
+    {
+      "epoch": 3.532830038740439,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005173775702791299,
+      "loss": 0.5309,
+      "step": 71130
+    },
+    {
+      "epoch": 3.533326711036058,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005173378364954803,
+      "loss": 0.5407,
+      "step": 71140
+    },
+    {
+      "epoch": 3.533823383331678,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0005172981027118308,
+      "loss": 0.5334,
+      "step": 71150
+    },
+    {
+      "epoch": 3.534320055627297,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005172583689281812,
+      "loss": 0.5472,
+      "step": 71160
+    },
+    {
+      "epoch": 3.5348167279229163,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005172186351445316,
+      "loss": 0.5391,
+      "step": 71170
+    },
+    {
+      "epoch": 3.535313400218536,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005171789013608822,
+      "loss": 0.5256,
+      "step": 71180
+    },
+    {
+      "epoch": 3.5358100725141552,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005171391675772325,
+      "loss": 0.5316,
+      "step": 71190
+    },
+    {
+      "epoch": 3.5363067448097745,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.000517099433793583,
+      "loss": 0.5215,
+      "step": 71200
+    },
+    {
+      "epoch": 3.5368034171053937,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005170597000099335,
+      "loss": 0.5602,
+      "step": 71210
+    },
+    {
+      "epoch": 3.5373000894010134,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005170199662262838,
+      "loss": 0.5639,
+      "step": 71220
+    },
+    {
+      "epoch": 3.5377967616966326,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005169802324426344,
+      "loss": 0.5334,
+      "step": 71230
+    },
+    {
+      "epoch": 3.538293433992252,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005169404986589849,
+      "loss": 0.5257,
+      "step": 71240
+    },
+    {
+      "epoch": 3.5387901062878715,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005169007648753353,
+      "loss": 0.5557,
+      "step": 71250
+    },
+    {
+      "epoch": 3.5392867785834907,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005168610310916857,
+      "loss": 0.526,
+      "step": 71260
+    },
+    {
+      "epoch": 3.53978345087911,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005168212973080361,
+      "loss": 0.5291,
+      "step": 71270
+    },
+    {
+      "epoch": 3.540280123174729,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005167815635243867,
+      "loss": 0.5306,
+      "step": 71280
+    },
+    {
+      "epoch": 3.5407767954703484,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0005167418297407371,
+      "loss": 0.5884,
+      "step": 71290
+    },
+    {
+      "epoch": 3.541273467765968,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005167020959570875,
+      "loss": 0.54,
+      "step": 71300
+    },
+    {
+      "epoch": 3.5417701400615873,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.000516662362173438,
+      "loss": 0.5544,
+      "step": 71310
+    },
+    {
+      "epoch": 3.542266812357207,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005166226283897884,
+      "loss": 0.5319,
+      "step": 71320
+    },
+    {
+      "epoch": 3.542763484652826,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005165828946061389,
+      "loss": 0.5144,
+      "step": 71330
+    },
+    {
+      "epoch": 3.5432601569484454,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005165431608224894,
+      "loss": 0.5351,
+      "step": 71340
+    },
+    {
+      "epoch": 3.5437568292440647,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005165034270388398,
+      "loss": 0.528,
+      "step": 71350
+    },
+    {
+      "epoch": 3.544253501539684,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005164636932551902,
+      "loss": 0.5571,
+      "step": 71360
+    },
+    {
+      "epoch": 3.5447501738353036,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005164239594715408,
+      "loss": 0.5561,
+      "step": 71370
+    },
+    {
+      "epoch": 3.545246846130923,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005163842256878911,
+      "loss": 0.5146,
+      "step": 71380
+    },
+    {
+      "epoch": 3.545743518426542,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005163444919042416,
+      "loss": 0.5854,
+      "step": 71390
+    },
+    {
+      "epoch": 3.5462401907221617,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005163047581205921,
+      "loss": 0.5365,
+      "step": 71400
+    },
+    {
+      "epoch": 3.546736863017781,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005162650243369425,
+      "loss": 0.5505,
+      "step": 71410
+    },
+    {
+      "epoch": 3.5472335353134,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005162252905532929,
+      "loss": 0.5291,
+      "step": 71420
+    },
+    {
+      "epoch": 3.5477302076090194,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005161855567696435,
+      "loss": 0.5225,
+      "step": 71430
+    },
+    {
+      "epoch": 3.548226879904639,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005161458229859939,
+      "loss": 0.5633,
+      "step": 71440
+    },
+    {
+      "epoch": 3.5487235522002583,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005161060892023443,
+      "loss": 0.5341,
+      "step": 71450
+    },
+    {
+      "epoch": 3.5492202244958775,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005160663554186947,
+      "loss": 0.5525,
+      "step": 71460
+    },
+    {
+      "epoch": 3.549716896791497,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005160266216350452,
+      "loss": 0.5279,
+      "step": 71470
+    },
+    {
+      "epoch": 3.5502135690871164,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005159868878513957,
+      "loss": 0.5233,
+      "step": 71480
+    },
+    {
+      "epoch": 3.5507102413827356,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005159471540677461,
+      "loss": 0.5602,
+      "step": 71490
+    },
+    {
+      "epoch": 3.551206913678355,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0005159074202840966,
+      "loss": 0.5283,
+      "step": 71500
+    },
+    {
+      "epoch": 3.5517035859739745,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.000515867686500447,
+      "loss": 0.5276,
+      "step": 71510
+    },
+    {
+      "epoch": 3.5522002582695937,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005158279527167974,
+      "loss": 0.5337,
+      "step": 71520
+    },
+    {
+      "epoch": 3.552696930565213,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.000515788218933148,
+      "loss": 0.5452,
+      "step": 71530
+    },
+    {
+      "epoch": 3.5531936028608326,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005157484851494984,
+      "loss": 0.5402,
+      "step": 71540
+    },
+    {
+      "epoch": 3.553690275156452,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005157087513658488,
+      "loss": 0.5368,
+      "step": 71550
+    },
+    {
+      "epoch": 3.554186947452071,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005156690175821993,
+      "loss": 0.5163,
+      "step": 71560
+    },
+    {
+      "epoch": 3.5546836197476903,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005156292837985497,
+      "loss": 0.5423,
+      "step": 71570
+    },
+    {
+      "epoch": 3.55518029204331,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005155895500149002,
+      "loss": 0.5379,
+      "step": 71580
+    },
+    {
+      "epoch": 3.555676964338929,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005155498162312507,
+      "loss": 0.5347,
+      "step": 71590
+    },
+    {
+      "epoch": 3.5561736366345484,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005155100824476011,
+      "loss": 0.5431,
+      "step": 71600
+    },
+    {
+      "epoch": 3.556670308930168,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005154703486639515,
+      "loss": 0.5104,
+      "step": 71610
+    },
+    {
+      "epoch": 3.5571669812257873,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.000515430614880302,
+      "loss": 0.5474,
+      "step": 71620
+    },
+    {
+      "epoch": 3.5576636535214066,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005153908810966525,
+      "loss": 0.552,
+      "step": 71630
+    },
+    {
+      "epoch": 3.558160325817026,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005153511473130029,
+      "loss": 0.534,
+      "step": 71640
+    },
+    {
+      "epoch": 3.558656998112645,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005153114135293533,
+      "loss": 0.5133,
+      "step": 71650
+    },
+    {
+      "epoch": 3.5591536704082647,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005152716797457038,
+      "loss": 0.5526,
+      "step": 71660
+    },
+    {
+      "epoch": 3.559650342703884,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005152319459620542,
+      "loss": 0.5265,
+      "step": 71670
+    },
+    {
+      "epoch": 3.5601470149995036,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005151922121784047,
+      "loss": 0.5408,
+      "step": 71680
+    },
+    {
+      "epoch": 3.560643687295123,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005151524783947552,
+      "loss": 0.5153,
+      "step": 71690
+    },
+    {
+      "epoch": 3.561140359590742,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005151127446111057,
+      "loss": 0.5274,
+      "step": 71700
+    },
+    {
+      "epoch": 3.5616370318863613,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.000515073010827456,
+      "loss": 0.5479,
+      "step": 71710
+    },
+    {
+      "epoch": 3.5621337041819805,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005150332770438065,
+      "loss": 0.5359,
+      "step": 71720
+    },
+    {
+      "epoch": 3.5626303764776,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.000514993543260157,
+      "loss": 0.5266,
+      "step": 71730
+    },
+    {
+      "epoch": 3.5631270487732194,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005149538094765074,
+      "loss": 0.5389,
+      "step": 71740
+    },
+    {
+      "epoch": 3.5636237210688386,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005149140756928579,
+      "loss": 0.5269,
+      "step": 71750
+    },
+    {
+      "epoch": 3.5641203933644583,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005148743419092083,
+      "loss": 0.5298,
+      "step": 71760
+    },
+    {
+      "epoch": 3.5646170656600775,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005148346081255587,
+      "loss": 0.5324,
+      "step": 71770
+    },
+    {
+      "epoch": 3.5651137379556967,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0005147948743419093,
+      "loss": 0.5094,
+      "step": 71780
+    },
+    {
+      "epoch": 3.565610410251316,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005147551405582597,
+      "loss": 0.541,
+      "step": 71790
+    },
+    {
+      "epoch": 3.5661070825469356,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0005147154067746101,
+      "loss": 0.5141,
+      "step": 71800
+    },
+    {
+      "epoch": 3.566603754842555,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005146756729909606,
+      "loss": 0.5217,
+      "step": 71810
+    },
+    {
+      "epoch": 3.567100427138174,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.000514635939207311,
+      "loss": 0.5366,
+      "step": 71820
+    },
+    {
+      "epoch": 3.5675970994337938,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005145962054236616,
+      "loss": 0.542,
+      "step": 71830
+    },
+    {
+      "epoch": 3.568093771729413,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005145564716400119,
+      "loss": 0.5197,
+      "step": 71840
+    },
+    {
+      "epoch": 3.5685904440250322,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0005145167378563624,
+      "loss": 0.5421,
+      "step": 71850
+    },
+    {
+      "epoch": 3.5690871163206515,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005144770040727129,
+      "loss": 0.5638,
+      "step": 71860
+    },
+    {
+      "epoch": 3.569583788616271,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005144372702890632,
+      "loss": 0.501,
+      "step": 71870
+    },
+    {
+      "epoch": 3.5700804609118904,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005143975365054138,
+      "loss": 0.5251,
+      "step": 71880
+    },
+    {
+      "epoch": 3.5705771332075096,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005143578027217643,
+      "loss": 0.5545,
+      "step": 71890
+    },
+    {
+      "epoch": 3.5710738055031293,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005143180689381146,
+      "loss": 0.5231,
+      "step": 71900
+    },
+    {
+      "epoch": 3.5715704777987485,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005142783351544651,
+      "loss": 0.5064,
+      "step": 71910
+    },
+    {
+      "epoch": 3.5720671500943677,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005142386013708155,
+      "loss": 0.5371,
+      "step": 71920
+    },
+    {
+      "epoch": 3.572563822389987,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.000514198867587166,
+      "loss": 0.5559,
+      "step": 71930
+    },
+    {
+      "epoch": 3.5730604946856066,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005141591338035165,
+      "loss": 0.5532,
+      "step": 71940
+    },
+    {
+      "epoch": 3.573557166981226,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005141194000198669,
+      "loss": 0.5337,
+      "step": 71950
+    },
+    {
+      "epoch": 3.574053839276845,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005140796662362173,
+      "loss": 0.5438,
+      "step": 71960
+    },
+    {
+      "epoch": 3.5745505115724647,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005140399324525678,
+      "loss": 0.507,
+      "step": 71970
+    },
+    {
+      "epoch": 3.575047183868084,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005140001986689183,
+      "loss": 0.5545,
+      "step": 71980
+    },
+    {
+      "epoch": 3.575543856163703,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005139604648852688,
+      "loss": 0.5334,
+      "step": 71990
+    },
+    {
+      "epoch": 3.5760405284593224,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005139207311016192,
+      "loss": 0.5445,
+      "step": 72000
+    },
+    {
+      "epoch": 3.5765372007549416,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005138809973179696,
+      "loss": 0.5251,
+      "step": 72010
+    },
+    {
+      "epoch": 3.5770338730505613,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0005138412635343201,
+      "loss": 0.5447,
+      "step": 72020
+    },
+    {
+      "epoch": 3.5775305453461805,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005138015297506706,
+      "loss": 0.5429,
+      "step": 72030
+    },
+    {
+      "epoch": 3.5780272176418,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.000513761795967021,
+      "loss": 0.5255,
+      "step": 72040
+    },
+    {
+      "epoch": 3.5785238899374194,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005137220621833715,
+      "loss": 0.5466,
+      "step": 72050
+    },
+    {
+      "epoch": 3.5790205622330387,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005136823283997218,
+      "loss": 0.5773,
+      "step": 72060
+    },
+    {
+      "epoch": 3.579517234528658,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005136425946160723,
+      "loss": 0.5379,
+      "step": 72070
+    },
+    {
+      "epoch": 3.580013906824277,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005136028608324229,
+      "loss": 0.5583,
+      "step": 72080
+    },
+    {
+      "epoch": 3.580510579119897,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005135631270487732,
+      "loss": 0.5559,
+      "step": 72090
+    },
+    {
+      "epoch": 3.581007251415516,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005135233932651237,
+      "loss": 0.5188,
+      "step": 72100
+    },
+    {
+      "epoch": 3.5815039237111352,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005134836594814741,
+      "loss": 0.5299,
+      "step": 72110
+    },
+    {
+      "epoch": 3.582000596006755,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005134439256978245,
+      "loss": 0.522,
+      "step": 72120
+    },
+    {
+      "epoch": 3.582497268302374,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005134041919141751,
+      "loss": 0.5496,
+      "step": 72130
+    },
+    {
+      "epoch": 3.5829939405979934,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005133644581305255,
+      "loss": 0.5553,
+      "step": 72140
+    },
+    {
+      "epoch": 3.5834906128936126,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.000513324724346876,
+      "loss": 0.5767,
+      "step": 72150
+    },
+    {
+      "epoch": 3.5839872851892323,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005132849905632264,
+      "loss": 0.5259,
+      "step": 72160
+    },
+    {
+      "epoch": 3.5844839574848515,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005132452567795768,
+      "loss": 0.546,
+      "step": 72170
+    },
+    {
+      "epoch": 3.5849806297804707,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005132055229959274,
+      "loss": 0.5524,
+      "step": 72180
+    },
+    {
+      "epoch": 3.5854773020760904,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0005131657892122778,
+      "loss": 0.5292,
+      "step": 72190
+    },
+    {
+      "epoch": 3.5859739743717096,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005131260554286282,
+      "loss": 0.5343,
+      "step": 72200
+    },
+    {
+      "epoch": 3.586470646667329,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005130863216449787,
+      "loss": 0.5208,
+      "step": 72210
+    },
+    {
+      "epoch": 3.586967318962948,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005130465878613291,
+      "loss": 0.5311,
+      "step": 72220
+    },
+    {
+      "epoch": 3.5874639912585677,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0005130068540776796,
+      "loss": 0.5472,
+      "step": 72230
+    },
+    {
+      "epoch": 3.587960663554187,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005129671202940301,
+      "loss": 0.5617,
+      "step": 72240
+    },
+    {
+      "epoch": 3.588457335849806,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005129273865103804,
+      "loss": 0.5349,
+      "step": 72250
+    },
+    {
+      "epoch": 3.588954008145426,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005128876527267309,
+      "loss": 0.5385,
+      "step": 72260
+    },
+    {
+      "epoch": 3.589450680441045,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005128479189430814,
+      "loss": 0.5568,
+      "step": 72270
+    },
+    {
+      "epoch": 3.5899473527366643,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005128081851594319,
+      "loss": 0.5277,
+      "step": 72280
+    },
+    {
+      "epoch": 3.5904440250322835,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005127684513757823,
+      "loss": 0.5236,
+      "step": 72290
+    },
+    {
+      "epoch": 3.5909406973279028,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005127287175921328,
+      "loss": 0.5263,
+      "step": 72300
+    },
+    {
+      "epoch": 3.5914373696235224,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005126889838084832,
+      "loss": 0.5068,
+      "step": 72310
+    },
+    {
+      "epoch": 3.5919340419191417,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005126492500248336,
+      "loss": 0.5226,
+      "step": 72320
+    },
+    {
+      "epoch": 3.5924307142147613,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.000512609516241184,
+      "loss": 0.526,
+      "step": 72330
+    },
+    {
+      "epoch": 3.5929273865103806,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005125697824575346,
+      "loss": 0.5258,
+      "step": 72340
+    },
+    {
+      "epoch": 3.593424058806,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.000512530048673885,
+      "loss": 0.5363,
+      "step": 72350
+    },
+    {
+      "epoch": 3.593920731101619,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005124903148902354,
+      "loss": 0.5559,
+      "step": 72360
+    },
+    {
+      "epoch": 3.5944174033972383,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005124505811065859,
+      "loss": 0.5435,
+      "step": 72370
+    },
+    {
+      "epoch": 3.594914075692858,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005124108473229364,
+      "loss": 0.5365,
+      "step": 72380
+    },
+    {
+      "epoch": 3.595410747988477,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005123711135392868,
+      "loss": 0.5493,
+      "step": 72390
+    },
+    {
+      "epoch": 3.595907420284097,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005123313797556373,
+      "loss": 0.4943,
+      "step": 72400
+    },
+    {
+      "epoch": 3.596404092579716,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005122916459719877,
+      "loss": 0.5471,
+      "step": 72410
+    },
+    {
+      "epoch": 3.5969007648753353,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005122519121883381,
+      "loss": 0.5154,
+      "step": 72420
+    },
+    {
+      "epoch": 3.5973974371709545,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005122121784046887,
+      "loss": 0.5237,
+      "step": 72430
+    },
+    {
+      "epoch": 3.5978941094665737,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005121724446210391,
+      "loss": 0.5129,
+      "step": 72440
+    },
+    {
+      "epoch": 3.5983907817621934,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005121327108373895,
+      "loss": 0.5398,
+      "step": 72450
+    },
+    {
+      "epoch": 3.5988874540578126,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.00051209297705374,
+      "loss": 0.5368,
+      "step": 72460
+    },
+    {
+      "epoch": 3.599384126353432,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005120532432700904,
+      "loss": 0.5295,
+      "step": 72470
+    },
+    {
+      "epoch": 3.5998807986490515,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005120135094864408,
+      "loss": 0.5352,
+      "step": 72480
+    },
+    {
+      "epoch": 3.6003774709446708,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005119737757027914,
+      "loss": 0.5417,
+      "step": 72490
+    },
+    {
+      "epoch": 3.60087414324029,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005119340419191418,
+      "loss": 0.5551,
+      "step": 72500
+    },
+    {
+      "epoch": 3.601370815535909,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005118943081354922,
+      "loss": 0.5224,
+      "step": 72510
+    },
+    {
+      "epoch": 3.601867487831529,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005118545743518426,
+      "loss": 0.5196,
+      "step": 72520
+    },
+    {
+      "epoch": 3.602364160127148,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005118148405681932,
+      "loss": 0.5537,
+      "step": 72530
+    },
+    {
+      "epoch": 3.6028608324227673,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005117751067845436,
+      "loss": 0.5389,
+      "step": 72540
+    },
+    {
+      "epoch": 3.603357504718387,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.000511735373000894,
+      "loss": 0.5216,
+      "step": 72550
+    },
+    {
+      "epoch": 3.6038541770140062,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005116956392172445,
+      "loss": 0.5322,
+      "step": 72560
+    },
+    {
+      "epoch": 3.6043508493096255,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005116559054335949,
+      "loss": 0.5167,
+      "step": 72570
+    },
+    {
+      "epoch": 3.6048475216052447,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005116161716499453,
+      "loss": 0.5366,
+      "step": 72580
+    },
+    {
+      "epoch": 3.6053441939008644,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005115764378662959,
+      "loss": 0.5253,
+      "step": 72590
+    },
+    {
+      "epoch": 3.6058408661964836,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005115367040826463,
+      "loss": 0.5323,
+      "step": 72600
+    },
+    {
+      "epoch": 3.606337538492103,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005114969702989967,
+      "loss": 0.5361,
+      "step": 72610
+    },
+    {
+      "epoch": 3.6068342107877225,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005114572365153472,
+      "loss": 0.5282,
+      "step": 72620
+    },
+    {
+      "epoch": 3.6073308830833417,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005114175027316976,
+      "loss": 0.5132,
+      "step": 72630
+    },
+    {
+      "epoch": 3.607827555378961,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005113777689480481,
+      "loss": 0.5314,
+      "step": 72640
+    },
+    {
+      "epoch": 3.60832422767458,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005113380351643986,
+      "loss": 0.5729,
+      "step": 72650
+    },
+    {
+      "epoch": 3.6088208999701994,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.000511298301380749,
+      "loss": 0.5214,
+      "step": 72660
+    },
+    {
+      "epoch": 3.609317572265819,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005112585675970994,
+      "loss": 0.5506,
+      "step": 72670
+    },
+    {
+      "epoch": 3.6098142445614383,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.00051121883381345,
+      "loss": 0.5271,
+      "step": 72680
+    },
+    {
+      "epoch": 3.610310916857058,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005111791000298004,
+      "loss": 0.5279,
+      "step": 72690
+    },
+    {
+      "epoch": 3.610807589152677,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005111393662461508,
+      "loss": 0.5279,
+      "step": 72700
+    },
+    {
+      "epoch": 3.6113042614482964,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005110996324625012,
+      "loss": 0.5447,
+      "step": 72710
+    },
+    {
+      "epoch": 3.6118009337439156,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0005110598986788517,
+      "loss": 0.5431,
+      "step": 72720
+    },
+    {
+      "epoch": 3.612297606039535,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0005110201648952023,
+      "loss": 0.5097,
+      "step": 72730
+    },
+    {
+      "epoch": 3.6127942783351545,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005109804311115526,
+      "loss": 0.5035,
+      "step": 72740
+    },
+    {
+      "epoch": 3.6132909506307738,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0005109406973279031,
+      "loss": 0.5387,
+      "step": 72750
+    },
+    {
+      "epoch": 3.6137876229263934,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005109009635442536,
+      "loss": 0.5378,
+      "step": 72760
+    },
+    {
+      "epoch": 3.6142842952220127,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0005108612297606039,
+      "loss": 0.5338,
+      "step": 72770
+    },
+    {
+      "epoch": 3.614780967517632,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005108214959769544,
+      "loss": 0.5323,
+      "step": 72780
+    },
+    {
+      "epoch": 3.615277639813251,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005107817621933049,
+      "loss": 0.54,
+      "step": 72790
+    },
+    {
+      "epoch": 3.6157743121088703,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005107420284096553,
+      "loss": 0.5103,
+      "step": 72800
+    },
+    {
+      "epoch": 3.61627098440449,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005107022946260058,
+      "loss": 0.5156,
+      "step": 72810
+    },
+    {
+      "epoch": 3.6167676567001092,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005106625608423562,
+      "loss": 0.5502,
+      "step": 72820
+    },
+    {
+      "epoch": 3.6172643289957285,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005106228270587066,
+      "loss": 0.5266,
+      "step": 72830
+    },
+    {
+      "epoch": 3.617761001291348,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005105830932750572,
+      "loss": 0.564,
+      "step": 72840
+    },
+    {
+      "epoch": 3.6182576735869674,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005105433594914076,
+      "loss": 0.55,
+      "step": 72850
+    },
+    {
+      "epoch": 3.6187543458825866,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.000510503625707758,
+      "loss": 0.5557,
+      "step": 72860
+    },
+    {
+      "epoch": 3.619251018178206,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005104638919241085,
+      "loss": 0.5395,
+      "step": 72870
+    },
+    {
+      "epoch": 3.6197476904738255,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005104241581404589,
+      "loss": 0.5113,
+      "step": 72880
+    },
+    {
+      "epoch": 3.6202443627694447,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005103844243568095,
+      "loss": 0.5233,
+      "step": 72890
+    },
+    {
+      "epoch": 3.620741035065064,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005103446905731599,
+      "loss": 0.5474,
+      "step": 72900
+    },
+    {
+      "epoch": 3.6212377073606836,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005103049567895103,
+      "loss": 0.5339,
+      "step": 72910
+    },
+    {
+      "epoch": 3.621734379656303,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0005102652230058608,
+      "loss": 0.5733,
+      "step": 72920
+    },
+    {
+      "epoch": 3.622231051951922,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005102254892222111,
+      "loss": 0.5275,
+      "step": 72930
+    },
+    {
+      "epoch": 3.6227277242475413,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005101857554385617,
+      "loss": 0.5442,
+      "step": 72940
+    },
+    {
+      "epoch": 3.623224396543161,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005101460216549122,
+      "loss": 0.5314,
+      "step": 72950
+    },
+    {
+      "epoch": 3.62372106883878,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005101062878712625,
+      "loss": 0.5269,
+      "step": 72960
+    },
+    {
+      "epoch": 3.6242177411343994,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.000510066554087613,
+      "loss": 0.5378,
+      "step": 72970
+    },
+    {
+      "epoch": 3.624714413430019,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005100268203039634,
+      "loss": 0.5608,
+      "step": 72980
+    },
+    {
+      "epoch": 3.6252110857256383,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005099870865203139,
+      "loss": 0.5212,
+      "step": 72990
+    },
+    {
+      "epoch": 3.6257077580212576,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005099473527366644,
+      "loss": 0.5278,
+      "step": 73000
+    },
+    {
+      "epoch": 3.626204430316877,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005099076189530148,
+      "loss": 0.5407,
+      "step": 73010
+    },
+    {
+      "epoch": 3.626701102612496,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005098678851693652,
+      "loss": 0.5108,
+      "step": 73020
+    },
+    {
+      "epoch": 3.6271977749081157,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005098281513857157,
+      "loss": 0.553,
+      "step": 73030
+    },
+    {
+      "epoch": 3.627694447203735,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005097884176020662,
+      "loss": 0.5371,
+      "step": 73040
+    },
+    {
+      "epoch": 3.6281911194993546,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005097486838184167,
+      "loss": 0.5203,
+      "step": 73050
+    },
+    {
+      "epoch": 3.628687791794974,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0005097089500347671,
+      "loss": 0.5108,
+      "step": 73060
+    },
+    {
+      "epoch": 3.629184464090593,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0005096692162511175,
+      "loss": 0.5054,
+      "step": 73070
+    },
+    {
+      "epoch": 3.6296811363862123,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.000509629482467468,
+      "loss": 0.5246,
+      "step": 73080
+    },
+    {
+      "epoch": 3.6301778086818315,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005095897486838185,
+      "loss": 0.5111,
+      "step": 73090
+    },
+    {
+      "epoch": 3.630674480977451,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005095500149001689,
+      "loss": 0.5068,
+      "step": 73100
+    },
+    {
+      "epoch": 3.6311711532730704,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005095102811165194,
+      "loss": 0.5182,
+      "step": 73110
+    },
+    {
+      "epoch": 3.63166782556869,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005094705473328697,
+      "loss": 0.5182,
+      "step": 73120
+    },
+    {
+      "epoch": 3.6321644978643093,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005094308135492202,
+      "loss": 0.5534,
+      "step": 73130
+    },
+    {
+      "epoch": 3.6326611701599285,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005093910797655708,
+      "loss": 0.553,
+      "step": 73140
+    },
+    {
+      "epoch": 3.6331578424555477,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005093513459819211,
+      "loss": 0.5526,
+      "step": 73150
+    },
+    {
+      "epoch": 3.633654514751167,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005093116121982716,
+      "loss": 0.5402,
+      "step": 73160
+    },
+    {
+      "epoch": 3.6341511870467866,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005092718784146221,
+      "loss": 0.515,
+      "step": 73170
+    },
+    {
+      "epoch": 3.634647859342406,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005092321446309725,
+      "loss": 0.5568,
+      "step": 73180
+    },
+    {
+      "epoch": 3.635144531638025,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.000509192410847323,
+      "loss": 0.5495,
+      "step": 73190
+    },
+    {
+      "epoch": 3.6356412039336448,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005091526770636734,
+      "loss": 0.5464,
+      "step": 73200
+    },
+    {
+      "epoch": 3.636137876229264,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005091129432800239,
+      "loss": 0.5332,
+      "step": 73210
+    },
+    {
+      "epoch": 3.636634548524883,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0005090732094963743,
+      "loss": 0.5506,
+      "step": 73220
+    },
+    {
+      "epoch": 3.6371312208205024,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0005090334757127247,
+      "loss": 0.5535,
+      "step": 73230
+    },
+    {
+      "epoch": 3.637627893116122,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005089937419290753,
+      "loss": 0.5442,
+      "step": 73240
+    },
+    {
+      "epoch": 3.6381245654117413,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005089540081454257,
+      "loss": 0.5157,
+      "step": 73250
+    },
+    {
+      "epoch": 3.6386212377073606,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005089142743617761,
+      "loss": 0.5133,
+      "step": 73260
+    },
+    {
+      "epoch": 3.6391179100029802,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0005088745405781266,
+      "loss": 0.5598,
+      "step": 73270
+    },
+    {
+      "epoch": 3.6396145822985995,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.000508834806794477,
+      "loss": 0.5106,
+      "step": 73280
+    },
+    {
+      "epoch": 3.6401112545942187,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005087950730108275,
+      "loss": 0.5276,
+      "step": 73290
+    },
+    {
+      "epoch": 3.640607926889838,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.000508755339227178,
+      "loss": 0.5294,
+      "step": 73300
+    },
+    {
+      "epoch": 3.6411045991854576,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005087156054435283,
+      "loss": 0.5231,
+      "step": 73310
+    },
+    {
+      "epoch": 3.641601271481077,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005086758716598788,
+      "loss": 0.5216,
+      "step": 73320
+    },
+    {
+      "epoch": 3.642097943776696,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0005086361378762293,
+      "loss": 0.536,
+      "step": 73330
+    },
+    {
+      "epoch": 3.6425946160723157,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005085964040925798,
+      "loss": 0.5273,
+      "step": 73340
+    },
+    {
+      "epoch": 3.643091288367935,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005085566703089302,
+      "loss": 0.5331,
+      "step": 73350
+    },
+    {
+      "epoch": 3.643587960663554,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005085169365252807,
+      "loss": 0.5148,
+      "step": 73360
+    },
+    {
+      "epoch": 3.6440846329591734,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005084772027416311,
+      "loss": 0.5313,
+      "step": 73370
+    },
+    {
+      "epoch": 3.6445813052547926,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0005084374689579815,
+      "loss": 0.5294,
+      "step": 73380
+    },
+    {
+      "epoch": 3.6450779775504123,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.000508397735174332,
+      "loss": 0.5306,
+      "step": 73390
+    },
+    {
+      "epoch": 3.6455746498460315,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0005083580013906825,
+      "loss": 0.5003,
+      "step": 73400
+    },
+    {
+      "epoch": 3.646071322141651,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005083182676070329,
+      "loss": 0.5395,
+      "step": 73410
+    },
+    {
+      "epoch": 3.6465679944372704,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0005082785338233833,
+      "loss": 0.5017,
+      "step": 73420
+    },
+    {
+      "epoch": 3.6470646667328896,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005082388000397338,
+      "loss": 0.5296,
+      "step": 73430
+    },
+    {
+      "epoch": 3.647561339028509,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0005081990662560843,
+      "loss": 0.5262,
+      "step": 73440
+    },
+    {
+      "epoch": 3.648058011324128,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0005081593324724347,
+      "loss": 0.5517,
+      "step": 73450
+    },
+    {
+      "epoch": 3.6485546836197478,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005081195986887852,
+      "loss": 0.5157,
+      "step": 73460
+    },
+    {
+      "epoch": 3.649051355915367,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005080798649051356,
+      "loss": 0.5272,
+      "step": 73470
+    },
+    {
+      "epoch": 3.6495480282109862,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.000508040131121486,
+      "loss": 0.5168,
+      "step": 73480
+    },
+    {
+      "epoch": 3.650044700506606,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005080003973378366,
+      "loss": 0.5559,
+      "step": 73490
+    },
+    {
+      "epoch": 3.650541372802225,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.000507960663554187,
+      "loss": 0.4941,
+      "step": 73500
+    },
+    {
+      "epoch": 3.6510380450978444,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005079209297705374,
+      "loss": 0.5256,
+      "step": 73510
+    },
+    {
+      "epoch": 3.6515347173934636,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005078811959868879,
+      "loss": 0.5269,
+      "step": 73520
+    },
+    {
+      "epoch": 3.6520313896890833,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005078414622032383,
+      "loss": 0.5148,
+      "step": 73530
+    },
+    {
+      "epoch": 3.6525280619847025,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005078017284195888,
+      "loss": 0.5282,
+      "step": 73540
+    },
+    {
+      "epoch": 3.6530247342803217,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005077619946359393,
+      "loss": 0.5345,
+      "step": 73550
+    },
+    {
+      "epoch": 3.6535214065759414,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005077222608522897,
+      "loss": 0.5298,
+      "step": 73560
+    },
+    {
+      "epoch": 3.6540180788715606,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005076825270686401,
+      "loss": 0.5444,
+      "step": 73570
+    },
+    {
+      "epoch": 3.65451475116718,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0005076427932849905,
+      "loss": 0.5165,
+      "step": 73580
+    },
+    {
+      "epoch": 3.655011423462799,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005076030595013411,
+      "loss": 0.5404,
+      "step": 73590
+    },
+    {
+      "epoch": 3.6555080957584187,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005075633257176915,
+      "loss": 0.5279,
+      "step": 73600
+    },
+    {
+      "epoch": 3.656004768054038,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0005075235919340419,
+      "loss": 0.5495,
+      "step": 73610
+    },
+    {
+      "epoch": 3.656501440349657,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005074838581503924,
+      "loss": 0.5339,
+      "step": 73620
+    },
+    {
+      "epoch": 3.656998112645277,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005074441243667429,
+      "loss": 0.4827,
+      "step": 73630
+    },
+    {
+      "epoch": 3.657494784940896,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005074043905830932,
+      "loss": 0.5288,
+      "step": 73640
+    },
+    {
+      "epoch": 3.6579914572365153,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005073646567994438,
+      "loss": 0.5439,
+      "step": 73650
+    },
+    {
+      "epoch": 3.6584881295321345,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005073249230157942,
+      "loss": 0.5207,
+      "step": 73660
+    },
+    {
+      "epoch": 3.658984801827754,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005072851892321446,
+      "loss": 0.5458,
+      "step": 73670
+    },
+    {
+      "epoch": 3.6594814741233734,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0005072454554484951,
+      "loss": 0.5007,
+      "step": 73680
+    },
+    {
+      "epoch": 3.6599781464189927,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0005072057216648456,
+      "loss": 0.5203,
+      "step": 73690
+    },
+    {
+      "epoch": 3.6604748187146123,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.000507165987881196,
+      "loss": 0.5408,
+      "step": 73700
+    },
+    {
+      "epoch": 3.6609714910102316,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005071262540975465,
+      "loss": 0.5383,
+      "step": 73710
+    },
+    {
+      "epoch": 3.661468163305851,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005070865203138969,
+      "loss": 0.5536,
+      "step": 73720
+    },
+    {
+      "epoch": 3.66196483560147,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005070467865302473,
+      "loss": 0.5546,
+      "step": 73730
+    },
+    {
+      "epoch": 3.6624615078970892,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005070070527465979,
+      "loss": 0.5166,
+      "step": 73740
+    },
+    {
+      "epoch": 3.662958180192709,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005069673189629483,
+      "loss": 0.5149,
+      "step": 73750
+    },
+    {
+      "epoch": 3.663454852488328,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005069275851792987,
+      "loss": 0.5278,
+      "step": 73760
+    },
+    {
+      "epoch": 3.663951524783948,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0005068878513956492,
+      "loss": 0.5433,
+      "step": 73770
+    },
+    {
+      "epoch": 3.664448197079567,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0005068481176119996,
+      "loss": 0.515,
+      "step": 73780
+    },
+    {
+      "epoch": 3.6649448693751863,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005068083838283502,
+      "loss": 0.5293,
+      "step": 73790
+    },
+    {
+      "epoch": 3.6654415416708055,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005067686500447005,
+      "loss": 0.5491,
+      "step": 73800
+    },
+    {
+      "epoch": 3.6659382139664247,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.000506728916261051,
+      "loss": 0.5319,
+      "step": 73810
+    },
+    {
+      "epoch": 3.6664348862620444,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0005066891824774015,
+      "loss": 0.5304,
+      "step": 73820
+    },
+    {
+      "epoch": 3.6669315585576636,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005066494486937518,
+      "loss": 0.5555,
+      "step": 73830
+    },
+    {
+      "epoch": 3.667428230853283,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005066097149101024,
+      "loss": 0.5552,
+      "step": 73840
+    },
+    {
+      "epoch": 3.6679249031489025,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005065699811264528,
+      "loss": 0.531,
+      "step": 73850
+    },
+    {
+      "epoch": 3.6684215754445217,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005065302473428032,
+      "loss": 0.5315,
+      "step": 73860
+    },
+    {
+      "epoch": 3.668918247740141,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0005064905135591537,
+      "loss": 0.5336,
+      "step": 73870
+    },
+    {
+      "epoch": 3.66941492003576,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005064507797755041,
+      "loss": 0.5378,
+      "step": 73880
+    },
+    {
+      "epoch": 3.66991159233138,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0005064110459918545,
+      "loss": 0.5245,
+      "step": 73890
+    },
+    {
+      "epoch": 3.670408264626999,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0005063713122082051,
+      "loss": 0.5215,
+      "step": 73900
+    },
+    {
+      "epoch": 3.6709049369226183,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0005063315784245555,
+      "loss": 0.526,
+      "step": 73910
+    },
+    {
+      "epoch": 3.671401609218238,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005062918446409059,
+      "loss": 0.5362,
+      "step": 73920
+    },
+    {
+      "epoch": 3.671898281513857,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005062521108572564,
+      "loss": 0.5355,
+      "step": 73930
+    },
+    {
+      "epoch": 3.6723949538094764,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005062123770736068,
+      "loss": 0.5274,
+      "step": 73940
+    },
+    {
+      "epoch": 3.6728916261050957,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0005061726432899574,
+      "loss": 0.5348,
+      "step": 73950
+    },
+    {
+      "epoch": 3.6733882984007153,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0005061329095063078,
+      "loss": 0.4846,
+      "step": 73960
+    },
+    {
+      "epoch": 3.6738849706963346,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005060931757226582,
+      "loss": 0.5432,
+      "step": 73970
+    },
+    {
+      "epoch": 3.674381642991954,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005060534419390087,
+      "loss": 0.5665,
+      "step": 73980
+    },
+    {
+      "epoch": 3.6748783152875735,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.000506013708155359,
+      "loss": 0.5178,
+      "step": 73990
+    },
+    {
+      "epoch": 3.6753749875831927,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0005059739743717096,
+      "loss": 0.5502,
+      "step": 74000
+    },
+    {
+      "epoch": 3.675871659878812,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005059342405880601,
+      "loss": 0.5316,
+      "step": 74010
+    },
+    {
+      "epoch": 3.676368332174431,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005058945068044104,
+      "loss": 0.521,
+      "step": 74020
+    },
+    {
+      "epoch": 3.676865004470051,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0005058547730207609,
+      "loss": 0.5134,
+      "step": 74030
+    },
+    {
+      "epoch": 3.67736167676567,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005058150392371115,
+      "loss": 0.5453,
+      "step": 74040
+    },
+    {
+      "epoch": 3.6778583490612893,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005057753054534618,
+      "loss": 0.5439,
+      "step": 74050
+    },
+    {
+      "epoch": 3.678355021356909,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005057355716698123,
+      "loss": 0.5378,
+      "step": 74060
+    },
+    {
+      "epoch": 3.678851693652528,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0005056958378861627,
+      "loss": 0.5359,
+      "step": 74070
+    },
+    {
+      "epoch": 3.6793483659481474,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005056561041025132,
+      "loss": 0.5216,
+      "step": 74080
+    },
+    {
+      "epoch": 3.6798450382437666,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005056163703188636,
+      "loss": 0.5172,
+      "step": 74090
+    },
+    {
+      "epoch": 3.680341710539386,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005055766365352141,
+      "loss": 0.5365,
+      "step": 74100
+    },
+    {
+      "epoch": 3.6808383828350055,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005055369027515646,
+      "loss": 0.5234,
+      "step": 74110
+    },
+    {
+      "epoch": 3.6813350551306248,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.000505497168967915,
+      "loss": 0.5581,
+      "step": 74120
+    },
+    {
+      "epoch": 3.6818317274262444,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005054574351842654,
+      "loss": 0.5291,
+      "step": 74130
+    },
+    {
+      "epoch": 3.6823283997218637,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.000505417701400616,
+      "loss": 0.5433,
+      "step": 74140
+    },
+    {
+      "epoch": 3.682825072017483,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005053779676169664,
+      "loss": 0.5088,
+      "step": 74150
+    },
+    {
+      "epoch": 3.683321744313102,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005053382338333168,
+      "loss": 0.552,
+      "step": 74160
+    },
+    {
+      "epoch": 3.6838184166087213,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005052985000496673,
+      "loss": 0.5277,
+      "step": 74170
+    },
+    {
+      "epoch": 3.684315088904341,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005052587662660176,
+      "loss": 0.5309,
+      "step": 74180
+    },
+    {
+      "epoch": 3.6848117611999602,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005052190324823681,
+      "loss": 0.5292,
+      "step": 74190
+    },
+    {
+      "epoch": 3.6853084334955795,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0005051792986987187,
+      "loss": 0.5,
+      "step": 74200
+    },
+    {
+      "epoch": 3.685805105791199,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.000505139564915069,
+      "loss": 0.5177,
+      "step": 74210
+    },
+    {
+      "epoch": 3.6863017780868184,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0005050998311314195,
+      "loss": 0.5462,
+      "step": 74220
+    },
+    {
+      "epoch": 3.6867984503824376,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00050506009734777,
+      "loss": 0.5618,
+      "step": 74230
+    },
+    {
+      "epoch": 3.687295122678057,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005050203635641204,
+      "loss": 0.5204,
+      "step": 74240
+    },
+    {
+      "epoch": 3.6877917949736765,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005049806297804709,
+      "loss": 0.5339,
+      "step": 74250
+    },
+    {
+      "epoch": 3.6882884672692957,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005049408959968213,
+      "loss": 0.5595,
+      "step": 74260
+    },
+    {
+      "epoch": 3.688785139564915,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0005049011622131718,
+      "loss": 0.5333,
+      "step": 74270
+    },
+    {
+      "epoch": 3.6892818118605346,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0005048614284295222,
+      "loss": 0.5175,
+      "step": 74280
+    },
+    {
+      "epoch": 3.689778484156154,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005048216946458726,
+      "loss": 0.5102,
+      "step": 74290
+    },
+    {
+      "epoch": 3.690275156451773,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005047819608622232,
+      "loss": 0.5456,
+      "step": 74300
+    },
+    {
+      "epoch": 3.6907718287473923,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005047422270785736,
+      "loss": 0.5277,
+      "step": 74310
+    },
+    {
+      "epoch": 3.691268501043012,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.000504702493294924,
+      "loss": 0.5348,
+      "step": 74320
+    },
+    {
+      "epoch": 3.691765173338631,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0005046627595112745,
+      "loss": 0.5225,
+      "step": 74330
+    },
+    {
+      "epoch": 3.6922618456342504,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005046230257276249,
+      "loss": 0.5521,
+      "step": 74340
+    },
+    {
+      "epoch": 3.69275851792987,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005045832919439754,
+      "loss": 0.5586,
+      "step": 74350
+    },
+    {
+      "epoch": 3.6932551902254893,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0005045435581603259,
+      "loss": 0.524,
+      "step": 74360
+    },
+    {
+      "epoch": 3.6937518625211085,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0005045038243766763,
+      "loss": 0.5222,
+      "step": 74370
+    },
+    {
+      "epoch": 3.6942485348167278,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0005044640905930267,
+      "loss": 0.5279,
+      "step": 74380
+    },
+    {
+      "epoch": 3.694745207112347,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005044243568093772,
+      "loss": 0.5513,
+      "step": 74390
+    },
+    {
+      "epoch": 3.6952418794079667,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005043846230257277,
+      "loss": 0.5075,
+      "step": 74400
+    },
+    {
+      "epoch": 3.695738551703586,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0005043448892420781,
+      "loss": 0.5501,
+      "step": 74410
+    },
+    {
+      "epoch": 3.6962352239992056,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0005043051554584286,
+      "loss": 0.5157,
+      "step": 74420
+    },
+    {
+      "epoch": 3.696731896294825,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.000504265421674779,
+      "loss": 0.5792,
+      "step": 74430
+    },
+    {
+      "epoch": 3.697228568590444,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005042256878911294,
+      "loss": 0.5353,
+      "step": 74440
+    },
+    {
+      "epoch": 3.6977252408860632,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005041859541074799,
+      "loss": 0.5202,
+      "step": 74450
+    },
+    {
+      "epoch": 3.6982219131816825,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005041462203238304,
+      "loss": 0.5369,
+      "step": 74460
+    },
+    {
+      "epoch": 3.698718585477302,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005041064865401808,
+      "loss": 0.5329,
+      "step": 74470
+    },
+    {
+      "epoch": 3.6992152577729214,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0005040667527565312,
+      "loss": 0.5314,
+      "step": 74480
+    },
+    {
+      "epoch": 3.699711930068541,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0005040270189728817,
+      "loss": 0.5389,
+      "step": 74490
+    },
+    {
+      "epoch": 3.7002086023641603,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005039872851892322,
+      "loss": 0.55,
+      "step": 74500
+    },
+    {
+      "epoch": 3.7007052746597795,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0005039475514055826,
+      "loss": 0.5334,
+      "step": 74510
+    },
+    {
+      "epoch": 3.7012019469553987,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005039078176219331,
+      "loss": 0.5203,
+      "step": 74520
+    },
+    {
+      "epoch": 3.701698619251018,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005038680838382835,
+      "loss": 0.5108,
+      "step": 74530
+    },
+    {
+      "epoch": 3.7021952915466376,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0005038283500546339,
+      "loss": 0.5133,
+      "step": 74540
+    },
+    {
+      "epoch": 3.702691963842257,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005037886162709845,
+      "loss": 0.5255,
+      "step": 74550
+    },
+    {
+      "epoch": 3.703188636137876,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0005037488824873349,
+      "loss": 0.4999,
+      "step": 74560
+    },
+    {
+      "epoch": 3.7036853084334957,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0005037091487036853,
+      "loss": 0.531,
+      "step": 74570
+    },
+    {
+      "epoch": 3.704181980729115,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0005036694149200358,
+      "loss": 0.5445,
+      "step": 74580
+    },
+    {
+      "epoch": 3.704678653024734,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0005036296811363862,
+      "loss": 0.5408,
+      "step": 74590
+    },
+    {
+      "epoch": 3.7051753253203534,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0005035899473527367,
+      "loss": 0.5694,
+      "step": 74600
+    },
+    {
+      "epoch": 3.705671997615973,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0005035502135690872,
+      "loss": 0.5388,
+      "step": 74610
+    },
+    {
+      "epoch": 3.7061686699115923,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0005035104797854376,
+      "loss": 0.5405,
+      "step": 74620
+    },
+    {
+      "epoch": 3.7066653422072116,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.000503470746001788,
+      "loss": 0.5164,
+      "step": 74630
+    },
+    {
+      "epoch": 3.7071620145028312,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0005034310122181385,
+      "loss": 0.5371,
+      "step": 74640
+    },
+    {
+      "epoch": 3.7076586867984505,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.000503391278434489,
+      "loss": 0.536,
+      "step": 74650
+    },
+    {
+      "epoch": 3.7081553590940697,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0005033515446508394,
+      "loss": 0.5447,
+      "step": 74660
+    },
+    {
+      "epoch": 3.708652031389689,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0005033118108671898,
+      "loss": 0.5733,
+      "step": 74670
+    },
+    {
+      "epoch": 3.7091487036853086,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0005032720770835403,
+      "loss": 0.5171,
+      "step": 74680
+    },
+    {
+      "epoch": 3.709645375980928,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005032323432998908,
+      "loss": 0.5562,
+      "step": 74690
+    },
+    {
+      "epoch": 3.710142048276547,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0005031926095162412,
+      "loss": 0.5388,
+      "step": 74700
+    },
+    {
+      "epoch": 3.7106387205721667,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005031528757325917,
+      "loss": 0.5385,
+      "step": 74710
+    },
+    {
+      "epoch": 3.711135392867786,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005031131419489421,
+      "loss": 0.5191,
+      "step": 74720
+    },
+    {
+      "epoch": 3.711632065163405,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005030734081652925,
+      "loss": 0.5243,
+      "step": 74730
+    },
+    {
+      "epoch": 3.7121287374590244,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.000503033674381643,
+      "loss": 0.5402,
+      "step": 74740
+    },
+    {
+      "epoch": 3.7126254097546436,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0005029939405979935,
+      "loss": 0.5371,
+      "step": 74750
+    },
+    {
+      "epoch": 3.7131220820502633,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0005029542068143439,
+      "loss": 0.5397,
+      "step": 74760
+    },
+    {
+      "epoch": 3.7136187543458825,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005029144730306944,
+      "loss": 0.5406,
+      "step": 74770
+    },
+    {
+      "epoch": 3.714115426641502,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005028747392470448,
+      "loss": 0.5853,
+      "step": 74780
+    },
+    {
+      "epoch": 3.7146120989371214,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0005028350054633952,
+      "loss": 0.5466,
+      "step": 74790
+    },
+    {
+      "epoch": 3.7151087712327406,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0005027952716797458,
+      "loss": 0.5634,
+      "step": 74800
+    },
+    {
+      "epoch": 3.71560544352836,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0005027555378960962,
+      "loss": 0.526,
+      "step": 74810
+    },
+    {
+      "epoch": 3.716102115823979,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005027158041124467,
+      "loss": 0.5252,
+      "step": 74820
+    },
+    {
+      "epoch": 3.7165987881195988,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0005026760703287971,
+      "loss": 0.5454,
+      "step": 74830
+    },
+    {
+      "epoch": 3.717095460415218,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0005026363365451475,
+      "loss": 0.5045,
+      "step": 74840
+    },
+    {
+      "epoch": 3.7175921327108377,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0005025966027614981,
+      "loss": 0.5203,
+      "step": 74850
+    },
+    {
+      "epoch": 3.718088805006457,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005025568689778484,
+      "loss": 0.5681,
+      "step": 74860
+    },
+    {
+      "epoch": 3.718585477302076,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005025171351941989,
+      "loss": 0.5234,
+      "step": 74870
+    },
+    {
+      "epoch": 3.7190821495976953,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005024774014105494,
+      "loss": 0.5626,
+      "step": 74880
+    },
+    {
+      "epoch": 3.7195788218933146,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0005024376676268997,
+      "loss": 0.5342,
+      "step": 74890
+    },
+    {
+      "epoch": 3.7200754941889342,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0005023979338432503,
+      "loss": 0.5172,
+      "step": 74900
+    },
+    {
+      "epoch": 3.7205721664845535,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0005023582000596008,
+      "loss": 0.5401,
+      "step": 74910
+    },
+    {
+      "epoch": 3.7210688387801727,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0005023184662759511,
+      "loss": 0.5451,
+      "step": 74920
+    },
+    {
+      "epoch": 3.7215655110757924,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0005022787324923016,
+      "loss": 0.534,
+      "step": 74930
+    },
+    {
+      "epoch": 3.7220621833714116,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.000502238998708652,
+      "loss": 0.5271,
+      "step": 74940
+    },
+    {
+      "epoch": 3.722558855667031,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005021992649250024,
+      "loss": 0.5351,
+      "step": 74950
+    },
+    {
+      "epoch": 3.72305552796265,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.000502159531141353,
+      "loss": 0.5425,
+      "step": 74960
+    },
+    {
+      "epoch": 3.7235522002582697,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005021197973577034,
+      "loss": 0.5562,
+      "step": 74970
+    },
+    {
+      "epoch": 3.724048872553889,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0005020800635740539,
+      "loss": 0.5429,
+      "step": 74980
+    },
+    {
+      "epoch": 3.724545544849508,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005020403297904043,
+      "loss": 0.5185,
+      "step": 74990
+    },
+    {
+      "epoch": 3.725042217145128,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0005020005960067548,
+      "loss": 0.5465,
+      "step": 75000
+    },
+    {
+      "epoch": 3.725538889440747,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0005019608622231053,
+      "loss": 0.5251,
+      "step": 75010
+    },
+    {
+      "epoch": 3.7260355617363663,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0005019211284394557,
+      "loss": 0.5146,
+      "step": 75020
+    },
+    {
+      "epoch": 3.7265322340319855,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005018813946558061,
+      "loss": 0.499,
+      "step": 75030
+    },
+    {
+      "epoch": 3.727028906327605,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005018416608721566,
+      "loss": 0.5298,
+      "step": 75040
+    },
+    {
+      "epoch": 3.7275255786232244,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0005018019270885069,
+      "loss": 0.5222,
+      "step": 75050
+    },
+    {
+      "epoch": 3.7280222509188436,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0005017621933048575,
+      "loss": 0.5474,
+      "step": 75060
+    },
+    {
+      "epoch": 3.7285189232144633,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.000501722459521208,
+      "loss": 0.5366,
+      "step": 75070
+    },
+    {
+      "epoch": 3.7290155955100825,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0005016827257375583,
+      "loss": 0.5366,
+      "step": 75080
+    },
+    {
+      "epoch": 3.7295122678057018,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0005016429919539088,
+      "loss": 0.5356,
+      "step": 75090
+    },
+    {
+      "epoch": 3.730008940101321,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0005016032581702594,
+      "loss": 0.5207,
+      "step": 75100
+    },
+    {
+      "epoch": 3.7305056123969402,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0005015635243866097,
+      "loss": 0.5548,
+      "step": 75110
+    },
+    {
+      "epoch": 3.73100228469256,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0005015237906029602,
+      "loss": 0.558,
+      "step": 75120
+    },
+    {
+      "epoch": 3.731498956988179,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0005014840568193106,
+      "loss": 0.5615,
+      "step": 75130
+    },
+    {
+      "epoch": 3.731995629283799,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0005014443230356611,
+      "loss": 0.5319,
+      "step": 75140
+    },
+    {
+      "epoch": 3.732492301579418,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005014045892520116,
+      "loss": 0.5452,
+      "step": 75150
+    },
+    {
+      "epoch": 3.7329889738750373,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.000501364855468362,
+      "loss": 0.5418,
+      "step": 75160
+    },
+    {
+      "epoch": 3.7334856461706565,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005013251216847125,
+      "loss": 0.5408,
+      "step": 75170
+    },
+    {
+      "epoch": 3.7339823184662757,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0005012853879010629,
+      "loss": 0.5412,
+      "step": 75180
+    },
+    {
+      "epoch": 3.7344789907618954,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0005012456541174133,
+      "loss": 0.5283,
+      "step": 75190
+    },
+    {
+      "epoch": 3.7349756630575146,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0005012059203337639,
+      "loss": 0.5539,
+      "step": 75200
+    },
+    {
+      "epoch": 3.7354723353531343,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005011661865501143,
+      "loss": 0.5243,
+      "step": 75210
+    },
+    {
+      "epoch": 3.7359690076487535,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0005011264527664647,
+      "loss": 0.5207,
+      "step": 75220
+    },
+    {
+      "epoch": 3.7364656799443727,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0005010867189828152,
+      "loss": 0.5092,
+      "step": 75230
+    },
+    {
+      "epoch": 3.736962352239992,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0005010469851991655,
+      "loss": 0.5549,
+      "step": 75240
+    },
+    {
+      "epoch": 3.737459024535611,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.000501007251415516,
+      "loss": 0.5267,
+      "step": 75250
+    },
+    {
+      "epoch": 3.737955696831231,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005009675176318666,
+      "loss": 0.504,
+      "step": 75260
+    },
+    {
+      "epoch": 3.73845236912685,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.000500927783848217,
+      "loss": 0.5222,
+      "step": 75270
+    },
+    {
+      "epoch": 3.7389490414224693,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005008880500645674,
+      "loss": 0.5661,
+      "step": 75280
+    },
+    {
+      "epoch": 3.739445713718089,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0005008483162809179,
+      "loss": 0.5216,
+      "step": 75290
+    },
+    {
+      "epoch": 3.739942386013708,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0005008085824972683,
+      "loss": 0.5375,
+      "step": 75300
+    },
+    {
+      "epoch": 3.7404390583093274,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005007688487136188,
+      "loss": 0.5243,
+      "step": 75310
+    },
+    {
+      "epoch": 3.7409357306049467,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0005007291149299692,
+      "loss": 0.5113,
+      "step": 75320
+    },
+    {
+      "epoch": 3.7414324029005663,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0005006893811463197,
+      "loss": 0.5147,
+      "step": 75330
+    },
+    {
+      "epoch": 3.7419290751961856,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0005006496473626701,
+      "loss": 0.535,
+      "step": 75340
+    },
+    {
+      "epoch": 3.742425747491805,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0005006099135790205,
+      "loss": 0.5412,
+      "step": 75350
+    },
+    {
+      "epoch": 3.7429224197874245,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0005005701797953711,
+      "loss": 0.5162,
+      "step": 75360
+    },
+    {
+      "epoch": 3.7434190920830437,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0005005304460117215,
+      "loss": 0.5369,
+      "step": 75370
+    },
+    {
+      "epoch": 3.743915764378663,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0005004907122280719,
+      "loss": 0.5315,
+      "step": 75380
+    },
+    {
+      "epoch": 3.744412436674282,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0005004509784444224,
+      "loss": 0.538,
+      "step": 75390
+    },
+    {
+      "epoch": 3.744909108969902,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0005004112446607728,
+      "loss": 0.5474,
+      "step": 75400
+    },
+    {
+      "epoch": 3.745405781265521,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0005003715108771233,
+      "loss": 0.5329,
+      "step": 75410
+    },
+    {
+      "epoch": 3.7459024535611403,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0005003317770934738,
+      "loss": 0.5152,
+      "step": 75420
+    },
+    {
+      "epoch": 3.74639912585676,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0005002920433098242,
+      "loss": 0.5356,
+      "step": 75430
+    },
+    {
+      "epoch": 3.746895798152379,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0005002523095261746,
+      "loss": 0.5321,
+      "step": 75440
+    },
+    {
+      "epoch": 3.7473924704479984,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0005002125757425251,
+      "loss": 0.5199,
+      "step": 75450
+    },
+    {
+      "epoch": 3.7478891427436176,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0005001728419588756,
+      "loss": 0.5606,
+      "step": 75460
+    },
+    {
+      "epoch": 3.748385815039237,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.000500133108175226,
+      "loss": 0.5587,
+      "step": 75470
+    },
+    {
+      "epoch": 3.7488824873348565,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005000933743915765,
+      "loss": 0.5451,
+      "step": 75480
+    },
+    {
+      "epoch": 3.7493791596304757,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0005000536406079269,
+      "loss": 0.5322,
+      "step": 75490
+    },
+    {
+      "epoch": 3.7498758319260954,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0005000139068242773,
+      "loss": 0.5317,
+      "step": 75500
+    },
+    {
+      "epoch": 3.7503725042217146,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0004999741730406278,
+      "loss": 0.5438,
+      "step": 75510
+    },
+    {
+      "epoch": 3.750869176517334,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0004999344392569783,
+      "loss": 0.5394,
+      "step": 75520
+    },
+    {
+      "epoch": 3.751365848812953,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0004998947054733287,
+      "loss": 0.5357,
+      "step": 75530
+    },
+    {
+      "epoch": 3.7518625211085723,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0004998549716896791,
+      "loss": 0.524,
+      "step": 75540
+    },
+    {
+      "epoch": 3.752359193404192,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0004998152379060296,
+      "loss": 0.5782,
+      "step": 75550
+    },
+    {
+      "epoch": 3.752855865699811,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004997755041223801,
+      "loss": 0.523,
+      "step": 75560
+    },
+    {
+      "epoch": 3.7533525379954304,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004997357703387305,
+      "loss": 0.5524,
+      "step": 75570
+    },
+    {
+      "epoch": 3.75384921029105,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.000499696036555081,
+      "loss": 0.5031,
+      "step": 75580
+    },
+    {
+      "epoch": 3.7543458825866693,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004996563027714314,
+      "loss": 0.5556,
+      "step": 75590
+    },
+    {
+      "epoch": 3.7548425548822886,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004996165689877818,
+      "loss": 0.5152,
+      "step": 75600
+    },
+    {
+      "epoch": 3.755339227177908,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0004995768352041324,
+      "loss": 0.5316,
+      "step": 75610
+    },
+    {
+      "epoch": 3.7558358994735275,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0004995371014204828,
+      "loss": 0.5134,
+      "step": 75620
+    },
+    {
+      "epoch": 3.7563325717691467,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0004994973676368332,
+      "loss": 0.5617,
+      "step": 75630
+    },
+    {
+      "epoch": 3.756829244064766,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0004994576338531837,
+      "loss": 0.528,
+      "step": 75640
+    },
+    {
+      "epoch": 3.7573259163603856,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004994179000695341,
+      "loss": 0.5361,
+      "step": 75650
+    },
+    {
+      "epoch": 3.757822588656005,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0004993781662858846,
+      "loss": 0.5295,
+      "step": 75660
+    },
+    {
+      "epoch": 3.758319260951624,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004993384325022351,
+      "loss": 0.5404,
+      "step": 75670
+    },
+    {
+      "epoch": 3.7588159332472433,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0004992986987185855,
+      "loss": 0.5221,
+      "step": 75680
+    },
+    {
+      "epoch": 3.759312605542863,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0004992589649349359,
+      "loss": 0.5524,
+      "step": 75690
+    },
+    {
+      "epoch": 3.759809277838482,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0004992192311512864,
+      "loss": 0.5441,
+      "step": 75700
+    },
+    {
+      "epoch": 3.7603059501341014,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0004991794973676369,
+      "loss": 0.5298,
+      "step": 75710
+    },
+    {
+      "epoch": 3.760802622429721,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0004991397635839874,
+      "loss": 0.5343,
+      "step": 75720
+    },
+    {
+      "epoch": 3.7612992947253403,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0004991000298003377,
+      "loss": 0.5252,
+      "step": 75730
+    },
+    {
+      "epoch": 3.7617959670209595,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004990602960166882,
+      "loss": 0.5356,
+      "step": 75740
+    },
+    {
+      "epoch": 3.7622926393165788,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0004990205622330387,
+      "loss": 0.5323,
+      "step": 75750
+    },
+    {
+      "epoch": 3.7627893116121984,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0004989808284493891,
+      "loss": 0.5198,
+      "step": 75760
+    },
+    {
+      "epoch": 3.7632859839078177,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0004989410946657396,
+      "loss": 0.5332,
+      "step": 75770
+    },
+    {
+      "epoch": 3.763782656203437,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00049890136088209,
+      "loss": 0.5421,
+      "step": 75780
+    },
+    {
+      "epoch": 3.7642793284990566,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004988616270984404,
+      "loss": 0.4969,
+      "step": 75790
+    },
+    {
+      "epoch": 3.764776000794676,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004988218933147909,
+      "loss": 0.5289,
+      "step": 75800
+    },
+    {
+      "epoch": 3.765272673090295,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0004987821595311414,
+      "loss": 0.5358,
+      "step": 75810
+    },
+    {
+      "epoch": 3.7657693453859142,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0004987424257474918,
+      "loss": 0.5283,
+      "step": 75820
+    },
+    {
+      "epoch": 3.7662660176815335,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0004987026919638423,
+      "loss": 0.5582,
+      "step": 75830
+    },
+    {
+      "epoch": 3.766762689977153,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004986629581801927,
+      "loss": 0.526,
+      "step": 75840
+    },
+    {
+      "epoch": 3.7672593622727724,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0004986232243965431,
+      "loss": 0.5216,
+      "step": 75850
+    },
+    {
+      "epoch": 3.767756034568392,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0004985834906128937,
+      "loss": 0.5394,
+      "step": 75860
+    },
+    {
+      "epoch": 3.7682527068640113,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0004985437568292441,
+      "loss": 0.5392,
+      "step": 75870
+    },
+    {
+      "epoch": 3.7687493791596305,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004985040230455946,
+      "loss": 0.5583,
+      "step": 75880
+    },
+    {
+      "epoch": 3.7692460514552497,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.000498464289261945,
+      "loss": 0.5372,
+      "step": 75890
+    },
+    {
+      "epoch": 3.769742723750869,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0004984245554782954,
+      "loss": 0.5437,
+      "step": 75900
+    },
+    {
+      "epoch": 3.7702393960464886,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.000498384821694646,
+      "loss": 0.5374,
+      "step": 75910
+    },
+    {
+      "epoch": 3.770736068342108,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0004983450879109963,
+      "loss": 0.5581,
+      "step": 75920
+    },
+    {
+      "epoch": 3.771232740637727,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0004983053541273468,
+      "loss": 0.5358,
+      "step": 75930
+    },
+    {
+      "epoch": 3.7717294129333467,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0004982656203436973,
+      "loss": 0.5459,
+      "step": 75940
+    },
+    {
+      "epoch": 3.772226085228966,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0004982258865600476,
+      "loss": 0.5242,
+      "step": 75950
+    },
+    {
+      "epoch": 3.772722757524585,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004981861527763982,
+      "loss": 0.5467,
+      "step": 75960
+    },
+    {
+      "epoch": 3.7732194298202044,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004981464189927487,
+      "loss": 0.5221,
+      "step": 75970
+    },
+    {
+      "epoch": 3.773716102115824,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.000498106685209099,
+      "loss": 0.5415,
+      "step": 75980
+    },
+    {
+      "epoch": 3.7742127744114433,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0004980669514254495,
+      "loss": 0.5212,
+      "step": 75990
+    },
+    {
+      "epoch": 3.7747094467070625,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004980272176417999,
+      "loss": 0.5402,
+      "step": 76000
+    },
+    {
+      "epoch": 3.775206119002682,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0004979874838581504,
+      "loss": 0.5187,
+      "step": 76010
+    },
+    {
+      "epoch": 3.7757027912983014,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0004979477500745009,
+      "loss": 0.5301,
+      "step": 76020
+    },
+    {
+      "epoch": 3.7761994635939207,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004979080162908513,
+      "loss": 0.5241,
+      "step": 76030
+    },
+    {
+      "epoch": 3.77669613588954,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0004978682825072018,
+      "loss": 0.5217,
+      "step": 76040
+    },
+    {
+      "epoch": 3.7771928081851596,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0004978285487235522,
+      "loss": 0.5234,
+      "step": 76050
+    },
+    {
+      "epoch": 3.777689480480779,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0004977888149399027,
+      "loss": 0.5231,
+      "step": 76060
+    },
+    {
+      "epoch": 3.778186152776398,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0004977490811562532,
+      "loss": 0.5186,
+      "step": 76070
+    },
+    {
+      "epoch": 3.7786828250720177,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0004977093473726036,
+      "loss": 0.5255,
+      "step": 76080
+    },
+    {
+      "epoch": 3.779179497367637,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.000497669613588954,
+      "loss": 0.5878,
+      "step": 76090
+    },
+    {
+      "epoch": 3.779676169663256,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0004976298798053045,
+      "loss": 0.5512,
+      "step": 76100
+    },
+    {
+      "epoch": 3.7801728419588754,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0004975901460216548,
+      "loss": 0.5578,
+      "step": 76110
+    },
+    {
+      "epoch": 3.780669514254495,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004975504122380054,
+      "loss": 0.5282,
+      "step": 76120
+    },
+    {
+      "epoch": 3.7811661865501143,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0004975106784543559,
+      "loss": 0.5399,
+      "step": 76130
+    },
+    {
+      "epoch": 3.7816628588457335,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0004974709446707062,
+      "loss": 0.5211,
+      "step": 76140
+    },
+    {
+      "epoch": 3.782159531141353,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004974312108870567,
+      "loss": 0.4971,
+      "step": 76150
+    },
+    {
+      "epoch": 3.7826562034369724,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004973914771034073,
+      "loss": 0.5228,
+      "step": 76160
+    },
+    {
+      "epoch": 3.7831528757325916,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0004973517433197577,
+      "loss": 0.5268,
+      "step": 76170
+    },
+    {
+      "epoch": 3.783649548028211,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004973120095361081,
+      "loss": 0.5358,
+      "step": 76180
+    },
+    {
+      "epoch": 3.78414622032383,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004972722757524585,
+      "loss": 0.5051,
+      "step": 76190
+    },
+    {
+      "epoch": 3.7846428926194497,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.000497232541968809,
+      "loss": 0.552,
+      "step": 76200
+    },
+    {
+      "epoch": 3.785139564915069,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004971928081851595,
+      "loss": 0.5174,
+      "step": 76210
+    },
+    {
+      "epoch": 3.7856362372106886,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004971530744015099,
+      "loss": 0.5515,
+      "step": 76220
+    },
+    {
+      "epoch": 3.786132909506308,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004971133406178604,
+      "loss": 0.4994,
+      "step": 76230
+    },
+    {
+      "epoch": 3.786629581801927,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0004970736068342108,
+      "loss": 0.5371,
+      "step": 76240
+    },
+    {
+      "epoch": 3.7871262540975463,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0004970338730505612,
+      "loss": 0.5338,
+      "step": 76250
+    },
+    {
+      "epoch": 3.7876229263931656,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0004969941392669118,
+      "loss": 0.536,
+      "step": 76260
+    },
+    {
+      "epoch": 3.7881195986887852,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0004969544054832622,
+      "loss": 0.5196,
+      "step": 76270
+    },
+    {
+      "epoch": 3.7886162709844045,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0004969146716996126,
+      "loss": 0.5345,
+      "step": 76280
+    },
+    {
+      "epoch": 3.7891129432800237,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004968749379159631,
+      "loss": 0.517,
+      "step": 76290
+    },
+    {
+      "epoch": 3.7896096155756434,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0004968352041323135,
+      "loss": 0.5606,
+      "step": 76300
+    },
+    {
+      "epoch": 3.7901062878712626,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.000496795470348664,
+      "loss": 0.5667,
+      "step": 76310
+    },
+    {
+      "epoch": 3.790602960166882,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0004967557365650145,
+      "loss": 0.5437,
+      "step": 76320
+    },
+    {
+      "epoch": 3.791099632462501,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0004967160027813649,
+      "loss": 0.5455,
+      "step": 76330
+    },
+    {
+      "epoch": 3.7915963047581207,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0004966762689977153,
+      "loss": 0.5301,
+      "step": 76340
+    },
+    {
+      "epoch": 3.79209297705374,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0004966365352140658,
+      "loss": 0.5488,
+      "step": 76350
+    },
+    {
+      "epoch": 3.792589649349359,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004965968014304163,
+      "loss": 0.5309,
+      "step": 76360
+    },
+    {
+      "epoch": 3.793086321644979,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004965570676467667,
+      "loss": 0.5257,
+      "step": 76370
+    },
+    {
+      "epoch": 3.793582993940598,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0004965173338631171,
+      "loss": 0.5463,
+      "step": 76380
+    },
+    {
+      "epoch": 3.7940796662362173,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004964776000794676,
+      "loss": 0.5615,
+      "step": 76390
+    },
+    {
+      "epoch": 3.7945763385318365,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.000496437866295818,
+      "loss": 0.5293,
+      "step": 76400
+    },
+    {
+      "epoch": 3.795073010827456,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004963981325121684,
+      "loss": 0.5131,
+      "step": 76410
+    },
+    {
+      "epoch": 3.7955696831230754,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.000496358398728519,
+      "loss": 0.5213,
+      "step": 76420
+    },
+    {
+      "epoch": 3.7960663554186946,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0004963186649448694,
+      "loss": 0.5412,
+      "step": 76430
+    },
+    {
+      "epoch": 3.7965630277143143,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0004962789311612198,
+      "loss": 0.5338,
+      "step": 76440
+    },
+    {
+      "epoch": 3.7970597000099335,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0004962391973775703,
+      "loss": 0.5537,
+      "step": 76450
+    },
+    {
+      "epoch": 3.7975563723055528,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0004961994635939207,
+      "loss": 0.5441,
+      "step": 76460
+    },
+    {
+      "epoch": 3.798053044601172,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0004961597298102712,
+      "loss": 0.5135,
+      "step": 76470
+    },
+    {
+      "epoch": 3.7985497168967917,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0004961199960266217,
+      "loss": 0.5597,
+      "step": 76480
+    },
+    {
+      "epoch": 3.799046389192411,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0004960802622429721,
+      "loss": 0.5418,
+      "step": 76490
+    },
+    {
+      "epoch": 3.79954306148803,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004960405284593225,
+      "loss": 0.5386,
+      "step": 76500
+    },
+    {
+      "epoch": 3.80003973378365,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.000496000794675673,
+      "loss": 0.5257,
+      "step": 76510
+    },
+    {
+      "epoch": 3.800536406079269,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0004959610608920235,
+      "loss": 0.5425,
+      "step": 76520
+    },
+    {
+      "epoch": 3.8010330783748882,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0004959213271083739,
+      "loss": 0.5369,
+      "step": 76530
+    },
+    {
+      "epoch": 3.8015297506705075,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0004958815933247244,
+      "loss": 0.5424,
+      "step": 76540
+    },
+    {
+      "epoch": 3.8020264229661267,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0004958418595410748,
+      "loss": 0.5137,
+      "step": 76550
+    },
+    {
+      "epoch": 3.8025230952617464,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0004958021257574252,
+      "loss": 0.5386,
+      "step": 76560
+    },
+    {
+      "epoch": 3.8030197675573656,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004957623919737758,
+      "loss": 0.514,
+      "step": 76570
+    },
+    {
+      "epoch": 3.8035164398529853,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0004957226581901262,
+      "loss": 0.535,
+      "step": 76580
+    },
+    {
+      "epoch": 3.8040131121486045,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004956829244064766,
+      "loss": 0.5503,
+      "step": 76590
+    },
+    {
+      "epoch": 3.8045097844442237,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000495643190622827,
+      "loss": 0.5156,
+      "step": 76600
+    },
+    {
+      "epoch": 3.805006456739843,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004956034568391775,
+      "loss": 0.5681,
+      "step": 76610
+    },
+    {
+      "epoch": 3.805503129035462,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0004955637230555281,
+      "loss": 0.5347,
+      "step": 76620
+    },
+    {
+      "epoch": 3.805999801331082,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0004955239892718784,
+      "loss": 0.5676,
+      "step": 76630
+    },
+    {
+      "epoch": 3.806496473626701,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004954842554882289,
+      "loss": 0.5308,
+      "step": 76640
+    },
+    {
+      "epoch": 3.8069931459223203,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004954445217045793,
+      "loss": 0.5314,
+      "step": 76650
+    },
+    {
+      "epoch": 3.80748981821794,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0004954047879209297,
+      "loss": 0.5216,
+      "step": 76660
+    },
+    {
+      "epoch": 3.807986490513559,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0004953650541372803,
+      "loss": 0.5433,
+      "step": 76670
+    },
+    {
+      "epoch": 3.8084831628091784,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0004953253203536307,
+      "loss": 0.5521,
+      "step": 76680
+    },
+    {
+      "epoch": 3.8089798351047977,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004952855865699811,
+      "loss": 0.5234,
+      "step": 76690
+    },
+    {
+      "epoch": 3.8094765074004173,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0004952458527863316,
+      "loss": 0.4972,
+      "step": 76700
+    },
+    {
+      "epoch": 3.8099731796960365,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.000495206119002682,
+      "loss": 0.5387,
+      "step": 76710
+    },
+    {
+      "epoch": 3.8104698519916558,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0004951663852190325,
+      "loss": 0.5169,
+      "step": 76720
+    },
+    {
+      "epoch": 3.8109665242872754,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.000495126651435383,
+      "loss": 0.5121,
+      "step": 76730
+    },
+    {
+      "epoch": 3.8114631965828947,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0004950869176517334,
+      "loss": 0.5538,
+      "step": 76740
+    },
+    {
+      "epoch": 3.811959868878514,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004950471838680838,
+      "loss": 0.529,
+      "step": 76750
+    },
+    {
+      "epoch": 3.812456541174133,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0004950074500844343,
+      "loss": 0.5366,
+      "step": 76760
+    },
+    {
+      "epoch": 3.812953213469753,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0004949677163007848,
+      "loss": 0.5343,
+      "step": 76770
+    },
+    {
+      "epoch": 3.813449885765372,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0004949279825171353,
+      "loss": 0.5588,
+      "step": 76780
+    },
+    {
+      "epoch": 3.8139465580609913,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0004948882487334856,
+      "loss": 0.5383,
+      "step": 76790
+    },
+    {
+      "epoch": 3.814443230356611,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0004948485149498361,
+      "loss": 0.5263,
+      "step": 76800
+    },
+    {
+      "epoch": 3.81493990265223,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0004948087811661867,
+      "loss": 0.5515,
+      "step": 76810
+    },
+    {
+      "epoch": 3.8154365749478494,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.000494769047382537,
+      "loss": 0.515,
+      "step": 76820
+    },
+    {
+      "epoch": 3.8159332472434686,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0004947293135988875,
+      "loss": 0.5327,
+      "step": 76830
+    },
+    {
+      "epoch": 3.816429919539088,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.000494689579815238,
+      "loss": 0.5308,
+      "step": 76840
+    },
+    {
+      "epoch": 3.8169265918347075,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0004946498460315883,
+      "loss": 0.5392,
+      "step": 76850
+    },
+    {
+      "epoch": 3.8174232641303267,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004946101122479388,
+      "loss": 0.5223,
+      "step": 76860
+    },
+    {
+      "epoch": 3.8179199364259464,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0004945703784642893,
+      "loss": 0.5221,
+      "step": 76870
+    },
+    {
+      "epoch": 3.8184166087215656,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0004945306446806397,
+      "loss": 0.5134,
+      "step": 76880
+    },
+    {
+      "epoch": 3.818913281017185,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0004944909108969902,
+      "loss": 0.5067,
+      "step": 76890
+    },
+    {
+      "epoch": 3.819409953312804,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0004944511771133406,
+      "loss": 0.5321,
+      "step": 76900
+    },
+    {
+      "epoch": 3.8199066256084233,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0004944114433296911,
+      "loss": 0.5426,
+      "step": 76910
+    },
+    {
+      "epoch": 3.820403297904043,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0004943717095460416,
+      "loss": 0.515,
+      "step": 76920
+    },
+    {
+      "epoch": 3.820899970199662,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.000494331975762392,
+      "loss": 0.5162,
+      "step": 76930
+    },
+    {
+      "epoch": 3.821396642495282,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0004942922419787425,
+      "loss": 0.5413,
+      "step": 76940
+    },
+    {
+      "epoch": 3.821893314790901,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0004942525081950929,
+      "loss": 0.5547,
+      "step": 76950
+    },
+    {
+      "epoch": 3.8223899870865203,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0004942127744114433,
+      "loss": 0.5419,
+      "step": 76960
+    },
+    {
+      "epoch": 3.8228866593821396,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004941730406277939,
+      "loss": 0.5224,
+      "step": 76970
+    },
+    {
+      "epoch": 3.823383331677759,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0004941333068441442,
+      "loss": 0.5259,
+      "step": 76980
+    },
+    {
+      "epoch": 3.8238800039733785,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0004940935730604947,
+      "loss": 0.5216,
+      "step": 76990
+    },
+    {
+      "epoch": 3.8243766762689977,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004940538392768452,
+      "loss": 0.5639,
+      "step": 77000
+    },
+    {
+      "epoch": 3.824873348564617,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004940141054931955,
+      "loss": 0.5427,
+      "step": 77010
+    },
+    {
+      "epoch": 3.8253700208602366,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0004939743717095461,
+      "loss": 0.5237,
+      "step": 77020
+    },
+    {
+      "epoch": 3.825866693155856,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004939346379258966,
+      "loss": 0.5508,
+      "step": 77030
+    },
+    {
+      "epoch": 3.826363365451475,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0004938949041422469,
+      "loss": 0.5257,
+      "step": 77040
+    },
+    {
+      "epoch": 3.8268600377470943,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0004938551703585974,
+      "loss": 0.5381,
+      "step": 77050
+    },
+    {
+      "epoch": 3.827356710042714,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0004938154365749478,
+      "loss": 0.5421,
+      "step": 77060
+    },
+    {
+      "epoch": 3.827853382338333,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0004937757027912984,
+      "loss": 0.525,
+      "step": 77070
+    },
+    {
+      "epoch": 3.8283500546339524,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0004937359690076488,
+      "loss": 0.5108,
+      "step": 77080
+    },
+    {
+      "epoch": 3.828846726929572,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0004936962352239992,
+      "loss": 0.5251,
+      "step": 77090
+    },
+    {
+      "epoch": 3.8293433992251913,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0004936565014403497,
+      "loss": 0.5566,
+      "step": 77100
+    },
+    {
+      "epoch": 3.8298400715208105,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004936167676567001,
+      "loss": 0.5185,
+      "step": 77110
+    },
+    {
+      "epoch": 3.8303367438164297,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0004935770338730506,
+      "loss": 0.5233,
+      "step": 77120
+    },
+    {
+      "epoch": 3.8308334161120494,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0004935373000894011,
+      "loss": 0.5356,
+      "step": 77130
+    },
+    {
+      "epoch": 3.8313300884076686,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004934975663057515,
+      "loss": 0.5334,
+      "step": 77140
+    },
+    {
+      "epoch": 3.831826760703288,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0004934578325221019,
+      "loss": 0.5327,
+      "step": 77150
+    },
+    {
+      "epoch": 3.8323234329989075,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0004934180987384524,
+      "loss": 0.5342,
+      "step": 77160
+    },
+    {
+      "epoch": 3.8328201052945268,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0004933783649548029,
+      "loss": 0.5361,
+      "step": 77170
+    },
+    {
+      "epoch": 3.833316777590146,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0004933386311711533,
+      "loss": 0.5545,
+      "step": 77180
+    },
+    {
+      "epoch": 3.8338134498857652,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0004932988973875038,
+      "loss": 0.5256,
+      "step": 77190
+    },
+    {
+      "epoch": 3.8343101221813845,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0004932591636038541,
+      "loss": 0.5355,
+      "step": 77200
+    },
+    {
+      "epoch": 3.834806794477004,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0004932194298202046,
+      "loss": 0.5366,
+      "step": 77210
+    },
+    {
+      "epoch": 3.8353034667726233,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0004931796960365552,
+      "loss": 0.5469,
+      "step": 77220
+    },
+    {
+      "epoch": 3.835800139068243,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0004931399622529056,
+      "loss": 0.5382,
+      "step": 77230
+    },
+    {
+      "epoch": 3.8362968113638622,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.000493100228469256,
+      "loss": 0.5042,
+      "step": 77240
+    },
+    {
+      "epoch": 3.8367934836594815,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0004930604946856064,
+      "loss": 0.5104,
+      "step": 77250
+    },
+    {
+      "epoch": 3.8372901559551007,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0004930207609019569,
+      "loss": 0.5509,
+      "step": 77260
+    },
+    {
+      "epoch": 3.83778682825072,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004929810271183074,
+      "loss": 0.5261,
+      "step": 77270
+    },
+    {
+      "epoch": 3.8382835005463396,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0004929412933346578,
+      "loss": 0.5002,
+      "step": 77280
+    },
+    {
+      "epoch": 3.838780172841959,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0004929015595510083,
+      "loss": 0.5322,
+      "step": 77290
+    },
+    {
+      "epoch": 3.8392768451375785,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0004928618257673587,
+      "loss": 0.5219,
+      "step": 77300
+    },
+    {
+      "epoch": 3.8397735174331977,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0004928220919837091,
+      "loss": 0.5465,
+      "step": 77310
+    },
+    {
+      "epoch": 3.840270189728817,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0004927823582000597,
+      "loss": 0.5752,
+      "step": 77320
+    },
+    {
+      "epoch": 3.840766862024436,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0004927426244164101,
+      "loss": 0.5246,
+      "step": 77330
+    },
+    {
+      "epoch": 3.8412635343200554,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0004927028906327605,
+      "loss": 0.5423,
+      "step": 77340
+    },
+    {
+      "epoch": 3.841760206615675,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.000492663156849111,
+      "loss": 0.5225,
+      "step": 77350
+    },
+    {
+      "epoch": 3.8422568789112943,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0004926234230654614,
+      "loss": 0.5334,
+      "step": 77360
+    },
+    {
+      "epoch": 3.8427535512069135,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0004925836892818119,
+      "loss": 0.505,
+      "step": 77370
+    },
+    {
+      "epoch": 3.843250223502533,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0004925439554981624,
+      "loss": 0.5224,
+      "step": 77380
+    },
+    {
+      "epoch": 3.8437468957981524,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004925042217145128,
+      "loss": 0.5149,
+      "step": 77390
+    },
+    {
+      "epoch": 3.8442435680937717,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004924644879308632,
+      "loss": 0.5327,
+      "step": 77400
+    },
+    {
+      "epoch": 3.844740240389391,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0004924247541472137,
+      "loss": 0.5212,
+      "step": 77410
+    },
+    {
+      "epoch": 3.8452369126850106,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0004923850203635642,
+      "loss": 0.5236,
+      "step": 77420
+    },
+    {
+      "epoch": 3.84573358498063,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0004923452865799146,
+      "loss": 0.5518,
+      "step": 77430
+    },
+    {
+      "epoch": 3.846230257276249,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004923055527962651,
+      "loss": 0.5539,
+      "step": 77440
+    },
+    {
+      "epoch": 3.8467269295718687,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004922658190126155,
+      "loss": 0.5373,
+      "step": 77450
+    },
+    {
+      "epoch": 3.847223601867488,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0004922260852289659,
+      "loss": 0.5529,
+      "step": 77460
+    },
+    {
+      "epoch": 3.847720274163107,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0004921863514453164,
+      "loss": 0.5682,
+      "step": 77470
+    },
+    {
+      "epoch": 3.8482169464587264,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0004921466176616669,
+      "loss": 0.5279,
+      "step": 77480
+    },
+    {
+      "epoch": 3.848713618754346,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004921068838780173,
+      "loss": 0.5535,
+      "step": 77490
+    },
+    {
+      "epoch": 3.8492102910499653,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004920671500943677,
+      "loss": 0.5249,
+      "step": 77500
+    },
+    {
+      "epoch": 3.8497069633455845,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0004920274163107182,
+      "loss": 0.5058,
+      "step": 77510
+    },
+    {
+      "epoch": 3.850203635641204,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0004919876825270687,
+      "loss": 0.5522,
+      "step": 77520
+    },
+    {
+      "epoch": 3.8507003079368234,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0004919479487434191,
+      "loss": 0.5331,
+      "step": 77530
+    },
+    {
+      "epoch": 3.8511969802324426,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0004919082149597696,
+      "loss": 0.5452,
+      "step": 77540
+    },
+    {
+      "epoch": 3.851693652528062,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.00049186848117612,
+      "loss": 0.5229,
+      "step": 77550
+    },
+    {
+      "epoch": 3.852190324823681,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0004918287473924704,
+      "loss": 0.5111,
+      "step": 77560
+    },
+    {
+      "epoch": 3.8526869971193007,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.000491789013608821,
+      "loss": 0.5245,
+      "step": 77570
+    },
+    {
+      "epoch": 3.85318366941492,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0004917492798251714,
+      "loss": 0.5325,
+      "step": 77580
+    },
+    {
+      "epoch": 3.8536803417105396,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0004917095460415218,
+      "loss": 0.5324,
+      "step": 77590
+    },
+    {
+      "epoch": 3.854177014006159,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0004916698122578723,
+      "loss": 0.523,
+      "step": 77600
+    },
+    {
+      "epoch": 3.854673686301778,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0004916300784742227,
+      "loss": 0.5764,
+      "step": 77610
+    },
+    {
+      "epoch": 3.8551703585973973,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0004915903446905732,
+      "loss": 0.5273,
+      "step": 77620
+    },
+    {
+      "epoch": 3.8556670308930165,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004915506109069237,
+      "loss": 0.5281,
+      "step": 77630
+    },
+    {
+      "epoch": 3.856163703188636,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0004915108771232741,
+      "loss": 0.5493,
+      "step": 77640
+    },
+    {
+      "epoch": 3.8566603754842554,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0004914711433396245,
+      "loss": 0.53,
+      "step": 77650
+    },
+    {
+      "epoch": 3.857157047779875,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0004914314095559749,
+      "loss": 0.5514,
+      "step": 77660
+    },
+    {
+      "epoch": 3.8576537200754943,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0004913916757723255,
+      "loss": 0.5638,
+      "step": 77670
+    },
+    {
+      "epoch": 3.8581503923711136,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.000491351941988676,
+      "loss": 0.5153,
+      "step": 77680
+    },
+    {
+      "epoch": 3.858647064666733,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0004913122082050263,
+      "loss": 0.5267,
+      "step": 77690
+    },
+    {
+      "epoch": 3.859143736962352,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004912724744213768,
+      "loss": 0.563,
+      "step": 77700
+    },
+    {
+      "epoch": 3.8596404092579717,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004912327406377273,
+      "loss": 0.5209,
+      "step": 77710
+    },
+    {
+      "epoch": 3.860137081553591,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0004911930068540776,
+      "loss": 0.5381,
+      "step": 77720
+    },
+    {
+      "epoch": 3.86063375384921,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0004911532730704282,
+      "loss": 0.5184,
+      "step": 77730
+    },
+    {
+      "epoch": 3.86113042614483,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0004911135392867786,
+      "loss": 0.5145,
+      "step": 77740
+    },
+    {
+      "epoch": 3.861627098440449,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.000491073805503129,
+      "loss": 0.5214,
+      "step": 77750
+    },
+    {
+      "epoch": 3.8621237707360683,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0004910340717194795,
+      "loss": 0.5344,
+      "step": 77760
+    },
+    {
+      "epoch": 3.8626204430316875,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00049099433793583,
+      "loss": 0.5234,
+      "step": 77770
+    },
+    {
+      "epoch": 3.863117115327307,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0004909546041521804,
+      "loss": 0.5339,
+      "step": 77780
+    },
+    {
+      "epoch": 3.8636137876229264,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0004909148703685309,
+      "loss": 0.5405,
+      "step": 77790
+    },
+    {
+      "epoch": 3.8641104599185456,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0004908751365848813,
+      "loss": 0.5536,
+      "step": 77800
+    },
+    {
+      "epoch": 3.8646071322141653,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004908354028012318,
+      "loss": 0.5374,
+      "step": 77810
+    },
+    {
+      "epoch": 3.8651038045097845,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0004907956690175823,
+      "loss": 0.5438,
+      "step": 77820
+    },
+    {
+      "epoch": 3.8656004768054038,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0004907559352339327,
+      "loss": 0.5263,
+      "step": 77830
+    },
+    {
+      "epoch": 3.866097149101023,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0004907162014502832,
+      "loss": 0.5228,
+      "step": 77840
+    },
+    {
+      "epoch": 3.8665938213966426,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0004906764676666335,
+      "loss": 0.5644,
+      "step": 77850
+    },
+    {
+      "epoch": 3.867090493692262,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.000490636733882984,
+      "loss": 0.5252,
+      "step": 77860
+    },
+    {
+      "epoch": 3.867587165987881,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0004905970000993346,
+      "loss": 0.5538,
+      "step": 77870
+    },
+    {
+      "epoch": 3.8680838382835008,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0004905572663156849,
+      "loss": 0.5115,
+      "step": 77880
+    },
+    {
+      "epoch": 3.86858051057912,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0004905175325320354,
+      "loss": 0.5239,
+      "step": 77890
+    },
+    {
+      "epoch": 3.8690771828747392,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004904777987483859,
+      "loss": 0.4906,
+      "step": 77900
+    },
+    {
+      "epoch": 3.8695738551703585,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0004904380649647362,
+      "loss": 0.551,
+      "step": 77910
+    },
+    {
+      "epoch": 3.8700705274659777,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0004903983311810867,
+      "loss": 0.5334,
+      "step": 77920
+    },
+    {
+      "epoch": 3.8705671997615974,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0004903585973974372,
+      "loss": 0.5479,
+      "step": 77930
+    },
+    {
+      "epoch": 3.8710638720572166,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0004903188636137876,
+      "loss": 0.5058,
+      "step": 77940
+    },
+    {
+      "epoch": 3.8715605443528363,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0004902791298301381,
+      "loss": 0.54,
+      "step": 77950
+    },
+    {
+      "epoch": 3.8720572166484555,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004902393960464885,
+      "loss": 0.533,
+      "step": 77960
+    },
+    {
+      "epoch": 3.8725538889440747,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.000490199662262839,
+      "loss": 0.5098,
+      "step": 77970
+    },
+    {
+      "epoch": 3.873050561239694,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004901599284791895,
+      "loss": 0.5435,
+      "step": 77980
+    },
+    {
+      "epoch": 3.873547233535313,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004901201946955399,
+      "loss": 0.5652,
+      "step": 77990
+    },
+    {
+      "epoch": 3.874043905830933,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004900804609118904,
+      "loss": 0.5213,
+      "step": 78000
+    },
+    {
+      "epoch": 3.874540578126552,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0004900407271282408,
+      "loss": 0.5487,
+      "step": 78010
+    },
+    {
+      "epoch": 3.8750372504221713,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0004900009933445912,
+      "loss": 0.5225,
+      "step": 78020
+    },
+    {
+      "epoch": 3.875533922717791,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0004899612595609418,
+      "loss": 0.5417,
+      "step": 78030
+    },
+    {
+      "epoch": 3.87603059501341,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0004899215257772922,
+      "loss": 0.5715,
+      "step": 78040
+    },
+    {
+      "epoch": 3.8765272673090294,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0004898817919936426,
+      "loss": 0.5074,
+      "step": 78050
+    },
+    {
+      "epoch": 3.8770239396046486,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004898420582099931,
+      "loss": 0.5138,
+      "step": 78060
+    },
+    {
+      "epoch": 3.8775206119002683,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0004898023244263434,
+      "loss": 0.513,
+      "step": 78070
+    },
+    {
+      "epoch": 3.8780172841958875,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.000489762590642694,
+      "loss": 0.5218,
+      "step": 78080
+    },
+    {
+      "epoch": 3.8785139564915068,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0004897228568590445,
+      "loss": 0.5516,
+      "step": 78090
+    },
+    {
+      "epoch": 3.8790106287871264,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0004896831230753948,
+      "loss": 0.5456,
+      "step": 78100
+    },
+    {
+      "epoch": 3.8795073010827457,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004896433892917453,
+      "loss": 0.5356,
+      "step": 78110
+    },
+    {
+      "epoch": 3.880003973378365,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004896036555080957,
+      "loss": 0.5203,
+      "step": 78120
+    },
+    {
+      "epoch": 3.880500645673984,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004895639217244463,
+      "loss": 0.5382,
+      "step": 78130
+    },
+    {
+      "epoch": 3.880997317969604,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0004895241879407967,
+      "loss": 0.5274,
+      "step": 78140
+    },
+    {
+      "epoch": 3.881493990265223,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004894844541571471,
+      "loss": 0.5278,
+      "step": 78150
+    },
+    {
+      "epoch": 3.8819906625608422,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0004894447203734976,
+      "loss": 0.5564,
+      "step": 78160
+    },
+    {
+      "epoch": 3.882487334856462,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000489404986589848,
+      "loss": 0.5444,
+      "step": 78170
+    },
+    {
+      "epoch": 3.882984007152081,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0004893652528061985,
+      "loss": 0.5267,
+      "step": 78180
+    },
+    {
+      "epoch": 3.8834806794477004,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.000489325519022549,
+      "loss": 0.5116,
+      "step": 78190
+    },
+    {
+      "epoch": 3.8839773517433196,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0004892857852388994,
+      "loss": 0.5024,
+      "step": 78200
+    },
+    {
+      "epoch": 3.8844740240389393,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0004892460514552498,
+      "loss": 0.5304,
+      "step": 78210
+    },
+    {
+      "epoch": 3.8849706963345585,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004892063176716003,
+      "loss": 0.5474,
+      "step": 78220
+    },
+    {
+      "epoch": 3.8854673686301777,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0004891665838879508,
+      "loss": 0.5243,
+      "step": 78230
+    },
+    {
+      "epoch": 3.8859640409257974,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004891268501043012,
+      "loss": 0.5531,
+      "step": 78240
+    },
+    {
+      "epoch": 3.8864607132214166,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0004890871163206517,
+      "loss": 0.5401,
+      "step": 78250
+    },
+    {
+      "epoch": 3.886957385517036,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0004890473825370021,
+      "loss": 0.5397,
+      "step": 78260
+    },
+    {
+      "epoch": 3.887454057812655,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0004890076487533525,
+      "loss": 0.5461,
+      "step": 78270
+    },
+    {
+      "epoch": 3.8879507301082743,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0004889679149697031,
+      "loss": 0.5115,
+      "step": 78280
+    },
+    {
+      "epoch": 3.888447402403894,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0004889281811860535,
+      "loss": 0.519,
+      "step": 78290
+    },
+    {
+      "epoch": 3.888944074699513,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0004888884474024039,
+      "loss": 0.5557,
+      "step": 78300
+    },
+    {
+      "epoch": 3.889440746995133,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004888487136187543,
+      "loss": 0.5311,
+      "step": 78310
+    },
+    {
+      "epoch": 3.889937419290752,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0004888089798351048,
+      "loss": 0.5394,
+      "step": 78320
+    },
+    {
+      "epoch": 3.8904340915863713,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004887692460514553,
+      "loss": 0.5154,
+      "step": 78330
+    },
+    {
+      "epoch": 3.8909307638819906,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0004887295122678057,
+      "loss": 0.5444,
+      "step": 78340
+    },
+    {
+      "epoch": 3.8914274361776098,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0004886897784841562,
+      "loss": 0.5237,
+      "step": 78350
+    },
+    {
+      "epoch": 3.8919241084732294,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004886500447005066,
+      "loss": 0.5175,
+      "step": 78360
+    },
+    {
+      "epoch": 3.8924207807688487,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.000488610310916857,
+      "loss": 0.5257,
+      "step": 78370
+    },
+    {
+      "epoch": 3.892917453064468,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0004885705771332076,
+      "loss": 0.5148,
+      "step": 78380
+    },
+    {
+      "epoch": 3.8934141253600876,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.000488530843349558,
+      "loss": 0.5445,
+      "step": 78390
+    },
+    {
+      "epoch": 3.893910797655707,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0004884911095659084,
+      "loss": 0.5255,
+      "step": 78400
+    },
+    {
+      "epoch": 3.894407469951326,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0004884513757822589,
+      "loss": 0.538,
+      "step": 78410
+    },
+    {
+      "epoch": 3.8949041422469453,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0004884116419986093,
+      "loss": 0.5602,
+      "step": 78420
+    },
+    {
+      "epoch": 3.895400814542565,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004883719082149598,
+      "loss": 0.5426,
+      "step": 78430
+    },
+    {
+      "epoch": 3.895897486838184,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0004883321744313103,
+      "loss": 0.5264,
+      "step": 78440
+    },
+    {
+      "epoch": 3.8963941591338034,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0004882924406476607,
+      "loss": 0.5425,
+      "step": 78450
+    },
+    {
+      "epoch": 3.896890831429423,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0004882527068640111,
+      "loss": 0.5515,
+      "step": 78460
+    },
+    {
+      "epoch": 3.8973875037250423,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0004882129730803616,
+      "loss": 0.5495,
+      "step": 78470
+    },
+    {
+      "epoch": 3.8978841760206615,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00048817323929671206,
+      "loss": 0.5271,
+      "step": 78480
+    },
+    {
+      "epoch": 3.8983808483162807,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0004881335055130625,
+      "loss": 0.5404,
+      "step": 78490
+    },
+    {
+      "epoch": 3.8988775206119004,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00048809377172941295,
+      "loss": 0.5227,
+      "step": 78500
+    },
+    {
+      "epoch": 3.8993741929075196,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0004880540379457634,
+      "loss": 0.5285,
+      "step": 78510
+    },
+    {
+      "epoch": 3.899870865203139,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.00048801430416211384,
+      "loss": 0.5331,
+      "step": 78520
+    },
+    {
+      "epoch": 3.9003675374987585,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0004879745703784643,
+      "loss": 0.5292,
+      "step": 78530
+    },
+    {
+      "epoch": 3.9008642097943778,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004879348365948148,
+      "loss": 0.4952,
+      "step": 78540
+    },
+    {
+      "epoch": 3.901360882089997,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0004878951028111652,
+      "loss": 0.5172,
+      "step": 78550
+    },
+    {
+      "epoch": 3.901857554385616,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00048785536902751567,
+      "loss": 0.5235,
+      "step": 78560
+    },
+    {
+      "epoch": 3.902354226681236,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00048781563524386614,
+      "loss": 0.5048,
+      "step": 78570
+    },
+    {
+      "epoch": 3.902850898976855,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0004877759014602166,
+      "loss": 0.5437,
+      "step": 78580
+    },
+    {
+      "epoch": 3.9033475712724743,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00048773616767656703,
+      "loss": 0.5178,
+      "step": 78590
+    },
+    {
+      "epoch": 3.903844243568094,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0004876964338929175,
+      "loss": 0.5071,
+      "step": 78600
+    },
+    {
+      "epoch": 3.9043409158637132,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.00048765670010926797,
+      "loss": 0.5203,
+      "step": 78610
+    },
+    {
+      "epoch": 3.9048375881593325,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00048761696632561833,
+      "loss": 0.5194,
+      "step": 78620
+    },
+    {
+      "epoch": 3.9053342604549517,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004875772325419688,
+      "loss": 0.515,
+      "step": 78630
+    },
+    {
+      "epoch": 3.905830932750571,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00048753749875831933,
+      "loss": 0.533,
+      "step": 78640
+    },
+    {
+      "epoch": 3.9063276050461906,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0004874977649746697,
+      "loss": 0.5588,
+      "step": 78650
+    },
+    {
+      "epoch": 3.90682427734181,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00048745803119102017,
+      "loss": 0.5296,
+      "step": 78660
+    },
+    {
+      "epoch": 3.9073209496374295,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00048741829740737064,
+      "loss": 0.5109,
+      "step": 78670
+    },
+    {
+      "epoch": 3.9078176219330487,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00048737856362372105,
+      "loss": 0.5399,
+      "step": 78680
+    },
+    {
+      "epoch": 3.908314294228668,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0004873388298400715,
+      "loss": 0.509,
+      "step": 78690
+    },
+    {
+      "epoch": 3.908810966524287,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.000487299096056422,
+      "loss": 0.52,
+      "step": 78700
+    },
+    {
+      "epoch": 3.9093076388199064,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00048725936227277247,
+      "loss": 0.5361,
+      "step": 78710
+    },
+    {
+      "epoch": 3.909804311115526,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0004872196284891229,
+      "loss": 0.5012,
+      "step": 78720
+    },
+    {
+      "epoch": 3.9103009834111453,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00048717989470547336,
+      "loss": 0.5238,
+      "step": 78730
+    },
+    {
+      "epoch": 3.9107976557067645,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00048714016092182383,
+      "loss": 0.5273,
+      "step": 78740
+    },
+    {
+      "epoch": 3.911294328002384,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00048710042713817424,
+      "loss": 0.532,
+      "step": 78750
+    },
+    {
+      "epoch": 3.9117910002980034,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004870606933545247,
+      "loss": 0.5236,
+      "step": 78760
+    },
+    {
+      "epoch": 3.9122876725936226,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0004870209595708752,
+      "loss": 0.5262,
+      "step": 78770
+    },
+    {
+      "epoch": 3.912784344889242,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0004869812257872256,
+      "loss": 0.5073,
+      "step": 78780
+    },
+    {
+      "epoch": 3.9132810171848615,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004869414920035761,
+      "loss": 0.5161,
+      "step": 78790
+    },
+    {
+      "epoch": 3.9137776894804808,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.00048690175821992655,
+      "loss": 0.5226,
+      "step": 78800
+    },
+    {
+      "epoch": 3.9142743617761,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0004868620244362769,
+      "loss": 0.5043,
+      "step": 78810
+    },
+    {
+      "epoch": 3.9147710340717197,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00048682229065262744,
+      "loss": 0.5119,
+      "step": 78820
+    },
+    {
+      "epoch": 3.915267706367339,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0004867825568689779,
+      "loss": 0.5354,
+      "step": 78830
+    },
+    {
+      "epoch": 3.915764378662958,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.00048674282308532827,
+      "loss": 0.5004,
+      "step": 78840
+    },
+    {
+      "epoch": 3.9162610509585773,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00048670308930167874,
+      "loss": 0.5131,
+      "step": 78850
+    },
+    {
+      "epoch": 3.916757723254197,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.00048666335551802927,
+      "loss": 0.5342,
+      "step": 78860
+    },
+    {
+      "epoch": 3.9172543955498162,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00048662362173437974,
+      "loss": 0.5517,
+      "step": 78870
+    },
+    {
+      "epoch": 3.9177510678454355,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0004865838879507301,
+      "loss": 0.5119,
+      "step": 78880
+    },
+    {
+      "epoch": 3.918247740141055,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.00048654415416708057,
+      "loss": 0.5285,
+      "step": 78890
+    },
+    {
+      "epoch": 3.9187444124366744,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00048650442038343104,
+      "loss": 0.5466,
+      "step": 78900
+    },
+    {
+      "epoch": 3.9192410847322936,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.00048646468659978146,
+      "loss": 0.5246,
+      "step": 78910
+    },
+    {
+      "epoch": 3.919737757027913,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00048642495281613193,
+      "loss": 0.5565,
+      "step": 78920
+    },
+    {
+      "epoch": 3.9202344293235325,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0004863852190324824,
+      "loss": 0.5123,
+      "step": 78930
+    },
+    {
+      "epoch": 3.9207311016191517,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004863454852488328,
+      "loss": 0.522,
+      "step": 78940
+    },
+    {
+      "epoch": 3.921227773914771,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0004863057514651833,
+      "loss": 0.5093,
+      "step": 78950
+    },
+    {
+      "epoch": 3.9217244462103906,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00048626601768153376,
+      "loss": 0.511,
+      "step": 78960
+    },
+    {
+      "epoch": 3.92222111850601,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004862262838978842,
+      "loss": 0.5285,
+      "step": 78970
+    },
+    {
+      "epoch": 3.922717790801629,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00048618655011423465,
+      "loss": 0.5343,
+      "step": 78980
+    },
+    {
+      "epoch": 3.9232144630972483,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0004861468163305851,
+      "loss": 0.5646,
+      "step": 78990
+    },
+    {
+      "epoch": 3.9237111353928675,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0004861070825469356,
+      "loss": 0.5277,
+      "step": 79000
+    },
+    {
+      "epoch": 3.924207807688487,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.000486067348763286,
+      "loss": 0.5393,
+      "step": 79010
+    },
+    {
+      "epoch": 3.9247044799841064,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0004860276149796365,
+      "loss": 0.5405,
+      "step": 79020
+    },
+    {
+      "epoch": 3.925201152279726,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.00048598788119598696,
+      "loss": 0.5444,
+      "step": 79030
+    },
+    {
+      "epoch": 3.9256978245753453,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0004859481474123373,
+      "loss": 0.5462,
+      "step": 79040
+    },
+    {
+      "epoch": 3.9261944968709646,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.00048590841362868784,
+      "loss": 0.5064,
+      "step": 79050
+    },
+    {
+      "epoch": 3.926691169166584,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0004858686798450383,
+      "loss": 0.519,
+      "step": 79060
+    },
+    {
+      "epoch": 3.927187841462203,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0004858289460613887,
+      "loss": 0.5443,
+      "step": 79070
+    },
+    {
+      "epoch": 3.9276845137578227,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00048578921227773915,
+      "loss": 0.511,
+      "step": 79080
+    },
+    {
+      "epoch": 3.928181186053442,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004857494784940897,
+      "loss": 0.5439,
+      "step": 79090
+    },
+    {
+      "epoch": 3.928677858349061,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.00048570974471044004,
+      "loss": 0.5368,
+      "step": 79100
+    },
+    {
+      "epoch": 3.929174530644681,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0004856700109267905,
+      "loss": 0.5425,
+      "step": 79110
+    },
+    {
+      "epoch": 3.9296712029403,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.000485630277143141,
+      "loss": 0.5535,
+      "step": 79120
+    },
+    {
+      "epoch": 3.9301678752359193,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0004855905433594914,
+      "loss": 0.5268,
+      "step": 79130
+    },
+    {
+      "epoch": 3.9306645475315385,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.00048555080957584187,
+      "loss": 0.523,
+      "step": 79140
+    },
+    {
+      "epoch": 3.931161219827158,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00048551107579219234,
+      "loss": 0.537,
+      "step": 79150
+    },
+    {
+      "epoch": 3.9316578921227774,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0004854713420085428,
+      "loss": 0.5496,
+      "step": 79160
+    },
+    {
+      "epoch": 3.9321545644183966,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00048543160822489323,
+      "loss": 0.5504,
+      "step": 79170
+    },
+    {
+      "epoch": 3.9326512367140163,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0004853918744412437,
+      "loss": 0.5735,
+      "step": 79180
+    },
+    {
+      "epoch": 3.9331479090096355,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00048535214065759417,
+      "loss": 0.5313,
+      "step": 79190
+    },
+    {
+      "epoch": 3.9336445813052547,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0004853124068739446,
+      "loss": 0.5333,
+      "step": 79200
+    },
+    {
+      "epoch": 3.934141253600874,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00048527267309029506,
+      "loss": 0.5218,
+      "step": 79210
+    },
+    {
+      "epoch": 3.9346379258964936,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00048523293930664553,
+      "loss": 0.5266,
+      "step": 79220
+    },
+    {
+      "epoch": 3.935134598192113,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0004851932055229959,
+      "loss": 0.5356,
+      "step": 79230
+    },
+    {
+      "epoch": 3.935631270487732,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0004851534717393464,
+      "loss": 0.5371,
+      "step": 79240
+    },
+    {
+      "epoch": 3.9361279427833518,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0004851137379556969,
+      "loss": 0.509,
+      "step": 79250
+    },
+    {
+      "epoch": 3.936624615078971,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00048507400417204725,
+      "loss": 0.5251,
+      "step": 79260
+    },
+    {
+      "epoch": 3.93712128737459,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0004850342703883977,
+      "loss": 0.5085,
+      "step": 79270
+    },
+    {
+      "epoch": 3.9376179596702094,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00048499453660474825,
+      "loss": 0.5571,
+      "step": 79280
+    },
+    {
+      "epoch": 3.9381146319658287,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0004849548028210986,
+      "loss": 0.5364,
+      "step": 79290
+    },
+    {
+      "epoch": 3.9386113042614483,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0004849150690374491,
+      "loss": 0.5451,
+      "step": 79300
+    },
+    {
+      "epoch": 3.9391079765570676,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.00048487533525379956,
+      "loss": 0.5003,
+      "step": 79310
+    },
+    {
+      "epoch": 3.9396046488526872,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0004848356014701501,
+      "loss": 0.5263,
+      "step": 79320
+    },
+    {
+      "epoch": 3.9401013211483065,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00048479586768650045,
+      "loss": 0.523,
+      "step": 79330
+    },
+    {
+      "epoch": 3.9405979934439257,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004847561339028509,
+      "loss": 0.5346,
+      "step": 79340
+    },
+    {
+      "epoch": 3.941094665739545,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0004847164001192014,
+      "loss": 0.5233,
+      "step": 79350
+    },
+    {
+      "epoch": 3.941591338035164,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0004846766663355518,
+      "loss": 0.5381,
+      "step": 79360
+    },
+    {
+      "epoch": 3.942088010330784,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0004846369325519023,
+      "loss": 0.5275,
+      "step": 79370
+    },
+    {
+      "epoch": 3.942584682626403,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00048459719876825275,
+      "loss": 0.5336,
+      "step": 79380
+    },
+    {
+      "epoch": 3.9430813549220227,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00048455746498460316,
+      "loss": 0.5401,
+      "step": 79390
+    },
+    {
+      "epoch": 3.943578027217642,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00048451773120095364,
+      "loss": 0.524,
+      "step": 79400
+    },
+    {
+      "epoch": 3.944074699513261,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0004844779974173041,
+      "loss": 0.514,
+      "step": 79410
+    },
+    {
+      "epoch": 3.9445713718088804,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004844382636336545,
+      "loss": 0.5181,
+      "step": 79420
+    },
+    {
+      "epoch": 3.9450680441044996,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.000484398529850005,
+      "loss": 0.5247,
+      "step": 79430
+    },
+    {
+      "epoch": 3.9455647164001193,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00048435879606635547,
+      "loss": 0.5204,
+      "step": 79440
+    },
+    {
+      "epoch": 3.9460613886957385,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.00048431906228270594,
+      "loss": 0.4897,
+      "step": 79450
+    },
+    {
+      "epoch": 3.9465580609913578,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0004842793284990563,
+      "loss": 0.5227,
+      "step": 79460
+    },
+    {
+      "epoch": 3.9470547332869774,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00048423959471540683,
+      "loss": 0.5644,
+      "step": 79470
+    },
+    {
+      "epoch": 3.9475514055825967,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004841998609317573,
+      "loss": 0.5391,
+      "step": 79480
+    },
+    {
+      "epoch": 3.948048077878216,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00048416012714810766,
+      "loss": 0.5092,
+      "step": 79490
+    },
+    {
+      "epoch": 3.948544750173835,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00048412039336445813,
+      "loss": 0.5233,
+      "step": 79500
+    },
+    {
+      "epoch": 3.9490414224694548,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00048408065958080866,
+      "loss": 0.5487,
+      "step": 79510
+    },
+    {
+      "epoch": 3.949538094765074,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.000484040925797159,
+      "loss": 0.5287,
+      "step": 79520
+    },
+    {
+      "epoch": 3.9500347670606932,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0004840011920135095,
+      "loss": 0.5458,
+      "step": 79530
+    },
+    {
+      "epoch": 3.950531439356313,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00048396145822985996,
+      "loss": 0.522,
+      "step": 79540
+    },
+    {
+      "epoch": 3.951028111651932,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0004839217244462104,
+      "loss": 0.5134,
+      "step": 79550
+    },
+    {
+      "epoch": 3.9515247839475514,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00048388199066256085,
+      "loss": 0.5682,
+      "step": 79560
+    },
+    {
+      "epoch": 3.9520214562431706,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0004838422568789113,
+      "loss": 0.4989,
+      "step": 79570
+    },
+    {
+      "epoch": 3.9525181285387903,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00048380252309526174,
+      "loss": 0.5383,
+      "step": 79580
+    },
+    {
+      "epoch": 3.9530148008344095,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0004837627893116122,
+      "loss": 0.5253,
+      "step": 79590
+    },
+    {
+      "epoch": 3.9535114731300287,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0004837230555279627,
+      "loss": 0.5204,
+      "step": 79600
+    },
+    {
+      "epoch": 3.9540081454256484,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00048368332174431316,
+      "loss": 0.5299,
+      "step": 79610
+    },
+    {
+      "epoch": 3.9545048177212676,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00048364358796066357,
+      "loss": 0.5158,
+      "step": 79620
+    },
+    {
+      "epoch": 3.955001490016887,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00048360385417701404,
+      "loss": 0.5451,
+      "step": 79630
+    },
+    {
+      "epoch": 3.955498162312506,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0004835641203933645,
+      "loss": 0.5338,
+      "step": 79640
+    },
+    {
+      "epoch": 3.9559948346081253,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00048352438660971493,
+      "loss": 0.5416,
+      "step": 79650
+    },
+    {
+      "epoch": 3.956491506903745,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0004834846528260654,
+      "loss": 0.5345,
+      "step": 79660
+    },
+    {
+      "epoch": 3.956988179199364,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004834449190424159,
+      "loss": 0.549,
+      "step": 79670
+    },
+    {
+      "epoch": 3.957484851494984,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00048340518525876624,
+      "loss": 0.5493,
+      "step": 79680
+    },
+    {
+      "epoch": 3.957981523790603,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00048336545147511676,
+      "loss": 0.5484,
+      "step": 79690
+    },
+    {
+      "epoch": 3.9584781960862223,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00048332571769146723,
+      "loss": 0.5111,
+      "step": 79700
+    },
+    {
+      "epoch": 3.9589748683818415,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004832859839078176,
+      "loss": 0.5525,
+      "step": 79710
+    },
+    {
+      "epoch": 3.9594715406774608,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00048324625012416807,
+      "loss": 0.516,
+      "step": 79720
+    },
+    {
+      "epoch": 3.9599682129730804,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00048320651634051854,
+      "loss": 0.5382,
+      "step": 79730
+    },
+    {
+      "epoch": 3.9604648852686997,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00048316678255686896,
+      "loss": 0.5643,
+      "step": 79740
+    },
+    {
+      "epoch": 3.9609615575643193,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00048312704877321943,
+      "loss": 0.5073,
+      "step": 79750
+    },
+    {
+      "epoch": 3.9614582298599386,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0004830873149895699,
+      "loss": 0.5077,
+      "step": 79760
+    },
+    {
+      "epoch": 3.961954902155558,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00048304758120592037,
+      "loss": 0.5649,
+      "step": 79770
+    },
+    {
+      "epoch": 3.962451574451177,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0004830078474222708,
+      "loss": 0.5168,
+      "step": 79780
+    },
+    {
+      "epoch": 3.9629482467467962,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.00048296811363862126,
+      "loss": 0.5375,
+      "step": 79790
+    },
+    {
+      "epoch": 3.963444919042416,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00048292837985497173,
+      "loss": 0.5512,
+      "step": 79800
+    },
+    {
+      "epoch": 3.963941591338035,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.00048288864607132215,
+      "loss": 0.5251,
+      "step": 79810
+    },
+    {
+      "epoch": 3.9644382636336544,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0004828489122876726,
+      "loss": 0.5405,
+      "step": 79820
+    },
+    {
+      "epoch": 3.964934935929274,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004828091785040231,
+      "loss": 0.5521,
+      "step": 79830
+    },
+    {
+      "epoch": 3.9654316082248933,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004827694447203735,
+      "loss": 0.5328,
+      "step": 79840
+    },
+    {
+      "epoch": 3.9659282805205125,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.000482729710936724,
+      "loss": 0.5432,
+      "step": 79850
+    },
+    {
+      "epoch": 3.9664249528161317,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00048268997715307445,
+      "loss": 0.542,
+      "step": 79860
+    },
+    {
+      "epoch": 3.9669216251117514,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0004826502433694248,
+      "loss": 0.5276,
+      "step": 79870
+    },
+    {
+      "epoch": 3.9674182974073706,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00048261050958577534,
+      "loss": 0.5499,
+      "step": 79880
+    },
+    {
+      "epoch": 3.96791496970299,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0004825707758021258,
+      "loss": 0.5069,
+      "step": 79890
+    },
+    {
+      "epoch": 3.9684116419986095,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0004825310420184763,
+      "loss": 0.5312,
+      "step": 79900
+    },
+    {
+      "epoch": 3.9689083142942287,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00048249130823482665,
+      "loss": 0.5259,
+      "step": 79910
+    },
+    {
+      "epoch": 3.969404986589848,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00048245157445117717,
+      "loss": 0.5166,
+      "step": 79920
+    },
+    {
+      "epoch": 3.969901658885467,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00048241184066752764,
+      "loss": 0.5314,
+      "step": 79930
+    },
+    {
+      "epoch": 3.970398331181087,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.000482372106883878,
+      "loss": 0.5214,
+      "step": 79940
+    },
+    {
+      "epoch": 3.970895003476706,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0004823323731002285,
+      "loss": 0.5146,
+      "step": 79950
+    },
+    {
+      "epoch": 3.9713916757723253,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.000482292639316579,
+      "loss": 0.538,
+      "step": 79960
+    },
+    {
+      "epoch": 3.971888348067945,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00048225290553292937,
+      "loss": 0.5395,
+      "step": 79970
+    },
+    {
+      "epoch": 3.9723850203635642,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.00048221317174927984,
+      "loss": 0.5594,
+      "step": 79980
+    },
+    {
+      "epoch": 3.9728816926591835,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0004821734379656303,
+      "loss": 0.5511,
+      "step": 79990
+    },
+    {
+      "epoch": 3.9733783649548027,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0004821337041819807,
+      "loss": 0.5274,
+      "step": 80000
+    },
+    {
+      "epoch": 3.973875037250422,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0004820939703983312,
+      "loss": 0.5179,
+      "step": 80010
+    },
+    {
+      "epoch": 3.9743717095460416,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00048205423661468167,
+      "loss": 0.5516,
+      "step": 80020
+    },
+    {
+      "epoch": 3.974868381841661,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0004820145028310321,
+      "loss": 0.5319,
+      "step": 80030
+    },
+    {
+      "epoch": 3.9753650541372805,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00048197476904738256,
+      "loss": 0.5173,
+      "step": 80040
+    },
+    {
+      "epoch": 3.9758617264328997,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00048193503526373303,
+      "loss": 0.5147,
+      "step": 80050
+    },
+    {
+      "epoch": 3.976358398728519,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0004818953014800835,
+      "loss": 0.5243,
+      "step": 80060
+    },
+    {
+      "epoch": 3.976855071024138,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0004818555676964339,
+      "loss": 0.4932,
+      "step": 80070
+    },
+    {
+      "epoch": 3.9773517433197574,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0004818158339127844,
+      "loss": 0.5149,
+      "step": 80080
+    },
+    {
+      "epoch": 3.977848415615377,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00048177610012913486,
+      "loss": 0.5575,
+      "step": 80090
+    },
+    {
+      "epoch": 3.9783450879109963,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0004817363663454852,
+      "loss": 0.5325,
+      "step": 80100
+    },
+    {
+      "epoch": 3.978841760206616,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00048169663256183575,
+      "loss": 0.5211,
+      "step": 80110
+    },
+    {
+      "epoch": 3.979338432502235,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0004816568987781862,
+      "loss": 0.4939,
+      "step": 80120
+    },
+    {
+      "epoch": 3.9798351047978544,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0004816171649945366,
+      "loss": 0.5304,
+      "step": 80130
+    },
+    {
+      "epoch": 3.9803317770934736,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00048157743121088705,
+      "loss": 0.5244,
+      "step": 80140
+    },
+    {
+      "epoch": 3.980828449389093,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004815376974272376,
+      "loss": 0.5435,
+      "step": 80150
+    },
+    {
+      "epoch": 3.9813251216847125,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00048149796364358794,
+      "loss": 0.5387,
+      "step": 80160
+    },
+    {
+      "epoch": 3.9818217939803318,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0004814582298599384,
+      "loss": 0.5371,
+      "step": 80170
+    },
+    {
+      "epoch": 3.982318466275951,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0004814184960762889,
+      "loss": 0.5091,
+      "step": 80180
+    },
+    {
+      "epoch": 3.9828151385715707,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004813787622926393,
+      "loss": 0.552,
+      "step": 80190
+    },
+    {
+      "epoch": 3.98331181086719,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00048133902850898977,
+      "loss": 0.4959,
+      "step": 80200
+    },
+    {
+      "epoch": 3.983808483162809,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00048129929472534024,
+      "loss": 0.5101,
+      "step": 80210
+    },
+    {
+      "epoch": 3.9843051554584283,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0004812595609416907,
+      "loss": 0.5362,
+      "step": 80220
+    },
+    {
+      "epoch": 3.984801827754048,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00048121982715804113,
+      "loss": 0.5268,
+      "step": 80230
+    },
+    {
+      "epoch": 3.9852985000496672,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0004811800933743916,
+      "loss": 0.5353,
+      "step": 80240
+    },
+    {
+      "epoch": 3.9857951723452865,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004811403595907421,
+      "loss": 0.5479,
+      "step": 80250
+    },
+    {
+      "epoch": 3.986291844640906,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0004811006258070925,
+      "loss": 0.5581,
+      "step": 80260
+    },
+    {
+      "epoch": 3.9867885169365254,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00048106089202344296,
+      "loss": 0.5255,
+      "step": 80270
+    },
+    {
+      "epoch": 3.9872851892321446,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00048102115823979344,
+      "loss": 0.5291,
+      "step": 80280
+    },
+    {
+      "epoch": 3.987781861527764,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00048098142445614385,
+      "loss": 0.5187,
+      "step": 80290
+    },
+    {
+      "epoch": 3.9882785338233835,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0004809416906724943,
+      "loss": 0.5122,
+      "step": 80300
+    },
+    {
+      "epoch": 3.9887752061190027,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004809019568888448,
+      "loss": 0.512,
+      "step": 80310
+    },
+    {
+      "epoch": 3.989271878414622,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00048086222310519516,
+      "loss": 0.5211,
+      "step": 80320
+    },
+    {
+      "epoch": 3.9897685507102416,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00048082248932154563,
+      "loss": 0.5828,
+      "step": 80330
+    },
+    {
+      "epoch": 3.990265223005861,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00048078275553789615,
+      "loss": 0.5629,
+      "step": 80340
+    },
+    {
+      "epoch": 3.99076189530148,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004807430217542466,
+      "loss": 0.544,
+      "step": 80350
+    },
+    {
+      "epoch": 3.9912585675970993,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.000480703287970597,
+      "loss": 0.5288,
+      "step": 80360
+    },
+    {
+      "epoch": 3.9917552398927185,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00048066355418694746,
+      "loss": 0.5364,
+      "step": 80370
+    },
+    {
+      "epoch": 3.992251912188338,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.000480623820403298,
+      "loss": 0.5254,
+      "step": 80380
+    },
+    {
+      "epoch": 3.9927485844839574,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.00048058408661964835,
+      "loss": 0.5298,
+      "step": 80390
+    },
+    {
+      "epoch": 3.993245256779577,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0004805443528359988,
+      "loss": 0.5143,
+      "step": 80400
+    },
+    {
+      "epoch": 3.9937419290751963,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0004805046190523493,
+      "loss": 0.5279,
+      "step": 80410
+    },
+    {
+      "epoch": 3.9942386013708155,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0004804648852686997,
+      "loss": 0.543,
+      "step": 80420
+    },
+    {
+      "epoch": 3.9947352736664348,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0004804251514850502,
+      "loss": 0.5356,
+      "step": 80430
+    },
+    {
+      "epoch": 3.995231945962054,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00048038541770140065,
+      "loss": 0.5297,
+      "step": 80440
+    },
+    {
+      "epoch": 3.9957286182576737,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00048034568391775107,
+      "loss": 0.5392,
+      "step": 80450
+    },
+    {
+      "epoch": 3.996225290553293,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00048030595013410154,
+      "loss": 0.5283,
+      "step": 80460
+    },
+    {
+      "epoch": 3.996721962848912,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.000480266216350452,
+      "loss": 0.5134,
+      "step": 80470
+    },
+    {
+      "epoch": 3.997218635144532,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00048022648256680243,
+      "loss": 0.5283,
+      "step": 80480
+    },
+    {
+      "epoch": 3.997715307440151,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004801867487831529,
+      "loss": 0.4861,
+      "step": 80490
+    },
+    {
+      "epoch": 3.9982119797357702,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.00048014701499950337,
+      "loss": 0.488,
+      "step": 80500
+    },
+    {
+      "epoch": 3.9987086520313895,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.00048010728121585384,
+      "loss": 0.5418,
+      "step": 80510
+    },
+    {
+      "epoch": 3.999205324327009,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.00048006754743220426,
+      "loss": 0.5178,
+      "step": 80520
+    },
+    {
+      "epoch": 3.9997019966226284,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00048002781364855473,
+      "loss": 0.563,
+      "step": 80530
+    },
+    {
+      "epoch": 4.000198668918248,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004799880798649052,
+      "loss": 0.5175,
+      "step": 80540
+    },
+    {
+      "epoch": 4.000695341213867,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00047994834608125557,
+      "loss": 0.5267,
+      "step": 80550
+    },
+    {
+      "epoch": 4.0011920135094865,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0004799086122976061,
+      "loss": 0.5212,
+      "step": 80560
+    },
+    {
+      "epoch": 4.001688685805106,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00047986887851395656,
+      "loss": 0.5415,
+      "step": 80570
+    },
+    {
+      "epoch": 4.002185358100725,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0004798291447303069,
+      "loss": 0.5447,
+      "step": 80580
+    },
+    {
+      "epoch": 4.002682030396344,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004797894109466574,
+      "loss": 0.5089,
+      "step": 80590
+    },
+    {
+      "epoch": 4.003178702691963,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00047974967716300787,
+      "loss": 0.5216,
+      "step": 80600
+    },
+    {
+      "epoch": 4.0036753749875835,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0004797099433793583,
+      "loss": 0.5156,
+      "step": 80610
+    },
+    {
+      "epoch": 4.004172047283203,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00047967020959570876,
+      "loss": 0.5018,
+      "step": 80620
+    },
+    {
+      "epoch": 4.004668719578822,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00047963047581205923,
+      "loss": 0.5122,
+      "step": 80630
+    },
+    {
+      "epoch": 4.005165391874441,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00047959074202840964,
+      "loss": 0.5377,
+      "step": 80640
+    },
+    {
+      "epoch": 4.00566206417006,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0004795510082447601,
+      "loss": 0.5088,
+      "step": 80650
+    },
+    {
+      "epoch": 4.00615873646568,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0004795112744611106,
+      "loss": 0.5172,
+      "step": 80660
+    },
+    {
+      "epoch": 4.006655408761299,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00047947154067746106,
+      "loss": 0.5396,
+      "step": 80670
+    },
+    {
+      "epoch": 4.007152081056919,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0004794318068938115,
+      "loss": 0.5247,
+      "step": 80680
+    },
+    {
+      "epoch": 4.007648753352538,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00047939207311016195,
+      "loss": 0.5022,
+      "step": 80690
+    },
+    {
+      "epoch": 4.0081454256481575,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0004793523393265124,
+      "loss": 0.499,
+      "step": 80700
+    },
+    {
+      "epoch": 4.008642097943777,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.00047931260554286284,
+      "loss": 0.4814,
+      "step": 80710
+    },
+    {
+      "epoch": 4.009138770239396,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0004792728717592133,
+      "loss": 0.5266,
+      "step": 80720
+    },
+    {
+      "epoch": 4.009635442535015,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0004792331379755638,
+      "loss": 0.5145,
+      "step": 80730
+    },
+    {
+      "epoch": 4.010132114830634,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00047919340419191414,
+      "loss": 0.4964,
+      "step": 80740
+    },
+    {
+      "epoch": 4.0106287871262545,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00047915367040826467,
+      "loss": 0.5007,
+      "step": 80750
+    },
+    {
+      "epoch": 4.011125459421874,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00047911393662461514,
+      "loss": 0.5183,
+      "step": 80760
+    },
+    {
+      "epoch": 4.011622131717493,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0004790742028409655,
+      "loss": 0.5162,
+      "step": 80770
+    },
+    {
+      "epoch": 4.012118804013112,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.00047903446905731597,
+      "loss": 0.5259,
+      "step": 80780
+    },
+    {
+      "epoch": 4.012615476308731,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0004789947352736665,
+      "loss": 0.5069,
+      "step": 80790
+    },
+    {
+      "epoch": 4.013112148604351,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00047895500149001697,
+      "loss": 0.4948,
+      "step": 80800
+    },
+    {
+      "epoch": 4.01360882089997,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00047891526770636733,
+      "loss": 0.5593,
+      "step": 80810
+    },
+    {
+      "epoch": 4.01410549319559,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004788755339227178,
+      "loss": 0.5431,
+      "step": 80820
+    },
+    {
+      "epoch": 4.014602165491209,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0004788358001390683,
+      "loss": 0.5352,
+      "step": 80830
+    },
+    {
+      "epoch": 4.015098837786828,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0004787960663554187,
+      "loss": 0.5245,
+      "step": 80840
+    },
+    {
+      "epoch": 4.015595510082448,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00047875633257176916,
+      "loss": 0.5241,
+      "step": 80850
+    },
+    {
+      "epoch": 4.016092182378067,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00047871659878811964,
+      "loss": 0.5128,
+      "step": 80860
+    },
+    {
+      "epoch": 4.016588854673686,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00047867686500447005,
+      "loss": 0.5192,
+      "step": 80870
+    },
+    {
+      "epoch": 4.017085526969305,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0004786371312208205,
+      "loss": 0.5221,
+      "step": 80880
+    },
+    {
+      "epoch": 4.017582199264925,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.000478597397437171,
+      "loss": 0.5048,
+      "step": 80890
+    },
+    {
+      "epoch": 4.018078871560545,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004785576636535214,
+      "loss": 0.5098,
+      "step": 80900
+    },
+    {
+      "epoch": 4.018575543856164,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0004785179298698719,
+      "loss": 0.5253,
+      "step": 80910
+    },
+    {
+      "epoch": 4.019072216151783,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.00047847819608622235,
+      "loss": 0.5029,
+      "step": 80920
+    },
+    {
+      "epoch": 4.019568888447402,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004784384623025727,
+      "loss": 0.5064,
+      "step": 80930
+    },
+    {
+      "epoch": 4.020065560743022,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.00047839872851892324,
+      "loss": 0.5097,
+      "step": 80940
+    },
+    {
+      "epoch": 4.020562233038641,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0004783589947352737,
+      "loss": 0.4999,
+      "step": 80950
+    },
+    {
+      "epoch": 4.02105890533426,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0004783192609516242,
+      "loss": 0.5068,
+      "step": 80960
+    },
+    {
+      "epoch": 4.02155557762988,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00047827952716797455,
+      "loss": 0.4947,
+      "step": 80970
+    },
+    {
+      "epoch": 4.022052249925499,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0004782397933843251,
+      "loss": 0.5178,
+      "step": 80980
+    },
+    {
+      "epoch": 4.022548922221119,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00047820005960067555,
+      "loss": 0.5396,
+      "step": 80990
+    },
+    {
+      "epoch": 4.023045594516738,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0004781603258170259,
+      "loss": 0.561,
+      "step": 81000
+    },
+    {
+      "epoch": 4.023542266812357,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0004781205920333764,
+      "loss": 0.5209,
+      "step": 81010
+    },
+    {
+      "epoch": 4.024038939107976,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0004780808582497269,
+      "loss": 0.5166,
+      "step": 81020
+    },
+    {
+      "epoch": 4.0245356114035955,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00047804112446607727,
+      "loss": 0.5519,
+      "step": 81030
+    },
+    {
+      "epoch": 4.025032283699216,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00047800139068242774,
+      "loss": 0.5737,
+      "step": 81040
+    },
+    {
+      "epoch": 4.025528955994835,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0004779616568987782,
+      "loss": 0.5398,
+      "step": 81050
+    },
+    {
+      "epoch": 4.026025628290454,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00047792192311512863,
+      "loss": 0.5005,
+      "step": 81060
+    },
+    {
+      "epoch": 4.026522300586073,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0004778821893314791,
+      "loss": 0.536,
+      "step": 81070
+    },
+    {
+      "epoch": 4.0270189728816925,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00047784245554782957,
+      "loss": 0.5223,
+      "step": 81080
+    },
+    {
+      "epoch": 4.027515645177312,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00047780272176418004,
+      "loss": 0.505,
+      "step": 81090
+    },
+    {
+      "epoch": 4.028012317472931,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00047776298798053046,
+      "loss": 0.499,
+      "step": 81100
+    },
+    {
+      "epoch": 4.028508989768551,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00047772325419688093,
+      "loss": 0.5178,
+      "step": 81110
+    },
+    {
+      "epoch": 4.02900566206417,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0004776835204132314,
+      "loss": 0.5239,
+      "step": 81120
+    },
+    {
+      "epoch": 4.0295023343597896,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004776437866295818,
+      "loss": 0.5163,
+      "step": 81130
+    },
+    {
+      "epoch": 4.029999006655409,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0004776040528459323,
+      "loss": 0.5398,
+      "step": 81140
+    },
+    {
+      "epoch": 4.030495678951028,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00047756431906228276,
+      "loss": 0.5174,
+      "step": 81150
+    },
+    {
+      "epoch": 4.030992351246647,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0004775245852786331,
+      "loss": 0.5257,
+      "step": 81160
+    },
+    {
+      "epoch": 4.0314890235422665,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00047748485149498365,
+      "loss": 0.493,
+      "step": 81170
+    },
+    {
+      "epoch": 4.031985695837887,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0004774451177113341,
+      "loss": 0.5354,
+      "step": 81180
+    },
+    {
+      "epoch": 4.032482368133506,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004774053839276845,
+      "loss": 0.5447,
+      "step": 81190
+    },
+    {
+      "epoch": 4.032979040429125,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.00047736565014403496,
+      "loss": 0.5177,
+      "step": 81200
+    },
+    {
+      "epoch": 4.033475712724744,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004773259163603855,
+      "loss": 0.5483,
+      "step": 81210
+    },
+    {
+      "epoch": 4.0339723850203635,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00047728618257673585,
+      "loss": 0.5108,
+      "step": 81220
+    },
+    {
+      "epoch": 4.034469057315983,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0004772464487930863,
+      "loss": 0.528,
+      "step": 81230
+    },
+    {
+      "epoch": 4.034965729611602,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0004772067150094368,
+      "loss": 0.5305,
+      "step": 81240
+    },
+    {
+      "epoch": 4.035462401907222,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0004771669812257873,
+      "loss": 0.5205,
+      "step": 81250
+    },
+    {
+      "epoch": 4.035959074202841,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0004771272474421377,
+      "loss": 0.5339,
+      "step": 81260
+    },
+    {
+      "epoch": 4.0364557464984605,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00047708751365848815,
+      "loss": 0.4839,
+      "step": 81270
+    },
+    {
+      "epoch": 4.03695241879408,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0004770477798748386,
+      "loss": 0.525,
+      "step": 81280
+    },
+    {
+      "epoch": 4.037449091089699,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00047700804609118904,
+      "loss": 0.5113,
+      "step": 81290
+    },
+    {
+      "epoch": 4.037945763385318,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0004769683123075395,
+      "loss": 0.5052,
+      "step": 81300
+    },
+    {
+      "epoch": 4.038442435680937,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00047692857852389,
+      "loss": 0.522,
+      "step": 81310
+    },
+    {
+      "epoch": 4.038939107976557,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0004768888447402404,
+      "loss": 0.5277,
+      "step": 81320
+    },
+    {
+      "epoch": 4.039435780272177,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00047684911095659087,
+      "loss": 0.5198,
+      "step": 81330
+    },
+    {
+      "epoch": 4.039932452567796,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00047680937717294134,
+      "loss": 0.4995,
+      "step": 81340
+    },
+    {
+      "epoch": 4.040429124863415,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.00047676964338929176,
+      "loss": 0.4999,
+      "step": 81350
+    },
+    {
+      "epoch": 4.040925797159034,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00047672990960564223,
+      "loss": 0.5285,
+      "step": 81360
+    },
+    {
+      "epoch": 4.041422469454654,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0004766901758219927,
+      "loss": 0.5227,
+      "step": 81370
+    },
+    {
+      "epoch": 4.041919141750273,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00047665044203834306,
+      "loss": 0.5269,
+      "step": 81380
+    },
+    {
+      "epoch": 4.042415814045892,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0004766107082546936,
+      "loss": 0.4916,
+      "step": 81390
+    },
+    {
+      "epoch": 4.042912486341512,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.00047657097447104406,
+      "loss": 0.5073,
+      "step": 81400
+    },
+    {
+      "epoch": 4.0434091586371315,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00047653124068739453,
+      "loss": 0.5292,
+      "step": 81410
+    },
+    {
+      "epoch": 4.043905830932751,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0004764915069037449,
+      "loss": 0.5153,
+      "step": 81420
+    },
+    {
+      "epoch": 4.04440250322837,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00047645177312009536,
+      "loss": 0.5215,
+      "step": 81430
+    },
+    {
+      "epoch": 4.044899175523989,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0004764120393364459,
+      "loss": 0.5353,
+      "step": 81440
+    },
+    {
+      "epoch": 4.045395847819608,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00047637230555279625,
+      "loss": 0.5035,
+      "step": 81450
+    },
+    {
+      "epoch": 4.045892520115228,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0004763325717691467,
+      "loss": 0.5241,
+      "step": 81460
+    },
+    {
+      "epoch": 4.046389192410848,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0004762928379854972,
+      "loss": 0.4885,
+      "step": 81470
+    },
+    {
+      "epoch": 4.046885864706467,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004762531042018476,
+      "loss": 0.5404,
+      "step": 81480
+    },
+    {
+      "epoch": 4.047382537002086,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004762133704181981,
+      "loss": 0.5116,
+      "step": 81490
+    },
+    {
+      "epoch": 4.047879209297705,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00047617363663454856,
+      "loss": 0.5134,
+      "step": 81500
+    },
+    {
+      "epoch": 4.048375881593325,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.00047613390285089897,
+      "loss": 0.511,
+      "step": 81510
+    },
+    {
+      "epoch": 4.048872553888944,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00047609416906724944,
+      "loss": 0.5603,
+      "step": 81520
+    },
+    {
+      "epoch": 4.049369226184563,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004760544352835999,
+      "loss": 0.5238,
+      "step": 81530
+    },
+    {
+      "epoch": 4.049865898480183,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0004760147014999504,
+      "loss": 0.5093,
+      "step": 81540
+    },
+    {
+      "epoch": 4.050362570775802,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0004759749677163008,
+      "loss": 0.5349,
+      "step": 81550
+    },
+    {
+      "epoch": 4.050859243071422,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0004759352339326513,
+      "loss": 0.5305,
+      "step": 81560
+    },
+    {
+      "epoch": 4.051355915367041,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.00047589550014900175,
+      "loss": 0.5223,
+      "step": 81570
+    },
+    {
+      "epoch": 4.05185258766266,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00047585576636535216,
+      "loss": 0.5055,
+      "step": 81580
+    },
+    {
+      "epoch": 4.052349259958279,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.00047581603258170263,
+      "loss": 0.5241,
+      "step": 81590
+    },
+    {
+      "epoch": 4.0528459322538986,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004757762987980531,
+      "loss": 0.5139,
+      "step": 81600
+    },
+    {
+      "epoch": 4.053342604549519,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00047573656501440347,
+      "loss": 0.5414,
+      "step": 81610
+    },
+    {
+      "epoch": 4.053839276845138,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.000475696831230754,
+      "loss": 0.5034,
+      "step": 81620
+    },
+    {
+      "epoch": 4.054335949140757,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00047565709744710447,
+      "loss": 0.513,
+      "step": 81630
+    },
+    {
+      "epoch": 4.054832621436376,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00047561736366345483,
+      "loss": 0.5004,
+      "step": 81640
+    },
+    {
+      "epoch": 4.055329293731996,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004755776298798053,
+      "loss": 0.5158,
+      "step": 81650
+    },
+    {
+      "epoch": 4.055825966027615,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0004755378960961558,
+      "loss": 0.5057,
+      "step": 81660
+    },
+    {
+      "epoch": 4.056322638323234,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0004754981623125062,
+      "loss": 0.503,
+      "step": 81670
+    },
+    {
+      "epoch": 4.056819310618853,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00047545842852885666,
+      "loss": 0.5453,
+      "step": 81680
+    },
+    {
+      "epoch": 4.057315982914473,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00047541869474520713,
+      "loss": 0.5194,
+      "step": 81690
+    },
+    {
+      "epoch": 4.057812655210093,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0004753789609615576,
+      "loss": 0.5501,
+      "step": 81700
+    },
+    {
+      "epoch": 4.058309327505712,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.000475339227177908,
+      "loss": 0.5369,
+      "step": 81710
+    },
+    {
+      "epoch": 4.058805999801331,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004752994933942585,
+      "loss": 0.5098,
+      "step": 81720
+    },
+    {
+      "epoch": 4.05930267209695,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00047525975961060896,
+      "loss": 0.5112,
+      "step": 81730
+    },
+    {
+      "epoch": 4.0597993443925695,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004752200258269594,
+      "loss": 0.5086,
+      "step": 81740
+    },
+    {
+      "epoch": 4.060296016688189,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00047518029204330985,
+      "loss": 0.5346,
+      "step": 81750
+    },
+    {
+      "epoch": 4.060792688983809,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0004751405582596603,
+      "loss": 0.5093,
+      "step": 81760
+    },
+    {
+      "epoch": 4.061289361279428,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00047510082447601074,
+      "loss": 0.4806,
+      "step": 81770
+    },
+    {
+      "epoch": 4.061786033575047,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0004750610906923612,
+      "loss": 0.5136,
+      "step": 81780
+    },
+    {
+      "epoch": 4.0622827058706665,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0004750213569087117,
+      "loss": 0.4983,
+      "step": 81790
+    },
+    {
+      "epoch": 4.062779378166286,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00047498162312506205,
+      "loss": 0.5402,
+      "step": 81800
+    },
+    {
+      "epoch": 4.063276050461905,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00047494188934141257,
+      "loss": 0.4821,
+      "step": 81810
+    },
+    {
+      "epoch": 4.063772722757524,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00047490215555776304,
+      "loss": 0.521,
+      "step": 81820
+    },
+    {
+      "epoch": 4.064269395053144,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004748624217741134,
+      "loss": 0.4949,
+      "step": 81830
+    },
+    {
+      "epoch": 4.064766067348764,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0004748226879904639,
+      "loss": 0.5308,
+      "step": 81840
+    },
+    {
+      "epoch": 4.065262739644383,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004747829542068144,
+      "loss": 0.5349,
+      "step": 81850
+    },
+    {
+      "epoch": 4.065759411940002,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004747432204231649,
+      "loss": 0.5077,
+      "step": 81860
+    },
+    {
+      "epoch": 4.066256084235621,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00047470348663951524,
+      "loss": 0.5036,
+      "step": 81870
+    },
+    {
+      "epoch": 4.0667527565312405,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0004746637528558657,
+      "loss": 0.5353,
+      "step": 81880
+    },
+    {
+      "epoch": 4.06724942882686,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00047462401907221623,
+      "loss": 0.5128,
+      "step": 81890
+    },
+    {
+      "epoch": 4.06774610112248,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0004745842852885666,
+      "loss": 0.5283,
+      "step": 81900
+    },
+    {
+      "epoch": 4.068242773418099,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00047454455150491707,
+      "loss": 0.532,
+      "step": 81910
+    },
+    {
+      "epoch": 4.068739445713718,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00047450481772126754,
+      "loss": 0.5092,
+      "step": 81920
+    },
+    {
+      "epoch": 4.0692361180093375,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00047446508393761796,
+      "loss": 0.5129,
+      "step": 81930
+    },
+    {
+      "epoch": 4.069732790304957,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00047442535015396843,
+      "loss": 0.5507,
+      "step": 81940
+    },
+    {
+      "epoch": 4.070229462600576,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0004743856163703189,
+      "loss": 0.5067,
+      "step": 81950
+    },
+    {
+      "epoch": 4.070726134896195,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0004743458825866693,
+      "loss": 0.5173,
+      "step": 81960
+    },
+    {
+      "epoch": 4.071222807191814,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0004743061488030198,
+      "loss": 0.5237,
+      "step": 81970
+    },
+    {
+      "epoch": 4.0717194794874345,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00047426641501937026,
+      "loss": 0.529,
+      "step": 81980
+    },
+    {
+      "epoch": 4.072216151783054,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00047422668123572073,
+      "loss": 0.497,
+      "step": 81990
+    },
+    {
+      "epoch": 4.072712824078673,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00047418694745207115,
+      "loss": 0.4832,
+      "step": 82000
+    },
+    {
+      "epoch": 4.073209496374292,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0004741472136684216,
+      "loss": 0.5452,
+      "step": 82010
+    },
+    {
+      "epoch": 4.073706168669911,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0004741074798847721,
+      "loss": 0.501,
+      "step": 82020
+    },
+    {
+      "epoch": 4.074202840965531,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00047406774610112245,
+      "loss": 0.5312,
+      "step": 82030
+    },
+    {
+      "epoch": 4.07469951326115,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.000474028012317473,
+      "loss": 0.521,
+      "step": 82040
+    },
+    {
+      "epoch": 4.07519618555677,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00047398827853382345,
+      "loss": 0.5259,
+      "step": 82050
+    },
+    {
+      "epoch": 4.075692857852389,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004739485447501738,
+      "loss": 0.5127,
+      "step": 82060
+    },
+    {
+      "epoch": 4.0761895301480084,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004739088109665243,
+      "loss": 0.5266,
+      "step": 82070
+    },
+    {
+      "epoch": 4.076686202443628,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0004738690771828748,
+      "loss": 0.5267,
+      "step": 82080
+    },
+    {
+      "epoch": 4.077182874739247,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00047382934339922517,
+      "loss": 0.5366,
+      "step": 82090
+    },
+    {
+      "epoch": 4.077679547034866,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00047378960961557564,
+      "loss": 0.505,
+      "step": 82100
+    },
+    {
+      "epoch": 4.078176219330485,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0004737498758319261,
+      "loss": 0.5062,
+      "step": 82110
+    },
+    {
+      "epoch": 4.0786728916261055,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00047371014204827653,
+      "loss": 0.5252,
+      "step": 82120
+    },
+    {
+      "epoch": 4.079169563921725,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.000473670408264627,
+      "loss": 0.5392,
+      "step": 82130
+    },
+    {
+      "epoch": 4.079666236217344,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0004736306744809775,
+      "loss": 0.5166,
+      "step": 82140
+    },
+    {
+      "epoch": 4.080162908512963,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00047359094069732795,
+      "loss": 0.5286,
+      "step": 82150
+    },
+    {
+      "epoch": 4.080659580808582,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00047355120691367836,
+      "loss": 0.4974,
+      "step": 82160
+    },
+    {
+      "epoch": 4.081156253104202,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00047351147313002883,
+      "loss": 0.5004,
+      "step": 82170
+    },
+    {
+      "epoch": 4.081652925399821,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0004734717393463793,
+      "loss": 0.5137,
+      "step": 82180
+    },
+    {
+      "epoch": 4.082149597695441,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0004734320055627297,
+      "loss": 0.4974,
+      "step": 82190
+    },
+    {
+      "epoch": 4.08264626999106,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0004733922717790802,
+      "loss": 0.5247,
+      "step": 82200
+    },
+    {
+      "epoch": 4.083142942286679,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.00047335253799543067,
+      "loss": 0.5155,
+      "step": 82210
+    },
+    {
+      "epoch": 4.083639614582299,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0004733128042117811,
+      "loss": 0.4963,
+      "step": 82220
+    },
+    {
+      "epoch": 4.084136286877918,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00047327307042813155,
+      "loss": 0.5233,
+      "step": 82230
+    },
+    {
+      "epoch": 4.084632959173537,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.000473233336644482,
+      "loss": 0.5229,
+      "step": 82240
+    },
+    {
+      "epoch": 4.085129631469156,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004731936028608324,
+      "loss": 0.491,
+      "step": 82250
+    },
+    {
+      "epoch": 4.085626303764776,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0004731538690771829,
+      "loss": 0.5364,
+      "step": 82260
+    },
+    {
+      "epoch": 4.086122976060396,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0004731141352935334,
+      "loss": 0.4996,
+      "step": 82270
+    },
+    {
+      "epoch": 4.086619648356015,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00047307440150988375,
+      "loss": 0.5088,
+      "step": 82280
+    },
+    {
+      "epoch": 4.087116320651634,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0004730346677262342,
+      "loss": 0.5234,
+      "step": 82290
+    },
+    {
+      "epoch": 4.087612992947253,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0004729949339425847,
+      "loss": 0.5018,
+      "step": 82300
+    },
+    {
+      "epoch": 4.088109665242873,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0004729552001589352,
+      "loss": 0.5146,
+      "step": 82310
+    },
+    {
+      "epoch": 4.088606337538492,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0004729154663752856,
+      "loss": 0.5191,
+      "step": 82320
+    },
+    {
+      "epoch": 4.089103009834112,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00047287573259163605,
+      "loss": 0.5376,
+      "step": 82330
+    },
+    {
+      "epoch": 4.089599682129731,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0004728359988079865,
+      "loss": 0.5175,
+      "step": 82340
+    },
+    {
+      "epoch": 4.09009635442535,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.00047279626502433694,
+      "loss": 0.5092,
+      "step": 82350
+    },
+    {
+      "epoch": 4.09059302672097,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0004727565312406874,
+      "loss": 0.5021,
+      "step": 82360
+    },
+    {
+      "epoch": 4.091089699016589,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004727167974570379,
+      "loss": 0.5344,
+      "step": 82370
+    },
+    {
+      "epoch": 4.091586371312208,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0004726770636733883,
+      "loss": 0.4935,
+      "step": 82380
+    },
+    {
+      "epoch": 4.092083043607827,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00047263732988973877,
+      "loss": 0.491,
+      "step": 82390
+    },
+    {
+      "epoch": 4.0925797159034465,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00047259759610608924,
+      "loss": 0.5281,
+      "step": 82400
+    },
+    {
+      "epoch": 4.093076388199067,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00047255786232243966,
+      "loss": 0.4877,
+      "step": 82410
+    },
+    {
+      "epoch": 4.093573060494686,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00047251812853879013,
+      "loss": 0.5048,
+      "step": 82420
+    },
+    {
+      "epoch": 4.094069732790305,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0004724783947551406,
+      "loss": 0.5371,
+      "step": 82430
+    },
+    {
+      "epoch": 4.094566405085924,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0004724386609714911,
+      "loss": 0.5233,
+      "step": 82440
+    },
+    {
+      "epoch": 4.0950630773815435,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0004723989271878415,
+      "loss": 0.5133,
+      "step": 82450
+    },
+    {
+      "epoch": 4.095559749677163,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00047235919340419196,
+      "loss": 0.5158,
+      "step": 82460
+    },
+    {
+      "epoch": 4.096056421972782,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00047231945962054243,
+      "loss": 0.517,
+      "step": 82470
+    },
+    {
+      "epoch": 4.096553094268402,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004722797258368928,
+      "loss": 0.4954,
+      "step": 82480
+    },
+    {
+      "epoch": 4.097049766564021,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0004722399920532433,
+      "loss": 0.5403,
+      "step": 82490
+    },
+    {
+      "epoch": 4.0975464388596405,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004722002582695938,
+      "loss": 0.5425,
+      "step": 82500
+    },
+    {
+      "epoch": 4.09804311115526,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00047216052448594416,
+      "loss": 0.524,
+      "step": 82510
+    },
+    {
+      "epoch": 4.098539783450879,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00047212079070229463,
+      "loss": 0.4932,
+      "step": 82520
+    },
+    {
+      "epoch": 4.099036455746498,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0004720810569186451,
+      "loss": 0.5,
+      "step": 82530
+    },
+    {
+      "epoch": 4.0995331280421174,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0004720413231349955,
+      "loss": 0.5181,
+      "step": 82540
+    },
+    {
+      "epoch": 4.100029800337738,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.000472001589351346,
+      "loss": 0.502,
+      "step": 82550
+    },
+    {
+      "epoch": 4.100526472633357,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00047196185556769646,
+      "loss": 0.5251,
+      "step": 82560
+    },
+    {
+      "epoch": 4.101023144928976,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004719221217840469,
+      "loss": 0.5354,
+      "step": 82570
+    },
+    {
+      "epoch": 4.101519817224595,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00047188238800039735,
+      "loss": 0.5215,
+      "step": 82580
+    },
+    {
+      "epoch": 4.1020164895202145,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0004718426542167478,
+      "loss": 0.5062,
+      "step": 82590
+    },
+    {
+      "epoch": 4.102513161815834,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004718029204330983,
+      "loss": 0.5057,
+      "step": 82600
+    },
+    {
+      "epoch": 4.103009834111453,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0004717631866494487,
+      "loss": 0.5163,
+      "step": 82610
+    },
+    {
+      "epoch": 4.103506506407073,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0004717234528657992,
+      "loss": 0.5181,
+      "step": 82620
+    },
+    {
+      "epoch": 4.104003178702692,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00047168371908214965,
+      "loss": 0.4976,
+      "step": 82630
+    },
+    {
+      "epoch": 4.1044998509983115,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00047164398529850007,
+      "loss": 0.5093,
+      "step": 82640
+    },
+    {
+      "epoch": 4.104996523293931,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00047160425151485054,
+      "loss": 0.5185,
+      "step": 82650
+    },
+    {
+      "epoch": 4.10549319558955,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.000471564517731201,
+      "loss": 0.4915,
+      "step": 82660
+    },
+    {
+      "epoch": 4.105989867885169,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00047152478394755137,
+      "loss": 0.5473,
+      "step": 82670
+    },
+    {
+      "epoch": 4.106486540180788,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0004714850501639019,
+      "loss": 0.5313,
+      "step": 82680
+    },
+    {
+      "epoch": 4.106983212476408,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00047144531638025237,
+      "loss": 0.5496,
+      "step": 82690
+    },
+    {
+      "epoch": 4.107479884772028,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00047140558259660273,
+      "loss": 0.5321,
+      "step": 82700
+    },
+    {
+      "epoch": 4.107976557067647,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0004713658488129532,
+      "loss": 0.5213,
+      "step": 82710
+    },
+    {
+      "epoch": 4.108473229363266,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00047132611502930373,
+      "loss": 0.5279,
+      "step": 82720
+    },
+    {
+      "epoch": 4.108969901658885,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004712863812456541,
+      "loss": 0.4876,
+      "step": 82730
+    },
+    {
+      "epoch": 4.109466573954505,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00047124664746200456,
+      "loss": 0.5223,
+      "step": 82740
+    },
+    {
+      "epoch": 4.109963246250124,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00047120691367835504,
+      "loss": 0.5293,
+      "step": 82750
+    },
+    {
+      "epoch": 4.110459918545743,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00047116717989470556,
+      "loss": 0.5532,
+      "step": 82760
+    },
+    {
+      "epoch": 4.110956590841363,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004711274461110559,
+      "loss": 0.5227,
+      "step": 82770
+    },
+    {
+      "epoch": 4.1114532631369825,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0004710877123274064,
+      "loss": 0.5163,
+      "step": 82780
+    },
+    {
+      "epoch": 4.111949935432602,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00047104797854375687,
+      "loss": 0.4817,
+      "step": 82790
+    },
+    {
+      "epoch": 4.112446607728221,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004710082447601073,
+      "loss": 0.5271,
+      "step": 82800
+    },
+    {
+      "epoch": 4.11294328002384,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00047096851097645775,
+      "loss": 0.5263,
+      "step": 82810
+    },
+    {
+      "epoch": 4.113439952319459,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0004709287771928082,
+      "loss": 0.489,
+      "step": 82820
+    },
+    {
+      "epoch": 4.113936624615079,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00047088904340915864,
+      "loss": 0.5237,
+      "step": 82830
+    },
+    {
+      "epoch": 4.114433296910699,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0004708493096255091,
+      "loss": 0.5245,
+      "step": 82840
+    },
+    {
+      "epoch": 4.114929969206318,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004708095758418596,
+      "loss": 0.5509,
+      "step": 82850
+    },
+    {
+      "epoch": 4.115426641501937,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.00047076984205820995,
+      "loss": 0.5182,
+      "step": 82860
+    },
+    {
+      "epoch": 4.115923313797556,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0004707301082745605,
+      "loss": 0.5243,
+      "step": 82870
+    },
+    {
+      "epoch": 4.116419986093176,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00047069037449091095,
+      "loss": 0.497,
+      "step": 82880
+    },
+    {
+      "epoch": 4.116916658388795,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004706506407072614,
+      "loss": 0.532,
+      "step": 82890
+    },
+    {
+      "epoch": 4.117413330684414,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0004706109069236118,
+      "loss": 0.5535,
+      "step": 82900
+    },
+    {
+      "epoch": 4.117910002980034,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0004705711731399623,
+      "loss": 0.522,
+      "step": 82910
+    },
+    {
+      "epoch": 4.118406675275653,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0004705314393563128,
+      "loss": 0.4981,
+      "step": 82920
+    },
+    {
+      "epoch": 4.118903347571273,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00047049170557266314,
+      "loss": 0.5042,
+      "step": 82930
+    },
+    {
+      "epoch": 4.119400019866892,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004704519717890136,
+      "loss": 0.5407,
+      "step": 82940
+    },
+    {
+      "epoch": 4.119896692162511,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00047041223800536414,
+      "loss": 0.495,
+      "step": 82950
+    },
+    {
+      "epoch": 4.12039336445813,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0004703725042217145,
+      "loss": 0.4886,
+      "step": 82960
+    },
+    {
+      "epoch": 4.1208900367537495,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00047033277043806497,
+      "loss": 0.5248,
+      "step": 82970
+    },
+    {
+      "epoch": 4.12138670904937,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00047029303665441544,
+      "loss": 0.5313,
+      "step": 82980
+    },
+    {
+      "epoch": 4.121883381344989,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.00047025330287076586,
+      "loss": 0.4949,
+      "step": 82990
+    },
+    {
+      "epoch": 4.122380053640608,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.00047021356908711633,
+      "loss": 0.4992,
+      "step": 83000
+    },
+    {
+      "epoch": 4.122876725936227,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004701738353034668,
+      "loss": 0.541,
+      "step": 83010
+    },
+    {
+      "epoch": 4.123373398231847,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0004701341015198172,
+      "loss": 0.5312,
+      "step": 83020
+    },
+    {
+      "epoch": 4.123870070527466,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004700943677361677,
+      "loss": 0.513,
+      "step": 83030
+    },
+    {
+      "epoch": 4.124366742823085,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00047005463395251816,
+      "loss": 0.5379,
+      "step": 83040
+    },
+    {
+      "epoch": 4.124863415118704,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00047001490016886863,
+      "loss": 0.508,
+      "step": 83050
+    },
+    {
+      "epoch": 4.125360087414324,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00046997516638521905,
+      "loss": 0.5071,
+      "step": 83060
+    },
+    {
+      "epoch": 4.125856759709944,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004699354326015695,
+      "loss": 0.4966,
+      "step": 83070
+    },
+    {
+      "epoch": 4.126353432005563,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00046989569881792,
+      "loss": 0.5498,
+      "step": 83080
+    },
+    {
+      "epoch": 4.126850104301182,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0004698559650342704,
+      "loss": 0.5128,
+      "step": 83090
+    },
+    {
+      "epoch": 4.127346776596801,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004698162312506209,
+      "loss": 0.534,
+      "step": 83100
+    },
+    {
+      "epoch": 4.1278434488924205,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00046977649746697135,
+      "loss": 0.5198,
+      "step": 83110
+    },
+    {
+      "epoch": 4.12834012118804,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0004697367636833217,
+      "loss": 0.5052,
+      "step": 83120
+    },
+    {
+      "epoch": 4.12883679348366,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004696970298996722,
+      "loss": 0.528,
+      "step": 83130
+    },
+    {
+      "epoch": 4.129333465779279,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0004696572961160227,
+      "loss": 0.5558,
+      "step": 83140
+    },
+    {
+      "epoch": 4.129830138074898,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004696175623323731,
+      "loss": 0.5222,
+      "step": 83150
+    },
+    {
+      "epoch": 4.1303268103705175,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00046957782854872355,
+      "loss": 0.5299,
+      "step": 83160
+    },
+    {
+      "epoch": 4.130823482666137,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.000469538094765074,
+      "loss": 0.5209,
+      "step": 83170
+    },
+    {
+      "epoch": 4.131320154961756,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00046949836098142444,
+      "loss": 0.528,
+      "step": 83180
+    },
+    {
+      "epoch": 4.131816827257375,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0004694586271977749,
+      "loss": 0.5078,
+      "step": 83190
+    },
+    {
+      "epoch": 4.132313499552995,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0004694188934141254,
+      "loss": 0.5392,
+      "step": 83200
+    },
+    {
+      "epoch": 4.1328101718486145,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00046937915963047585,
+      "loss": 0.4914,
+      "step": 83210
+    },
+    {
+      "epoch": 4.133306844144234,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00046933942584682627,
+      "loss": 0.5129,
+      "step": 83220
+    },
+    {
+      "epoch": 4.133803516439853,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00046929969206317674,
+      "loss": 0.531,
+      "step": 83230
+    },
+    {
+      "epoch": 4.134300188735472,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004692599582795272,
+      "loss": 0.5036,
+      "step": 83240
+    },
+    {
+      "epoch": 4.1347968610310915,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.00046922022449587763,
+      "loss": 0.5092,
+      "step": 83250
+    },
+    {
+      "epoch": 4.135293533326711,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0004691804907122281,
+      "loss": 0.5096,
+      "step": 83260
+    },
+    {
+      "epoch": 4.135790205622331,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00046914075692857857,
+      "loss": 0.5065,
+      "step": 83270
+    },
+    {
+      "epoch": 4.13628687791795,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.000469101023144929,
+      "loss": 0.5523,
+      "step": 83280
+    },
+    {
+      "epoch": 4.136783550213569,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00046906128936127946,
+      "loss": 0.5279,
+      "step": 83290
+    },
+    {
+      "epoch": 4.1372802225091885,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.00046902155557762993,
+      "loss": 0.5352,
+      "step": 83300
+    },
+    {
+      "epoch": 4.137776894804808,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004689818217939803,
+      "loss": 0.5355,
+      "step": 83310
+    },
+    {
+      "epoch": 4.138273567100427,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0004689420880103308,
+      "loss": 0.563,
+      "step": 83320
+    },
+    {
+      "epoch": 4.138770239396046,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0004689023542266813,
+      "loss": 0.5084,
+      "step": 83330
+    },
+    {
+      "epoch": 4.139266911691665,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00046886262044303176,
+      "loss": 0.5208,
+      "step": 83340
+    },
+    {
+      "epoch": 4.1397635839872855,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0004688228866593821,
+      "loss": 0.5138,
+      "step": 83350
+    },
+    {
+      "epoch": 4.140260256282905,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00046878315287573265,
+      "loss": 0.5155,
+      "step": 83360
+    },
+    {
+      "epoch": 4.140756928578524,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0004687434190920831,
+      "loss": 0.5194,
+      "step": 83370
+    },
+    {
+      "epoch": 4.141253600874143,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0004687036853084335,
+      "loss": 0.5232,
+      "step": 83380
+    },
+    {
+      "epoch": 4.141750273169762,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00046866395152478396,
+      "loss": 0.5192,
+      "step": 83390
+    },
+    {
+      "epoch": 4.142246945465382,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0004686242177411344,
+      "loss": 0.532,
+      "step": 83400
+    },
+    {
+      "epoch": 4.142743617761001,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00046858448395748484,
+      "loss": 0.5369,
+      "step": 83410
+    },
+    {
+      "epoch": 4.143240290056621,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0004685447501738353,
+      "loss": 0.5194,
+      "step": 83420
+    },
+    {
+      "epoch": 4.14373696235224,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0004685050163901858,
+      "loss": 0.5188,
+      "step": 83430
+    },
+    {
+      "epoch": 4.144233634647859,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0004684652826065362,
+      "loss": 0.4746,
+      "step": 83440
+    },
+    {
+      "epoch": 4.144730306943479,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004684255488228867,
+      "loss": 0.5332,
+      "step": 83450
+    },
+    {
+      "epoch": 4.145226979239098,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00046838581503923715,
+      "loss": 0.4914,
+      "step": 83460
+    },
+    {
+      "epoch": 4.145723651534717,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00046834608125558756,
+      "loss": 0.5398,
+      "step": 83470
+    },
+    {
+      "epoch": 4.146220323830336,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00046830634747193803,
+      "loss": 0.5198,
+      "step": 83480
+    },
+    {
+      "epoch": 4.1467169961259565,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0004682666136882885,
+      "loss": 0.5076,
+      "step": 83490
+    },
+    {
+      "epoch": 4.147213668421576,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.000468226879904639,
+      "loss": 0.5263,
+      "step": 83500
+    },
+    {
+      "epoch": 4.147710340717195,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004681871461209894,
+      "loss": 0.5078,
+      "step": 83510
+    },
+    {
+      "epoch": 4.148207013012814,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00046814741233733987,
+      "loss": 0.5411,
+      "step": 83520
+    },
+    {
+      "epoch": 4.148703685308433,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00046810767855369034,
+      "loss": 0.5165,
+      "step": 83530
+    },
+    {
+      "epoch": 4.149200357604053,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0004680679447700407,
+      "loss": 0.5239,
+      "step": 83540
+    },
+    {
+      "epoch": 4.149697029899672,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0004680282109863912,
+      "loss": 0.5162,
+      "step": 83550
+    },
+    {
+      "epoch": 4.150193702195292,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004679884772027417,
+      "loss": 0.5268,
+      "step": 83560
+    },
+    {
+      "epoch": 4.150690374490911,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00046794874341909206,
+      "loss": 0.4695,
+      "step": 83570
+    },
+    {
+      "epoch": 4.15118704678653,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00046790900963544253,
+      "loss": 0.5221,
+      "step": 83580
+    },
+    {
+      "epoch": 4.15168371908215,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00046786927585179306,
+      "loss": 0.5106,
+      "step": 83590
+    },
+    {
+      "epoch": 4.152180391377769,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0004678295420681434,
+      "loss": 0.5034,
+      "step": 83600
+    },
+    {
+      "epoch": 4.152677063673388,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004677898082844939,
+      "loss": 0.5582,
+      "step": 83610
+    },
+    {
+      "epoch": 4.153173735969007,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00046775007450084436,
+      "loss": 0.5301,
+      "step": 83620
+    },
+    {
+      "epoch": 4.153670408264627,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0004677103407171949,
+      "loss": 0.5215,
+      "step": 83630
+    },
+    {
+      "epoch": 4.154167080560247,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00046767060693354525,
+      "loss": 0.512,
+      "step": 83640
+    },
+    {
+      "epoch": 4.154663752855866,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0004676308731498957,
+      "loss": 0.5255,
+      "step": 83650
+    },
+    {
+      "epoch": 4.155160425151485,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004675911393662462,
+      "loss": 0.5147,
+      "step": 83660
+    },
+    {
+      "epoch": 4.155657097447104,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0004675514055825966,
+      "loss": 0.5296,
+      "step": 83670
+    },
+    {
+      "epoch": 4.1561537697427235,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0004675116717989471,
+      "loss": 0.5289,
+      "step": 83680
+    },
+    {
+      "epoch": 4.156650442038343,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00046747193801529755,
+      "loss": 0.5453,
+      "step": 83690
+    },
+    {
+      "epoch": 4.157147114333963,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00046743220423164797,
+      "loss": 0.5377,
+      "step": 83700
+    },
+    {
+      "epoch": 4.157643786629582,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00046739247044799844,
+      "loss": 0.4831,
+      "step": 83710
+    },
+    {
+      "epoch": 4.158140458925201,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004673527366643489,
+      "loss": 0.5165,
+      "step": 83720
+    },
+    {
+      "epoch": 4.158637131220821,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0004673130028806993,
+      "loss": 0.532,
+      "step": 83730
+    },
+    {
+      "epoch": 4.15913380351644,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004672732690970498,
+      "loss": 0.5285,
+      "step": 83740
+    },
+    {
+      "epoch": 4.159630475812059,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0004672335353134003,
+      "loss": 0.5117,
+      "step": 83750
+    },
+    {
+      "epoch": 4.160127148107678,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00046719380152975064,
+      "loss": 0.5011,
+      "step": 83760
+    },
+    {
+      "epoch": 4.1606238204032975,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0004671540677461011,
+      "loss": 0.541,
+      "step": 83770
+    },
+    {
+      "epoch": 4.161120492698918,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00046711433396245163,
+      "loss": 0.5228,
+      "step": 83780
+    },
+    {
+      "epoch": 4.161617164994537,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0004670746001788021,
+      "loss": 0.4957,
+      "step": 83790
+    },
+    {
+      "epoch": 4.162113837290156,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00046703486639515247,
+      "loss": 0.4934,
+      "step": 83800
+    },
+    {
+      "epoch": 4.162610509585775,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00046699513261150294,
+      "loss": 0.5274,
+      "step": 83810
+    },
+    {
+      "epoch": 4.1631071818813945,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00046695539882785346,
+      "loss": 0.5108,
+      "step": 83820
+    },
+    {
+      "epoch": 4.163603854177014,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00046691566504420383,
+      "loss": 0.5514,
+      "step": 83830
+    },
+    {
+      "epoch": 4.164100526472633,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0004668759312605543,
+      "loss": 0.5133,
+      "step": 83840
+    },
+    {
+      "epoch": 4.164597198768253,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00046683619747690477,
+      "loss": 0.5386,
+      "step": 83850
+    },
+    {
+      "epoch": 4.165093871063872,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0004667964636932552,
+      "loss": 0.5187,
+      "step": 83860
+    },
+    {
+      "epoch": 4.1655905433594915,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00046675672990960566,
+      "loss": 0.5123,
+      "step": 83870
+    },
+    {
+      "epoch": 4.166087215655111,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00046671699612595613,
+      "loss": 0.5165,
+      "step": 83880
+    },
+    {
+      "epoch": 4.16658388795073,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00046667726234230655,
+      "loss": 0.5761,
+      "step": 83890
+    },
+    {
+      "epoch": 4.167080560246349,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.000466637528558657,
+      "loss": 0.5479,
+      "step": 83900
+    },
+    {
+      "epoch": 4.167577232541968,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0004665977947750075,
+      "loss": 0.5462,
+      "step": 83910
+    },
+    {
+      "epoch": 4.1680739048375886,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0004665580609913579,
+      "loss": 0.5091,
+      "step": 83920
+    },
+    {
+      "epoch": 4.168570577133208,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0004665183272077084,
+      "loss": 0.5157,
+      "step": 83930
+    },
+    {
+      "epoch": 4.169067249428827,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00046647859342405885,
+      "loss": 0.5151,
+      "step": 83940
+    },
+    {
+      "epoch": 4.169563921724446,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004664388596404093,
+      "loss": 0.5114,
+      "step": 83950
+    },
+    {
+      "epoch": 4.1700605940200655,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00046639912585675974,
+      "loss": 0.5499,
+      "step": 83960
+    },
+    {
+      "epoch": 4.170557266315685,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0004663593920731102,
+      "loss": 0.5134,
+      "step": 83970
+    },
+    {
+      "epoch": 4.171053938611304,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0004663196582894607,
+      "loss": 0.53,
+      "step": 83980
+    },
+    {
+      "epoch": 4.171550610906924,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00046627992450581104,
+      "loss": 0.5353,
+      "step": 83990
+    },
+    {
+      "epoch": 4.172047283202543,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004662401907221615,
+      "loss": 0.5044,
+      "step": 84000
+    },
+    {
+      "epoch": 4.1725439554981625,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00046620045693851204,
+      "loss": 0.5158,
+      "step": 84010
+    },
+    {
+      "epoch": 4.173040627793782,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004661607231548624,
+      "loss": 0.5087,
+      "step": 84020
+    },
+    {
+      "epoch": 4.173537300089401,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0004661209893712129,
+      "loss": 0.5213,
+      "step": 84030
+    },
+    {
+      "epoch": 4.17403397238502,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00046608125558756335,
+      "loss": 0.5164,
+      "step": 84040
+    },
+    {
+      "epoch": 4.174530644680639,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.00046604152180391376,
+      "loss": 0.5075,
+      "step": 84050
+    },
+    {
+      "epoch": 4.175027316976259,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00046600178802026423,
+      "loss": 0.4927,
+      "step": 84060
+    },
+    {
+      "epoch": 4.175523989271879,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004659620542366147,
+      "loss": 0.5201,
+      "step": 84070
+    },
+    {
+      "epoch": 4.176020661567498,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004659223204529652,
+      "loss": 0.5072,
+      "step": 84080
+    },
+    {
+      "epoch": 4.176517333863117,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0004658825866693156,
+      "loss": 0.5163,
+      "step": 84090
+    },
+    {
+      "epoch": 4.177014006158736,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00046584285288566607,
+      "loss": 0.5076,
+      "step": 84100
+    },
+    {
+      "epoch": 4.177510678454356,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00046580311910201654,
+      "loss": 0.4845,
+      "step": 84110
+    },
+    {
+      "epoch": 4.178007350749975,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00046576338531836695,
+      "loss": 0.5261,
+      "step": 84120
+    },
+    {
+      "epoch": 4.178504023045594,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0004657236515347174,
+      "loss": 0.5091,
+      "step": 84130
+    },
+    {
+      "epoch": 4.179000695341214,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0004656839177510679,
+      "loss": 0.5522,
+      "step": 84140
+    },
+    {
+      "epoch": 4.179497367636833,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0004656441839674183,
+      "loss": 0.5152,
+      "step": 84150
+    },
+    {
+      "epoch": 4.179994039932453,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004656044501837688,
+      "loss": 0.5298,
+      "step": 84160
+    },
+    {
+      "epoch": 4.180490712228072,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00046556471640011926,
+      "loss": 0.514,
+      "step": 84170
+    },
+    {
+      "epoch": 4.180987384523691,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0004655249826164696,
+      "loss": 0.5026,
+      "step": 84180
+    },
+    {
+      "epoch": 4.18148405681931,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00046548524883282015,
+      "loss": 0.5257,
+      "step": 84190
+    },
+    {
+      "epoch": 4.18198072911493,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0004654455150491706,
+      "loss": 0.5155,
+      "step": 84200
+    },
+    {
+      "epoch": 4.18247740141055,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.000465405781265521,
+      "loss": 0.5175,
+      "step": 84210
+    },
+    {
+      "epoch": 4.182974073706169,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00046536604748187145,
+      "loss": 0.5337,
+      "step": 84220
+    },
+    {
+      "epoch": 4.183470746001788,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0004653263136982219,
+      "loss": 0.5357,
+      "step": 84230
+    },
+    {
+      "epoch": 4.183967418297407,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00046528657991457245,
+      "loss": 0.5226,
+      "step": 84240
+    },
+    {
+      "epoch": 4.184464090593027,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0004652468461309228,
+      "loss": 0.5126,
+      "step": 84250
+    },
+    {
+      "epoch": 4.184960762888646,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0004652071123472733,
+      "loss": 0.5091,
+      "step": 84260
+    },
+    {
+      "epoch": 4.185457435184265,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00046516737856362375,
+      "loss": 0.5189,
+      "step": 84270
+    },
+    {
+      "epoch": 4.185954107479885,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00046512764477997417,
+      "loss": 0.5327,
+      "step": 84280
+    },
+    {
+      "epoch": 4.186450779775504,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00046508791099632464,
+      "loss": 0.5098,
+      "step": 84290
+    },
+    {
+      "epoch": 4.186947452071124,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0004650481772126751,
+      "loss": 0.5159,
+      "step": 84300
+    },
+    {
+      "epoch": 4.187444124366743,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00046500844342902553,
+      "loss": 0.5371,
+      "step": 84310
+    },
+    {
+      "epoch": 4.187940796662362,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.000464968709645376,
+      "loss": 0.5184,
+      "step": 84320
+    },
+    {
+      "epoch": 4.188437468957981,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0004649289758617265,
+      "loss": 0.4931,
+      "step": 84330
+    },
+    {
+      "epoch": 4.1889341412536005,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0004648892420780769,
+      "loss": 0.5019,
+      "step": 84340
+    },
+    {
+      "epoch": 4.189430813549221,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00046484950829442736,
+      "loss": 0.5593,
+      "step": 84350
+    },
+    {
+      "epoch": 4.18992748584484,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00046480977451077783,
+      "loss": 0.5058,
+      "step": 84360
+    },
+    {
+      "epoch": 4.190424158140459,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0004647700407271282,
+      "loss": 0.526,
+      "step": 84370
+    },
+    {
+      "epoch": 4.190920830436078,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0004647303069434787,
+      "loss": 0.5303,
+      "step": 84380
+    },
+    {
+      "epoch": 4.1914175027316976,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0004646905731598292,
+      "loss": 0.5006,
+      "step": 84390
+    },
+    {
+      "epoch": 4.191914175027317,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00046465083937617966,
+      "loss": 0.5384,
+      "step": 84400
+    },
+    {
+      "epoch": 4.192410847322936,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.00046461110559253003,
+      "loss": 0.539,
+      "step": 84410
+    },
+    {
+      "epoch": 4.192907519618556,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00046457137180888055,
+      "loss": 0.5146,
+      "step": 84420
+    },
+    {
+      "epoch": 4.193404191914175,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.000464531638025231,
+      "loss": 0.5103,
+      "step": 84430
+    },
+    {
+      "epoch": 4.193900864209795,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004644919042415814,
+      "loss": 0.5426,
+      "step": 84440
+    },
+    {
+      "epoch": 4.194397536505414,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00046445217045793186,
+      "loss": 0.5138,
+      "step": 84450
+    },
+    {
+      "epoch": 4.194894208801033,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0004644124366742824,
+      "loss": 0.521,
+      "step": 84460
+    },
+    {
+      "epoch": 4.195390881096652,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00046437270289063275,
+      "loss": 0.4961,
+      "step": 84470
+    },
+    {
+      "epoch": 4.1958875533922715,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004643329691069832,
+      "loss": 0.5285,
+      "step": 84480
+    },
+    {
+      "epoch": 4.196384225687891,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0004642932353233337,
+      "loss": 0.5304,
+      "step": 84490
+    },
+    {
+      "epoch": 4.196880897983511,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0004642535015396841,
+      "loss": 0.5183,
+      "step": 84500
+    },
+    {
+      "epoch": 4.19737757027913,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0004642137677560346,
+      "loss": 0.5008,
+      "step": 84510
+    },
+    {
+      "epoch": 4.197874242574749,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00046417403397238505,
+      "loss": 0.527,
+      "step": 84520
+    },
+    {
+      "epoch": 4.1983709148703685,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0004641343001887355,
+      "loss": 0.5472,
+      "step": 84530
+    },
+    {
+      "epoch": 4.198867587165988,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00046409456640508594,
+      "loss": 0.5314,
+      "step": 84540
+    },
+    {
+      "epoch": 4.199364259461607,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004640548326214364,
+      "loss": 0.5252,
+      "step": 84550
+    },
+    {
+      "epoch": 4.199860931757226,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0004640150988377869,
+      "loss": 0.5077,
+      "step": 84560
+    },
+    {
+      "epoch": 4.200357604052846,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0004639753650541373,
+      "loss": 0.5551,
+      "step": 84570
+    },
+    {
+      "epoch": 4.2008542763484655,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00046393563127048777,
+      "loss": 0.5278,
+      "step": 84580
+    },
+    {
+      "epoch": 4.201350948644085,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00046389589748683824,
+      "loss": 0.5052,
+      "step": 84590
+    },
+    {
+      "epoch": 4.201847620939704,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0004638561637031886,
+      "loss": 0.5159,
+      "step": 84600
+    },
+    {
+      "epoch": 4.202344293235323,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00046381642991953913,
+      "loss": 0.536,
+      "step": 84610
+    },
+    {
+      "epoch": 4.202840965530942,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0004637766961358896,
+      "loss": 0.5078,
+      "step": 84620
+    },
+    {
+      "epoch": 4.203337637826562,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00046373696235223996,
+      "loss": 0.5414,
+      "step": 84630
+    },
+    {
+      "epoch": 4.203834310122182,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00046369722856859044,
+      "loss": 0.5545,
+      "step": 84640
+    },
+    {
+      "epoch": 4.204330982417801,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00046365749478494096,
+      "loss": 0.5656,
+      "step": 84650
+    },
+    {
+      "epoch": 4.20482765471342,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004636177610012913,
+      "loss": 0.5252,
+      "step": 84660
+    },
+    {
+      "epoch": 4.2053243270090395,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004635780272176418,
+      "loss": 0.5175,
+      "step": 84670
+    },
+    {
+      "epoch": 4.205820999304659,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00046353829343399227,
+      "loss": 0.5261,
+      "step": 84680
+    },
+    {
+      "epoch": 4.206317671600278,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0004634985596503428,
+      "loss": 0.5269,
+      "step": 84690
+    },
+    {
+      "epoch": 4.206814343895897,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00046345882586669315,
+      "loss": 0.5254,
+      "step": 84700
+    },
+    {
+      "epoch": 4.207311016191517,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0004634190920830436,
+      "loss": 0.5144,
+      "step": 84710
+    },
+    {
+      "epoch": 4.2078076884871365,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004633793582993941,
+      "loss": 0.5286,
+      "step": 84720
+    },
+    {
+      "epoch": 4.208304360782756,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0004633396245157445,
+      "loss": 0.5184,
+      "step": 84730
+    },
+    {
+      "epoch": 4.208801033078375,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.000463299890732095,
+      "loss": 0.5175,
+      "step": 84740
+    },
+    {
+      "epoch": 4.209297705373994,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00046326015694844546,
+      "loss": 0.4948,
+      "step": 84750
+    },
+    {
+      "epoch": 4.209794377669613,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004632204231647959,
+      "loss": 0.5161,
+      "step": 84760
+    },
+    {
+      "epoch": 4.210291049965233,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.00046318068938114635,
+      "loss": 0.5169,
+      "step": 84770
+    },
+    {
+      "epoch": 4.210787722260852,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004631409555974968,
+      "loss": 0.5402,
+      "step": 84780
+    },
+    {
+      "epoch": 4.211284394556472,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00046310122181384723,
+      "loss": 0.5178,
+      "step": 84790
+    },
+    {
+      "epoch": 4.211781066852091,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0004630614880301977,
+      "loss": 0.5004,
+      "step": 84800
+    },
+    {
+      "epoch": 4.21227773914771,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0004630217542465482,
+      "loss": 0.5539,
+      "step": 84810
+    },
+    {
+      "epoch": 4.21277441144333,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00046298202046289854,
+      "loss": 0.5342,
+      "step": 84820
+    },
+    {
+      "epoch": 4.213271083738949,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.000462942286679249,
+      "loss": 0.4905,
+      "step": 84830
+    },
+    {
+      "epoch": 4.213767756034568,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00046290255289559954,
+      "loss": 0.5354,
+      "step": 84840
+    },
+    {
+      "epoch": 4.214264428330187,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00046286281911195,
+      "loss": 0.522,
+      "step": 84850
+    },
+    {
+      "epoch": 4.214761100625807,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00046282308532830037,
+      "loss": 0.5274,
+      "step": 84860
+    },
+    {
+      "epoch": 4.215257772921427,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.00046278335154465084,
+      "loss": 0.4981,
+      "step": 84870
+    },
+    {
+      "epoch": 4.215754445217046,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00046274361776100137,
+      "loss": 0.4926,
+      "step": 84880
+    },
+    {
+      "epoch": 4.216251117512665,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00046270388397735173,
+      "loss": 0.5332,
+      "step": 84890
+    },
+    {
+      "epoch": 4.216747789808284,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004626641501937022,
+      "loss": 0.5076,
+      "step": 84900
+    },
+    {
+      "epoch": 4.217244462103904,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0004626244164100527,
+      "loss": 0.514,
+      "step": 84910
+    },
+    {
+      "epoch": 4.217741134399523,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0004625846826264031,
+      "loss": 0.5241,
+      "step": 84920
+    },
+    {
+      "epoch": 4.218237806695143,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00046254494884275356,
+      "loss": 0.5204,
+      "step": 84930
+    },
+    {
+      "epoch": 4.218734478990762,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00046250521505910403,
+      "loss": 0.5244,
+      "step": 84940
+    },
+    {
+      "epoch": 4.219231151286381,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00046246548127545445,
+      "loss": 0.5049,
+      "step": 84950
+    },
+    {
+      "epoch": 4.219727823582001,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0004624257474918049,
+      "loss": 0.5101,
+      "step": 84960
+    },
+    {
+      "epoch": 4.22022449587762,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004623860137081554,
+      "loss": 0.5321,
+      "step": 84970
+    },
+    {
+      "epoch": 4.220721168173239,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00046234627992450586,
+      "loss": 0.5455,
+      "step": 84980
+    },
+    {
+      "epoch": 4.221217840468858,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0004623065461408563,
+      "loss": 0.4932,
+      "step": 84990
+    },
+    {
+      "epoch": 4.221714512764478,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00046226681235720675,
+      "loss": 0.5315,
+      "step": 85000
+    },
+    {
+      "epoch": 4.222211185060098,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0004622270785735572,
+      "loss": 0.5236,
+      "step": 85010
+    },
+    {
+      "epoch": 4.222707857355717,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.00046218734478990764,
+      "loss": 0.5112,
+      "step": 85020
+    },
+    {
+      "epoch": 4.223204529651336,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0004621476110062581,
+      "loss": 0.513,
+      "step": 85030
+    },
+    {
+      "epoch": 4.223701201946955,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0004621078772226086,
+      "loss": 0.4923,
+      "step": 85040
+    },
+    {
+      "epoch": 4.2241978742425745,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00046206814343895895,
+      "loss": 0.5019,
+      "step": 85050
+    },
+    {
+      "epoch": 4.224694546538194,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0004620284096553095,
+      "loss": 0.4975,
+      "step": 85060
+    },
+    {
+      "epoch": 4.225191218833814,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00046198867587165994,
+      "loss": 0.5384,
+      "step": 85070
+    },
+    {
+      "epoch": 4.225687891129433,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004619489420880103,
+      "loss": 0.5359,
+      "step": 85080
+    },
+    {
+      "epoch": 4.226184563425052,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0004619092083043608,
+      "loss": 0.5235,
+      "step": 85090
+    },
+    {
+      "epoch": 4.226681235720672,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.00046186947452071125,
+      "loss": 0.5044,
+      "step": 85100
+    },
+    {
+      "epoch": 4.227177908016291,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00046182974073706167,
+      "loss": 0.5421,
+      "step": 85110
+    },
+    {
+      "epoch": 4.22767458031191,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00046179000695341214,
+      "loss": 0.5294,
+      "step": 85120
+    },
+    {
+      "epoch": 4.228171252607529,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0004617502731697626,
+      "loss": 0.5175,
+      "step": 85130
+    },
+    {
+      "epoch": 4.228667924903149,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0004617105393861131,
+      "loss": 0.5488,
+      "step": 85140
+    },
+    {
+      "epoch": 4.229164597198769,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0004616708056024635,
+      "loss": 0.4906,
+      "step": 85150
+    },
+    {
+      "epoch": 4.229661269494388,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00046163107181881397,
+      "loss": 0.5223,
+      "step": 85160
+    },
+    {
+      "epoch": 4.230157941790007,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00046159133803516444,
+      "loss": 0.5071,
+      "step": 85170
+    },
+    {
+      "epoch": 4.230654614085626,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00046155160425151486,
+      "loss": 0.5137,
+      "step": 85180
+    },
+    {
+      "epoch": 4.2311512863812455,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00046151187046786533,
+      "loss": 0.5056,
+      "step": 85190
+    },
+    {
+      "epoch": 4.231647958676865,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0004614721366842158,
+      "loss": 0.505,
+      "step": 85200
+    },
+    {
+      "epoch": 4.232144630972484,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0004614324029005662,
+      "loss": 0.5275,
+      "step": 85210
+    },
+    {
+      "epoch": 4.232641303268104,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0004613926691169167,
+      "loss": 0.5189,
+      "step": 85220
+    },
+    {
+      "epoch": 4.233137975563723,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00046135293533326716,
+      "loss": 0.5391,
+      "step": 85230
+    },
+    {
+      "epoch": 4.2336346478593425,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0004613132015496175,
+      "loss": 0.5331,
+      "step": 85240
+    },
+    {
+      "epoch": 4.234131320154962,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00046127346776596805,
+      "loss": 0.5216,
+      "step": 85250
+    },
+    {
+      "epoch": 4.234627992450581,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0004612337339823185,
+      "loss": 0.5352,
+      "step": 85260
+    },
+    {
+      "epoch": 4.2351246647462,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0004611940001986689,
+      "loss": 0.5044,
+      "step": 85270
+    },
+    {
+      "epoch": 4.235621337041819,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.00046115426641501936,
+      "loss": 0.5219,
+      "step": 85280
+    },
+    {
+      "epoch": 4.2361180093374395,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004611145326313699,
+      "loss": 0.5064,
+      "step": 85290
+    },
+    {
+      "epoch": 4.236614681633059,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00046107479884772035,
+      "loss": 0.5503,
+      "step": 85300
+    },
+    {
+      "epoch": 4.237111353928678,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0004610350650640707,
+      "loss": 0.5075,
+      "step": 85310
+    },
+    {
+      "epoch": 4.237608026224297,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0004609953312804212,
+      "loss": 0.5491,
+      "step": 85320
+    },
+    {
+      "epoch": 4.2381046985199164,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0004609555974967717,
+      "loss": 0.5304,
+      "step": 85330
+    },
+    {
+      "epoch": 4.238601370815536,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0004609158637131221,
+      "loss": 0.5417,
+      "step": 85340
+    },
+    {
+      "epoch": 4.239098043111155,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.00046087612992947255,
+      "loss": 0.5318,
+      "step": 85350
+    },
+    {
+      "epoch": 4.239594715406775,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.000460836396145823,
+      "loss": 0.5274,
+      "step": 85360
+    },
+    {
+      "epoch": 4.240091387702394,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00046079666236217343,
+      "loss": 0.5192,
+      "step": 85370
+    },
+    {
+      "epoch": 4.2405880599980135,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0004607569285785239,
+      "loss": 0.5049,
+      "step": 85380
+    },
+    {
+      "epoch": 4.241084732293633,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0004607171947948744,
+      "loss": 0.5139,
+      "step": 85390
+    },
+    {
+      "epoch": 4.241581404589252,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0004606774610112248,
+      "loss": 0.5182,
+      "step": 85400
+    },
+    {
+      "epoch": 4.242078076884871,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00046063772722757527,
+      "loss": 0.5096,
+      "step": 85410
+    },
+    {
+      "epoch": 4.24257474918049,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00046059799344392574,
+      "loss": 0.5349,
+      "step": 85420
+    },
+    {
+      "epoch": 4.24307142147611,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0004605582596602762,
+      "loss": 0.5377,
+      "step": 85430
+    },
+    {
+      "epoch": 4.24356809377173,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0004605185258766266,
+      "loss": 0.5205,
+      "step": 85440
+    },
+    {
+      "epoch": 4.244064766067349,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0004604787920929771,
+      "loss": 0.5251,
+      "step": 85450
+    },
+    {
+      "epoch": 4.244561438362968,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00046043905830932757,
+      "loss": 0.5031,
+      "step": 85460
+    },
+    {
+      "epoch": 4.245058110658587,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00046039932452567793,
+      "loss": 0.5041,
+      "step": 85470
+    },
+    {
+      "epoch": 4.245554782954207,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00046035959074202846,
+      "loss": 0.5095,
+      "step": 85480
+    },
+    {
+      "epoch": 4.246051455249826,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00046031985695837893,
+      "loss": 0.5163,
+      "step": 85490
+    },
+    {
+      "epoch": 4.246548127545445,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0004602801231747293,
+      "loss": 0.5149,
+      "step": 85500
+    },
+    {
+      "epoch": 4.247044799841065,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00046024038939107976,
+      "loss": 0.523,
+      "step": 85510
+    },
+    {
+      "epoch": 4.247541472136684,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0004602006556074303,
+      "loss": 0.5251,
+      "step": 85520
+    },
+    {
+      "epoch": 4.248038144432304,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.00046016092182378065,
+      "loss": 0.4988,
+      "step": 85530
+    },
+    {
+      "epoch": 4.248534816727923,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004601211880401311,
+      "loss": 0.4956,
+      "step": 85540
+    },
+    {
+      "epoch": 4.249031489023542,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0004600814542564816,
+      "loss": 0.5315,
+      "step": 85550
+    },
+    {
+      "epoch": 4.249528161319161,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000460041720472832,
+      "loss": 0.5234,
+      "step": 85560
+    },
+    {
+      "epoch": 4.250024833614781,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0004600019866891825,
+      "loss": 0.5087,
+      "step": 85570
+    },
+    {
+      "epoch": 4.250521505910401,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00045996225290553295,
+      "loss": 0.4927,
+      "step": 85580
+    },
+    {
+      "epoch": 4.25101817820602,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0004599225191218834,
+      "loss": 0.5326,
+      "step": 85590
+    },
+    {
+      "epoch": 4.251514850501639,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.00045988278533823384,
+      "loss": 0.4948,
+      "step": 85600
+    },
+    {
+      "epoch": 4.252011522797258,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0004598430515545843,
+      "loss": 0.521,
+      "step": 85610
+    },
+    {
+      "epoch": 4.252508195092878,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0004598033177709348,
+      "loss": 0.4986,
+      "step": 85620
+    },
+    {
+      "epoch": 4.253004867388497,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0004597635839872852,
+      "loss": 0.512,
+      "step": 85630
+    },
+    {
+      "epoch": 4.253501539684116,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0004597238502036357,
+      "loss": 0.5244,
+      "step": 85640
+    },
+    {
+      "epoch": 4.253998211979736,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00045968411641998614,
+      "loss": 0.4869,
+      "step": 85650
+    },
+    {
+      "epoch": 4.254494884275355,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004596443826363365,
+      "loss": 0.5067,
+      "step": 85660
+    },
+    {
+      "epoch": 4.254991556570975,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00045960464885268703,
+      "loss": 0.5391,
+      "step": 85670
+    },
+    {
+      "epoch": 4.255488228866594,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0004595649150690375,
+      "loss": 0.4976,
+      "step": 85680
+    },
+    {
+      "epoch": 4.255984901162213,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00045952518128538787,
+      "loss": 0.5479,
+      "step": 85690
+    },
+    {
+      "epoch": 4.256481573457832,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00045948544750173834,
+      "loss": 0.5283,
+      "step": 85700
+    },
+    {
+      "epoch": 4.2569782457534515,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.00045944571371808886,
+      "loss": 0.5109,
+      "step": 85710
+    },
+    {
+      "epoch": 4.257474918049072,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.00045940597993443934,
+      "loss": 0.5107,
+      "step": 85720
+    },
+    {
+      "epoch": 4.257971590344691,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0004593662461507897,
+      "loss": 0.5134,
+      "step": 85730
+    },
+    {
+      "epoch": 4.25846826264031,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00045932651236714017,
+      "loss": 0.5107,
+      "step": 85740
+    },
+    {
+      "epoch": 4.258964934935929,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0004592867785834907,
+      "loss": 0.5183,
+      "step": 85750
+    },
+    {
+      "epoch": 4.2594616072315485,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00045924704479984106,
+      "loss": 0.54,
+      "step": 85760
+    },
+    {
+      "epoch": 4.259958279527168,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00045920731101619153,
+      "loss": 0.5028,
+      "step": 85770
+    },
+    {
+      "epoch": 4.260454951822787,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.000459167577232542,
+      "loss": 0.5201,
+      "step": 85780
+    },
+    {
+      "epoch": 4.260951624118407,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0004591278434488924,
+      "loss": 0.4954,
+      "step": 85790
+    },
+    {
+      "epoch": 4.261448296414026,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004590881096652429,
+      "loss": 0.5304,
+      "step": 85800
+    },
+    {
+      "epoch": 4.261944968709646,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00045904837588159336,
+      "loss": 0.5535,
+      "step": 85810
+    },
+    {
+      "epoch": 4.262441641005265,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0004590086420979438,
+      "loss": 0.5143,
+      "step": 85820
+    },
+    {
+      "epoch": 4.262938313300884,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00045896890831429425,
+      "loss": 0.5171,
+      "step": 85830
+    },
+    {
+      "epoch": 4.263434985596503,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0004589291745306447,
+      "loss": 0.5081,
+      "step": 85840
+    },
+    {
+      "epoch": 4.2639316578921225,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00045888944074699514,
+      "loss": 0.5469,
+      "step": 85850
+    },
+    {
+      "epoch": 4.264428330187743,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0004588497069633456,
+      "loss": 0.5263,
+      "step": 85860
+    },
+    {
+      "epoch": 4.264925002483362,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0004588099731796961,
+      "loss": 0.5449,
+      "step": 85870
+    },
+    {
+      "epoch": 4.265421674778981,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00045877023939604655,
+      "loss": 0.5172,
+      "step": 85880
+    },
+    {
+      "epoch": 4.2659183470746,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00045873050561239697,
+      "loss": 0.5357,
+      "step": 85890
+    },
+    {
+      "epoch": 4.2664150193702195,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00045869077182874744,
+      "loss": 0.5344,
+      "step": 85900
+    },
+    {
+      "epoch": 4.266911691665839,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0004586510380450979,
+      "loss": 0.5317,
+      "step": 85910
+    },
+    {
+      "epoch": 4.267408363961458,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004586113042614483,
+      "loss": 0.5079,
+      "step": 85920
+    },
+    {
+      "epoch": 4.267905036257077,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00045857157047779875,
+      "loss": 0.4963,
+      "step": 85930
+    },
+    {
+      "epoch": 4.268401708552697,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00045853183669414927,
+      "loss": 0.5111,
+      "step": 85940
+    },
+    {
+      "epoch": 4.2688983808483165,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00045849210291049963,
+      "loss": 0.5146,
+      "step": 85950
+    },
+    {
+      "epoch": 4.269395053143936,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0004584523691268501,
+      "loss": 0.5579,
+      "step": 85960
+    },
+    {
+      "epoch": 4.269891725439555,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0004584126353432006,
+      "loss": 0.5281,
+      "step": 85970
+    },
+    {
+      "epoch": 4.270388397735174,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.000458372901559551,
+      "loss": 0.5497,
+      "step": 85980
+    },
+    {
+      "epoch": 4.270885070030793,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.00045833316777590147,
+      "loss": 0.4916,
+      "step": 85990
+    },
+    {
+      "epoch": 4.271381742326413,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00045829343399225194,
+      "loss": 0.5393,
+      "step": 86000
+    },
+    {
+      "epoch": 4.271878414622033,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00045825370020860235,
+      "loss": 0.545,
+      "step": 86010
+    },
+    {
+      "epoch": 4.272375086917652,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0004582139664249528,
+      "loss": 0.5428,
+      "step": 86020
+    },
+    {
+      "epoch": 4.272871759213271,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0004581742326413033,
+      "loss": 0.5207,
+      "step": 86030
+    },
+    {
+      "epoch": 4.2733684315088905,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00045813449885765377,
+      "loss": 0.5163,
+      "step": 86040
+    },
+    {
+      "epoch": 4.27386510380451,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0004580947650740042,
+      "loss": 0.515,
+      "step": 86050
+    },
+    {
+      "epoch": 4.274361776100129,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00045805503129035466,
+      "loss": 0.549,
+      "step": 86060
+    },
+    {
+      "epoch": 4.274858448395748,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00045801529750670513,
+      "loss": 0.535,
+      "step": 86070
+    },
+    {
+      "epoch": 4.275355120691368,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00045797556372305555,
+      "loss": 0.5278,
+      "step": 86080
+    },
+    {
+      "epoch": 4.2758517929869875,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.000457935829939406,
+      "loss": 0.5455,
+      "step": 86090
+    },
+    {
+      "epoch": 4.276348465282607,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0004578960961557565,
+      "loss": 0.5381,
+      "step": 86100
+    },
+    {
+      "epoch": 4.276845137578226,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00045785636237210685,
+      "loss": 0.526,
+      "step": 86110
+    },
+    {
+      "epoch": 4.277341809873845,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0004578166285884574,
+      "loss": 0.5111,
+      "step": 86120
+    },
+    {
+      "epoch": 4.277838482169464,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00045777689480480785,
+      "loss": 0.507,
+      "step": 86130
+    },
+    {
+      "epoch": 4.278335154465084,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0004577371610211582,
+      "loss": 0.5182,
+      "step": 86140
+    },
+    {
+      "epoch": 4.278831826760703,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0004576974272375087,
+      "loss": 0.5155,
+      "step": 86150
+    },
+    {
+      "epoch": 4.279328499056323,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0004576576934538592,
+      "loss": 0.5027,
+      "step": 86160
+    },
+    {
+      "epoch": 4.279825171351942,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004576179596702097,
+      "loss": 0.5301,
+      "step": 86170
+    },
+    {
+      "epoch": 4.280321843647561,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00045757822588656004,
+      "loss": 0.5158,
+      "step": 86180
+    },
+    {
+      "epoch": 4.280818515943181,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0004575384921029105,
+      "loss": 0.5059,
+      "step": 86190
+    },
+    {
+      "epoch": 4.2813151882388,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.000457498758319261,
+      "loss": 0.5142,
+      "step": 86200
+    },
+    {
+      "epoch": 4.281811860534419,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0004574590245356114,
+      "loss": 0.5515,
+      "step": 86210
+    },
+    {
+      "epoch": 4.282308532830038,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0004574192907519619,
+      "loss": 0.5063,
+      "step": 86220
+    },
+    {
+      "epoch": 4.282805205125658,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00045737955696831234,
+      "loss": 0.4999,
+      "step": 86230
+    },
+    {
+      "epoch": 4.283301877421278,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00045733982318466276,
+      "loss": 0.5308,
+      "step": 86240
+    },
+    {
+      "epoch": 4.283798549716897,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00045730008940101323,
+      "loss": 0.526,
+      "step": 86250
+    },
+    {
+      "epoch": 4.284295222012516,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0004572603556173637,
+      "loss": 0.5213,
+      "step": 86260
+    },
+    {
+      "epoch": 4.284791894308135,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0004572206218337141,
+      "loss": 0.539,
+      "step": 86270
+    },
+    {
+      "epoch": 4.285288566603755,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0004571808880500646,
+      "loss": 0.5036,
+      "step": 86280
+    },
+    {
+      "epoch": 4.285785238899374,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00045714115426641506,
+      "loss": 0.5239,
+      "step": 86290
+    },
+    {
+      "epoch": 4.286281911194994,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.00045710142048276543,
+      "loss": 0.5123,
+      "step": 86300
+    },
+    {
+      "epoch": 4.286778583490613,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00045706168669911595,
+      "loss": 0.5036,
+      "step": 86310
+    },
+    {
+      "epoch": 4.287275255786232,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004570219529154664,
+      "loss": 0.5138,
+      "step": 86320
+    },
+    {
+      "epoch": 4.287771928081852,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0004569822191318169,
+      "loss": 0.4928,
+      "step": 86330
+    },
+    {
+      "epoch": 4.288268600377471,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00045694248534816726,
+      "loss": 0.5393,
+      "step": 86340
+    },
+    {
+      "epoch": 4.28876527267309,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0004569027515645178,
+      "loss": 0.5234,
+      "step": 86350
+    },
+    {
+      "epoch": 4.289261944968709,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00045686301778086826,
+      "loss": 0.5304,
+      "step": 86360
+    },
+    {
+      "epoch": 4.289758617264329,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0004568232839972186,
+      "loss": 0.5225,
+      "step": 86370
+    },
+    {
+      "epoch": 4.290255289559949,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0004567835502135691,
+      "loss": 0.5119,
+      "step": 86380
+    },
+    {
+      "epoch": 4.290751961855568,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0004567438164299196,
+      "loss": 0.5613,
+      "step": 86390
+    },
+    {
+      "epoch": 4.291248634151187,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00045670408264627,
+      "loss": 0.538,
+      "step": 86400
+    },
+    {
+      "epoch": 4.291745306446806,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00045666434886262045,
+      "loss": 0.5313,
+      "step": 86410
+    },
+    {
+      "epoch": 4.2922419787424255,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0004566246150789709,
+      "loss": 0.5336,
+      "step": 86420
+    },
+    {
+      "epoch": 4.292738651038045,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00045658488129532134,
+      "loss": 0.4929,
+      "step": 86430
+    },
+    {
+      "epoch": 4.293235323333665,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0004565451475116718,
+      "loss": 0.5121,
+      "step": 86440
+    },
+    {
+      "epoch": 4.293731995629284,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0004565054137280223,
+      "loss": 0.54,
+      "step": 86450
+    },
+    {
+      "epoch": 4.294228667924903,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0004564656799443727,
+      "loss": 0.5073,
+      "step": 86460
+    },
+    {
+      "epoch": 4.2947253402205225,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00045642594616072317,
+      "loss": 0.5126,
+      "step": 86470
+    },
+    {
+      "epoch": 4.295222012516142,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00045638621237707364,
+      "loss": 0.5324,
+      "step": 86480
+    },
+    {
+      "epoch": 4.295718684811761,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0004563464785934241,
+      "loss": 0.52,
+      "step": 86490
+    },
+    {
+      "epoch": 4.29621535710738,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00045630674480977453,
+      "loss": 0.5243,
+      "step": 86500
+    },
+    {
+      "epoch": 4.296712029403,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.000456267011026125,
+      "loss": 0.513,
+      "step": 86510
+    },
+    {
+      "epoch": 4.29720870169862,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00045622727724247547,
+      "loss": 0.4888,
+      "step": 86520
+    },
+    {
+      "epoch": 4.297705373994239,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00045618754345882584,
+      "loss": 0.5306,
+      "step": 86530
+    },
+    {
+      "epoch": 4.298202046289858,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00045614780967517636,
+      "loss": 0.5291,
+      "step": 86540
+    },
+    {
+      "epoch": 4.298698718585477,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00045610807589152683,
+      "loss": 0.5164,
+      "step": 86550
+    },
+    {
+      "epoch": 4.2991953908810965,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0004560683421078772,
+      "loss": 0.5422,
+      "step": 86560
+    },
+    {
+      "epoch": 4.299692063176716,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00045602860832422767,
+      "loss": 0.5098,
+      "step": 86570
+    },
+    {
+      "epoch": 4.300188735472336,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0004559888745405782,
+      "loss": 0.5123,
+      "step": 86580
+    },
+    {
+      "epoch": 4.300685407767955,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00045594914075692855,
+      "loss": 0.4986,
+      "step": 86590
+    },
+    {
+      "epoch": 4.301182080063574,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.000455909406973279,
+      "loss": 0.5292,
+      "step": 86600
+    },
+    {
+      "epoch": 4.3016787523591935,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0004558696731896295,
+      "loss": 0.5057,
+      "step": 86610
+    },
+    {
+      "epoch": 4.302175424654813,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00045582993940598,
+      "loss": 0.5015,
+      "step": 86620
+    },
+    {
+      "epoch": 4.302672096950432,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0004557902056223304,
+      "loss": 0.5086,
+      "step": 86630
+    },
+    {
+      "epoch": 4.303168769246051,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00045575047183868086,
+      "loss": 0.5087,
+      "step": 86640
+    },
+    {
+      "epoch": 4.30366544154167,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00045571073805503133,
+      "loss": 0.4946,
+      "step": 86650
+    },
+    {
+      "epoch": 4.3041621138372905,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00045567100427138175,
+      "loss": 0.5068,
+      "step": 86660
+    },
+    {
+      "epoch": 4.30465878613291,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0004556312704877322,
+      "loss": 0.4953,
+      "step": 86670
+    },
+    {
+      "epoch": 4.305155458428529,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0004555915367040827,
+      "loss": 0.5288,
+      "step": 86680
+    },
+    {
+      "epoch": 4.305652130724148,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0004555518029204331,
+      "loss": 0.507,
+      "step": 86690
+    },
+    {
+      "epoch": 4.306148803019767,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0004555120691367836,
+      "loss": 0.5239,
+      "step": 86700
+    },
+    {
+      "epoch": 4.306645475315387,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00045547233535313405,
+      "loss": 0.5222,
+      "step": 86710
+    },
+    {
+      "epoch": 4.307142147611006,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00045543260156948447,
+      "loss": 0.5368,
+      "step": 86720
+    },
+    {
+      "epoch": 4.307638819906626,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00045539286778583494,
+      "loss": 0.5294,
+      "step": 86730
+    },
+    {
+      "epoch": 4.308135492202245,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0004553531340021854,
+      "loss": 0.5032,
+      "step": 86740
+    },
+    {
+      "epoch": 4.3086321644978645,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00045531340021853577,
+      "loss": 0.5411,
+      "step": 86750
+    },
+    {
+      "epoch": 4.309128836793484,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0004552736664348863,
+      "loss": 0.5447,
+      "step": 86760
+    },
+    {
+      "epoch": 4.309625509089103,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00045523393265123677,
+      "loss": 0.5329,
+      "step": 86770
+    },
+    {
+      "epoch": 4.310122181384722,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00045519419886758724,
+      "loss": 0.4954,
+      "step": 86780
+    },
+    {
+      "epoch": 4.310618853680341,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0004551544650839376,
+      "loss": 0.504,
+      "step": 86790
+    },
+    {
+      "epoch": 4.311115525975961,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0004551147313002881,
+      "loss": 0.5425,
+      "step": 86800
+    },
+    {
+      "epoch": 4.311612198271581,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0004550749975166386,
+      "loss": 0.5113,
+      "step": 86810
+    },
+    {
+      "epoch": 4.3121088705672,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00045503526373298896,
+      "loss": 0.5203,
+      "step": 86820
+    },
+    {
+      "epoch": 4.312605542862819,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00045499552994933943,
+      "loss": 0.5085,
+      "step": 86830
+    },
+    {
+      "epoch": 4.313102215158438,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0004549557961656899,
+      "loss": 0.5188,
+      "step": 86840
+    },
+    {
+      "epoch": 4.313598887454058,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0004549160623820403,
+      "loss": 0.5094,
+      "step": 86850
+    },
+    {
+      "epoch": 4.314095559749677,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004548763285983908,
+      "loss": 0.5057,
+      "step": 86860
+    },
+    {
+      "epoch": 4.314592232045296,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00045483659481474126,
+      "loss": 0.4869,
+      "step": 86870
+    },
+    {
+      "epoch": 4.315088904340916,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0004547968610310917,
+      "loss": 0.5005,
+      "step": 86880
+    },
+    {
+      "epoch": 4.315585576636535,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00045475712724744215,
+      "loss": 0.5301,
+      "step": 86890
+    },
+    {
+      "epoch": 4.316082248932155,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0004547173934637926,
+      "loss": 0.53,
+      "step": 86900
+    },
+    {
+      "epoch": 4.316578921227774,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00045467765968014304,
+      "loss": 0.5241,
+      "step": 86910
+    },
+    {
+      "epoch": 4.317075593523393,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0004546379258964935,
+      "loss": 0.5442,
+      "step": 86920
+    },
+    {
+      "epoch": 4.317572265819012,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.000454598192112844,
+      "loss": 0.4976,
+      "step": 86930
+    },
+    {
+      "epoch": 4.3180689381146315,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00045455845832919446,
+      "loss": 0.5249,
+      "step": 86940
+    },
+    {
+      "epoch": 4.318565610410252,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0004545187245455449,
+      "loss": 0.5007,
+      "step": 86950
+    },
+    {
+      "epoch": 4.319062282705871,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00045447899076189534,
+      "loss": 0.528,
+      "step": 86960
+    },
+    {
+      "epoch": 4.31955895500149,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0004544392569782458,
+      "loss": 0.5246,
+      "step": 86970
+    },
+    {
+      "epoch": 4.320055627297109,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004543995231945962,
+      "loss": 0.505,
+      "step": 86980
+    },
+    {
+      "epoch": 4.320552299592729,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0004543597894109467,
+      "loss": 0.4948,
+      "step": 86990
+    },
+    {
+      "epoch": 4.321048971888348,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0004543200556272972,
+      "loss": 0.5177,
+      "step": 87000
+    },
+    {
+      "epoch": 4.321545644183967,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00045428032184364754,
+      "loss": 0.5274,
+      "step": 87010
+    },
+    {
+      "epoch": 4.322042316479587,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.000454240588059998,
+      "loss": 0.5335,
+      "step": 87020
+    },
+    {
+      "epoch": 4.322538988775206,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00045420085427634854,
+      "loss": 0.5448,
+      "step": 87030
+    },
+    {
+      "epoch": 4.323035661070826,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0004541611204926989,
+      "loss": 0.5349,
+      "step": 87040
+    },
+    {
+      "epoch": 4.323532333366445,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00045412138670904937,
+      "loss": 0.4996,
+      "step": 87050
+    },
+    {
+      "epoch": 4.324029005662064,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00045408165292539984,
+      "loss": 0.516,
+      "step": 87060
+    },
+    {
+      "epoch": 4.324525677957683,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0004540419191417503,
+      "loss": 0.5177,
+      "step": 87070
+    },
+    {
+      "epoch": 4.3250223502533025,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00045400218535810073,
+      "loss": 0.5055,
+      "step": 87080
+    },
+    {
+      "epoch": 4.325519022548923,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0004539624515744512,
+      "loss": 0.5124,
+      "step": 87090
+    },
+    {
+      "epoch": 4.326015694844542,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00045392271779080167,
+      "loss": 0.5161,
+      "step": 87100
+    },
+    {
+      "epoch": 4.326512367140161,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0004538829840071521,
+      "loss": 0.5411,
+      "step": 87110
+    },
+    {
+      "epoch": 4.32700903943578,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00045384325022350256,
+      "loss": 0.5097,
+      "step": 87120
+    },
+    {
+      "epoch": 4.3275057117313995,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00045380351643985303,
+      "loss": 0.5181,
+      "step": 87130
+    },
+    {
+      "epoch": 4.328002384027019,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.00045376378265620345,
+      "loss": 0.5099,
+      "step": 87140
+    },
+    {
+      "epoch": 4.328499056322638,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0004537240488725539,
+      "loss": 0.5018,
+      "step": 87150
+    },
+    {
+      "epoch": 4.328995728618258,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0004536843150889044,
+      "loss": 0.5034,
+      "step": 87160
+    },
+    {
+      "epoch": 4.329492400913877,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00045364458130525475,
+      "loss": 0.5121,
+      "step": 87170
+    },
+    {
+      "epoch": 4.3299890732094966,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0004536048475216053,
+      "loss": 0.5178,
+      "step": 87180
+    },
+    {
+      "epoch": 4.330485745505116,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00045356511373795575,
+      "loss": 0.5198,
+      "step": 87190
+    },
+    {
+      "epoch": 4.330982417800735,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0004535253799543061,
+      "loss": 0.5295,
+      "step": 87200
+    },
+    {
+      "epoch": 4.331479090096354,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0004534856461706566,
+      "loss": 0.5136,
+      "step": 87210
+    },
+    {
+      "epoch": 4.3319757623919735,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0004534459123870071,
+      "loss": 0.5361,
+      "step": 87220
+    },
+    {
+      "epoch": 4.332472434687594,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004534061786033576,
+      "loss": 0.5476,
+      "step": 87230
+    },
+    {
+      "epoch": 4.332969106983213,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00045336644481970795,
+      "loss": 0.511,
+      "step": 87240
+    },
+    {
+      "epoch": 4.333465779278832,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0004533267110360584,
+      "loss": 0.5044,
+      "step": 87250
+    },
+    {
+      "epoch": 4.333962451574451,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00045328697725240894,
+      "loss": 0.5091,
+      "step": 87260
+    },
+    {
+      "epoch": 4.3344591238700705,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0004532472434687593,
+      "loss": 0.5155,
+      "step": 87270
+    },
+    {
+      "epoch": 4.33495579616569,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004532075096851098,
+      "loss": 0.523,
+      "step": 87280
+    },
+    {
+      "epoch": 4.335452468461309,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00045316777590146025,
+      "loss": 0.5077,
+      "step": 87290
+    },
+    {
+      "epoch": 4.335949140756928,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.00045312804211781067,
+      "loss": 0.5133,
+      "step": 87300
+    },
+    {
+      "epoch": 4.336445813052548,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.00045308830833416114,
+      "loss": 0.504,
+      "step": 87310
+    },
+    {
+      "epoch": 4.3369424853481675,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004530485745505116,
+      "loss": 0.5102,
+      "step": 87320
+    },
+    {
+      "epoch": 4.337439157643787,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.000453008840766862,
+      "loss": 0.528,
+      "step": 87330
+    },
+    {
+      "epoch": 4.337935829939406,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0004529691069832125,
+      "loss": 0.5072,
+      "step": 87340
+    },
+    {
+      "epoch": 4.338432502235025,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00045292937319956297,
+      "loss": 0.5125,
+      "step": 87350
+    },
+    {
+      "epoch": 4.338929174530644,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00045288963941591333,
+      "loss": 0.517,
+      "step": 87360
+    },
+    {
+      "epoch": 4.339425846826264,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00045284990563226386,
+      "loss": 0.5224,
+      "step": 87370
+    },
+    {
+      "epoch": 4.339922519121884,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00045281017184861433,
+      "loss": 0.5199,
+      "step": 87380
+    },
+    {
+      "epoch": 4.340419191417503,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0004527704380649648,
+      "loss": 0.5326,
+      "step": 87390
+    },
+    {
+      "epoch": 4.340915863713122,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00045273070428131516,
+      "loss": 0.5324,
+      "step": 87400
+    },
+    {
+      "epoch": 4.341412536008741,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004526909704976657,
+      "loss": 0.5096,
+      "step": 87410
+    },
+    {
+      "epoch": 4.341909208304361,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.00045265123671401616,
+      "loss": 0.526,
+      "step": 87420
+    },
+    {
+      "epoch": 4.34240588059998,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004526115029303665,
+      "loss": 0.5329,
+      "step": 87430
+    },
+    {
+      "epoch": 4.342902552895599,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.000452571769146717,
+      "loss": 0.5089,
+      "step": 87440
+    },
+    {
+      "epoch": 4.343399225191219,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004525320353630675,
+      "loss": 0.4873,
+      "step": 87450
+    },
+    {
+      "epoch": 4.3438958974868385,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0004524923015794179,
+      "loss": 0.5222,
+      "step": 87460
+    },
+    {
+      "epoch": 4.344392569782458,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00045245256779576835,
+      "loss": 0.5066,
+      "step": 87470
+    },
+    {
+      "epoch": 4.344889242078077,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0004524128340121188,
+      "loss": 0.5389,
+      "step": 87480
+    },
+    {
+      "epoch": 4.345385914373696,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00045237310022846924,
+      "loss": 0.5425,
+      "step": 87490
+    },
+    {
+      "epoch": 4.345882586669315,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0004523333664448197,
+      "loss": 0.5212,
+      "step": 87500
+    },
+    {
+      "epoch": 4.346379258964935,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004522936326611702,
+      "loss": 0.5493,
+      "step": 87510
+    },
+    {
+      "epoch": 4.346875931260554,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00045225389887752066,
+      "loss": 0.5508,
+      "step": 87520
+    },
+    {
+      "epoch": 4.347372603556174,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0004522141650938711,
+      "loss": 0.5234,
+      "step": 87530
+    },
+    {
+      "epoch": 4.347869275851793,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00045217443131022154,
+      "loss": 0.5188,
+      "step": 87540
+    },
+    {
+      "epoch": 4.348365948147412,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.000452134697526572,
+      "loss": 0.5179,
+      "step": 87550
+    },
+    {
+      "epoch": 4.348862620443032,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00045209496374292243,
+      "loss": 0.4952,
+      "step": 87560
+    },
+    {
+      "epoch": 4.349359292738651,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0004520552299592729,
+      "loss": 0.5231,
+      "step": 87570
+    },
+    {
+      "epoch": 4.34985596503427,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004520154961756234,
+      "loss": 0.5128,
+      "step": 87580
+    },
+    {
+      "epoch": 4.350352637329889,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0004519757623919738,
+      "loss": 0.4745,
+      "step": 87590
+    },
+    {
+      "epoch": 4.350849309625509,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.00045193602860832426,
+      "loss": 0.5124,
+      "step": 87600
+    },
+    {
+      "epoch": 4.351345981921129,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00045189629482467474,
+      "loss": 0.514,
+      "step": 87610
+    },
+    {
+      "epoch": 4.351842654216748,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0004518565610410251,
+      "loss": 0.5281,
+      "step": 87620
+    },
+    {
+      "epoch": 4.352339326512367,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00045181682725737557,
+      "loss": 0.5177,
+      "step": 87630
+    },
+    {
+      "epoch": 4.352835998807986,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0004517770934737261,
+      "loss": 0.5217,
+      "step": 87640
+    },
+    {
+      "epoch": 4.3533326711036056,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00045173735969007646,
+      "loss": 0.536,
+      "step": 87650
+    },
+    {
+      "epoch": 4.353829343399225,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00045169762590642693,
+      "loss": 0.5201,
+      "step": 87660
+    },
+    {
+      "epoch": 4.354326015694845,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0004516578921227774,
+      "loss": 0.531,
+      "step": 87670
+    },
+    {
+      "epoch": 4.354822687990464,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0004516181583391279,
+      "loss": 0.5379,
+      "step": 87680
+    },
+    {
+      "epoch": 4.355319360286083,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0004515784245554783,
+      "loss": 0.5195,
+      "step": 87690
+    },
+    {
+      "epoch": 4.355816032581703,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00045153869077182876,
+      "loss": 0.533,
+      "step": 87700
+    },
+    {
+      "epoch": 4.356312704877322,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00045149895698817923,
+      "loss": 0.5403,
+      "step": 87710
+    },
+    {
+      "epoch": 4.356809377172941,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00045145922320452965,
+      "loss": 0.5178,
+      "step": 87720
+    },
+    {
+      "epoch": 4.35730604946856,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0004514194894208801,
+      "loss": 0.525,
+      "step": 87730
+    },
+    {
+      "epoch": 4.35780272176418,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0004513797556372306,
+      "loss": 0.5164,
+      "step": 87740
+    },
+    {
+      "epoch": 4.3582993940598,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.000451340021853581,
+      "loss": 0.5465,
+      "step": 87750
+    },
+    {
+      "epoch": 4.358796066355419,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004513002880699315,
+      "loss": 0.5248,
+      "step": 87760
+    },
+    {
+      "epoch": 4.359292738651038,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00045126055428628195,
+      "loss": 0.5562,
+      "step": 87770
+    },
+    {
+      "epoch": 4.359789410946657,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00045122082050263237,
+      "loss": 0.5066,
+      "step": 87780
+    },
+    {
+      "epoch": 4.3602860832422765,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.00045118108671898284,
+      "loss": 0.4942,
+      "step": 87790
+    },
+    {
+      "epoch": 4.360782755537896,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0004511413529353333,
+      "loss": 0.5248,
+      "step": 87800
+    },
+    {
+      "epoch": 4.361279427833516,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004511016191516837,
+      "loss": 0.5215,
+      "step": 87810
+    },
+    {
+      "epoch": 4.361776100129135,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0004510618853680342,
+      "loss": 0.5075,
+      "step": 87820
+    },
+    {
+      "epoch": 4.362272772424754,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00045102215158438467,
+      "loss": 0.5233,
+      "step": 87830
+    },
+    {
+      "epoch": 4.3627694447203735,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00045098241780073514,
+      "loss": 0.5262,
+      "step": 87840
+    },
+    {
+      "epoch": 4.363266117015993,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004509426840170855,
+      "loss": 0.5257,
+      "step": 87850
+    },
+    {
+      "epoch": 4.363762789311612,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00045090295023343603,
+      "loss": 0.5001,
+      "step": 87860
+    },
+    {
+      "epoch": 4.364259461607231,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0004508632164497865,
+      "loss": 0.531,
+      "step": 87870
+    },
+    {
+      "epoch": 4.364756133902851,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00045082348266613687,
+      "loss": 0.5427,
+      "step": 87880
+    },
+    {
+      "epoch": 4.365252806198471,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00045078374888248734,
+      "loss": 0.5221,
+      "step": 87890
+    },
+    {
+      "epoch": 4.36574947849409,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0004507440150988378,
+      "loss": 0.5203,
+      "step": 87900
+    },
+    {
+      "epoch": 4.366246150789709,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0004507042813151882,
+      "loss": 0.5099,
+      "step": 87910
+    },
+    {
+      "epoch": 4.366742823085328,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0004506645475315387,
+      "loss": 0.5242,
+      "step": 87920
+    },
+    {
+      "epoch": 4.3672394953809475,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.00045062481374788917,
+      "loss": 0.5017,
+      "step": 87930
+    },
+    {
+      "epoch": 4.367736167676567,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0004505850799642396,
+      "loss": 0.4999,
+      "step": 87940
+    },
+    {
+      "epoch": 4.368232839972187,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00045054534618059006,
+      "loss": 0.5277,
+      "step": 87950
+    },
+    {
+      "epoch": 4.368729512267806,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00045050561239694053,
+      "loss": 0.5373,
+      "step": 87960
+    },
+    {
+      "epoch": 4.369226184563425,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.000450465878613291,
+      "loss": 0.5431,
+      "step": 87970
+    },
+    {
+      "epoch": 4.3697228568590445,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0004504261448296414,
+      "loss": 0.5276,
+      "step": 87980
+    },
+    {
+      "epoch": 4.370219529154664,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004503864110459919,
+      "loss": 0.5384,
+      "step": 87990
+    },
+    {
+      "epoch": 4.370716201450283,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00045034667726234236,
+      "loss": 0.5363,
+      "step": 88000
+    },
+    {
+      "epoch": 4.371212873745902,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0004503069434786928,
+      "loss": 0.5175,
+      "step": 88010
+    },
+    {
+      "epoch": 4.371709546041521,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00045026720969504325,
+      "loss": 0.5321,
+      "step": 88020
+    },
+    {
+      "epoch": 4.3722062183371415,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0004502274759113937,
+      "loss": 0.5007,
+      "step": 88030
+    },
+    {
+      "epoch": 4.372702890632761,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0004501877421277441,
+      "loss": 0.5155,
+      "step": 88040
+    },
+    {
+      "epoch": 4.37319956292838,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004501480083440946,
+      "loss": 0.5093,
+      "step": 88050
+    },
+    {
+      "epoch": 4.373696235223999,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0004501082745604451,
+      "loss": 0.5215,
+      "step": 88060
+    },
+    {
+      "epoch": 4.374192907519618,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.00045006854077679544,
+      "loss": 0.4951,
+      "step": 88070
+    },
+    {
+      "epoch": 4.374689579815238,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0004500288069931459,
+      "loss": 0.5061,
+      "step": 88080
+    },
+    {
+      "epoch": 4.375186252110857,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00044998907320949644,
+      "loss": 0.5246,
+      "step": 88090
+    },
+    {
+      "epoch": 4.375682924406477,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0004499493394258468,
+      "loss": 0.5417,
+      "step": 88100
+    },
+    {
+      "epoch": 4.376179596702096,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0004499096056421973,
+      "loss": 0.552,
+      "step": 88110
+    },
+    {
+      "epoch": 4.3766762689977154,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00044986987185854774,
+      "loss": 0.5474,
+      "step": 88120
+    },
+    {
+      "epoch": 4.377172941293335,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00044983013807489827,
+      "loss": 0.5339,
+      "step": 88130
+    },
+    {
+      "epoch": 4.377669613588954,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00044979040429124863,
+      "loss": 0.5036,
+      "step": 88140
+    },
+    {
+      "epoch": 4.378166285884573,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0004497506705075991,
+      "loss": 0.5204,
+      "step": 88150
+    },
+    {
+      "epoch": 4.378662958180192,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0004497109367239496,
+      "loss": 0.5123,
+      "step": 88160
+    },
+    {
+      "epoch": 4.3791596304758125,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0004496712029403,
+      "loss": 0.5498,
+      "step": 88170
+    },
+    {
+      "epoch": 4.379656302771432,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00044963146915665046,
+      "loss": 0.5111,
+      "step": 88180
+    },
+    {
+      "epoch": 4.380152975067051,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00044959173537300094,
+      "loss": 0.5254,
+      "step": 88190
+    },
+    {
+      "epoch": 4.38064964736267,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00044955200158935135,
+      "loss": 0.5242,
+      "step": 88200
+    },
+    {
+      "epoch": 4.381146319658289,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0004495122678057018,
+      "loss": 0.5093,
+      "step": 88210
+    },
+    {
+      "epoch": 4.381642991953909,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0004494725340220523,
+      "loss": 0.5439,
+      "step": 88220
+    },
+    {
+      "epoch": 4.382139664249528,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00044943280023840266,
+      "loss": 0.5357,
+      "step": 88230
+    },
+    {
+      "epoch": 4.382636336545147,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0004493930664547532,
+      "loss": 0.5,
+      "step": 88240
+    },
+    {
+      "epoch": 4.383133008840767,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00044935333267110366,
+      "loss": 0.5414,
+      "step": 88250
+    },
+    {
+      "epoch": 4.383629681136386,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0004493135988874541,
+      "loss": 0.5203,
+      "step": 88260
+    },
+    {
+      "epoch": 4.384126353432006,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0004492738651038045,
+      "loss": 0.5286,
+      "step": 88270
+    },
+    {
+      "epoch": 4.384623025727625,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.000449234131320155,
+      "loss": 0.5242,
+      "step": 88280
+    },
+    {
+      "epoch": 4.385119698023244,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0004491943975365055,
+      "loss": 0.5384,
+      "step": 88290
+    },
+    {
+      "epoch": 4.385616370318863,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00044915466375285585,
+      "loss": 0.5252,
+      "step": 88300
+    },
+    {
+      "epoch": 4.3861130426144825,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0004491149299692063,
+      "loss": 0.5113,
+      "step": 88310
+    },
+    {
+      "epoch": 4.386609714910103,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.00044907519618555685,
+      "loss": 0.5088,
+      "step": 88320
+    },
+    {
+      "epoch": 4.387106387205722,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0004490354624019072,
+      "loss": 0.5234,
+      "step": 88330
+    },
+    {
+      "epoch": 4.387603059501341,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0004489957286182577,
+      "loss": 0.488,
+      "step": 88340
+    },
+    {
+      "epoch": 4.38809973179696,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00044895599483460815,
+      "loss": 0.5555,
+      "step": 88350
+    },
+    {
+      "epoch": 4.38859640409258,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00044891626105095857,
+      "loss": 0.5122,
+      "step": 88360
+    },
+    {
+      "epoch": 4.389093076388199,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00044887652726730904,
+      "loss": 0.5252,
+      "step": 88370
+    },
+    {
+      "epoch": 4.389589748683818,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004488367934836595,
+      "loss": 0.538,
+      "step": 88380
+    },
+    {
+      "epoch": 4.390086420979438,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00044879705970000993,
+      "loss": 0.4978,
+      "step": 88390
+    },
+    {
+      "epoch": 4.390583093275057,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004487573259163604,
+      "loss": 0.5182,
+      "step": 88400
+    },
+    {
+      "epoch": 4.391079765570677,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00044871759213271087,
+      "loss": 0.5699,
+      "step": 88410
+    },
+    {
+      "epoch": 4.391576437866296,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00044867785834906134,
+      "loss": 0.5379,
+      "step": 88420
+    },
+    {
+      "epoch": 4.392073110161915,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00044863812456541176,
+      "loss": 0.5178,
+      "step": 88430
+    },
+    {
+      "epoch": 4.392569782457534,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00044859839078176223,
+      "loss": 0.5005,
+      "step": 88440
+    },
+    {
+      "epoch": 4.3930664547531535,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0004485586569981127,
+      "loss": 0.516,
+      "step": 88450
+    },
+    {
+      "epoch": 4.393563127048774,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004485189232144631,
+      "loss": 0.538,
+      "step": 88460
+    },
+    {
+      "epoch": 4.394059799344393,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0004484791894308136,
+      "loss": 0.5167,
+      "step": 88470
+    },
+    {
+      "epoch": 4.394556471640012,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00044843945564716406,
+      "loss": 0.5185,
+      "step": 88480
+    },
+    {
+      "epoch": 4.395053143935631,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0004483997218635144,
+      "loss": 0.508,
+      "step": 88490
+    },
+    {
+      "epoch": 4.3955498162312505,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004483599880798649,
+      "loss": 0.5249,
+      "step": 88500
+    },
+    {
+      "epoch": 4.39604648852687,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0004483202542962154,
+      "loss": 0.4992,
+      "step": 88510
+    },
+    {
+      "epoch": 4.396543160822489,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004482805205125658,
+      "loss": 0.5404,
+      "step": 88520
+    },
+    {
+      "epoch": 4.397039833118109,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00044824078672891626,
+      "loss": 0.5283,
+      "step": 88530
+    },
+    {
+      "epoch": 4.397536505413728,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00044820105294526673,
+      "loss": 0.4998,
+      "step": 88540
+    },
+    {
+      "epoch": 4.3980331777093475,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00044816131916161715,
+      "loss": 0.5101,
+      "step": 88550
+    },
+    {
+      "epoch": 4.398529850004967,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004481215853779676,
+      "loss": 0.5001,
+      "step": 88560
+    },
+    {
+      "epoch": 4.399026522300586,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0004480818515943181,
+      "loss": 0.5308,
+      "step": 88570
+    },
+    {
+      "epoch": 4.399523194596205,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00044804211781066856,
+      "loss": 0.5087,
+      "step": 88580
+    },
+    {
+      "epoch": 4.4000198668918244,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.000448002384027019,
+      "loss": 0.5077,
+      "step": 88590
+    },
+    {
+      "epoch": 4.400516539187445,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00044796265024336945,
+      "loss": 0.5311,
+      "step": 88600
+    },
+    {
+      "epoch": 4.401013211483064,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0004479229164597199,
+      "loss": 0.5168,
+      "step": 88610
+    },
+    {
+      "epoch": 4.401509883778683,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00044788318267607034,
+      "loss": 0.5317,
+      "step": 88620
+    },
+    {
+      "epoch": 4.402006556074302,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0004478434488924208,
+      "loss": 0.5134,
+      "step": 88630
+    },
+    {
+      "epoch": 4.4025032283699215,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004478037151087713,
+      "loss": 0.5415,
+      "step": 88640
+    },
+    {
+      "epoch": 4.402999900665541,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0004477639813251217,
+      "loss": 0.4927,
+      "step": 88650
+    },
+    {
+      "epoch": 4.40349657296116,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00044772424754147217,
+      "loss": 0.5059,
+      "step": 88660
+    },
+    {
+      "epoch": 4.40399324525678,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00044768451375782264,
+      "loss": 0.5184,
+      "step": 88670
+    },
+    {
+      "epoch": 4.404489917552399,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.000447644779974173,
+      "loss": 0.5118,
+      "step": 88680
+    },
+    {
+      "epoch": 4.4049865898480185,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00044760504619052353,
+      "loss": 0.5293,
+      "step": 88690
+    },
+    {
+      "epoch": 4.405483262143638,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.000447565312406874,
+      "loss": 0.5356,
+      "step": 88700
+    },
+    {
+      "epoch": 4.405979934439257,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00044752557862322447,
+      "loss": 0.5143,
+      "step": 88710
+    },
+    {
+      "epoch": 4.406476606734876,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00044748584483957483,
+      "loss": 0.5378,
+      "step": 88720
+    },
+    {
+      "epoch": 4.406973279030495,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00044744611105592536,
+      "loss": 0.5285,
+      "step": 88730
+    },
+    {
+      "epoch": 4.407469951326115,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00044740637727227583,
+      "loss": 0.5536,
+      "step": 88740
+    },
+    {
+      "epoch": 4.407966623621735,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0004473666434886262,
+      "loss": 0.5143,
+      "step": 88750
+    },
+    {
+      "epoch": 4.408463295917354,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00044732690970497666,
+      "loss": 0.4971,
+      "step": 88760
+    },
+    {
+      "epoch": 4.408959968212973,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00044728717592132714,
+      "loss": 0.5375,
+      "step": 88770
+    },
+    {
+      "epoch": 4.409456640508592,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00044724744213767755,
+      "loss": 0.5175,
+      "step": 88780
+    },
+    {
+      "epoch": 4.409953312804212,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.000447207708354028,
+      "loss": 0.5282,
+      "step": 88790
+    },
+    {
+      "epoch": 4.410449985099831,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0004471679745703785,
+      "loss": 0.5248,
+      "step": 88800
+    },
+    {
+      "epoch": 4.41094665739545,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0004471282407867289,
+      "loss": 0.5178,
+      "step": 88810
+    },
+    {
+      "epoch": 4.41144332969107,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0004470885070030794,
+      "loss": 0.4972,
+      "step": 88820
+    },
+    {
+      "epoch": 4.4119400019866895,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.00044704877321942986,
+      "loss": 0.5102,
+      "step": 88830
+    },
+    {
+      "epoch": 4.412436674282309,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0004470090394357803,
+      "loss": 0.5329,
+      "step": 88840
+    },
+    {
+      "epoch": 4.412933346577928,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00044696930565213074,
+      "loss": 0.5102,
+      "step": 88850
+    },
+    {
+      "epoch": 4.413430018873547,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004469295718684812,
+      "loss": 0.4967,
+      "step": 88860
+    },
+    {
+      "epoch": 4.413926691169166,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0004468898380848317,
+      "loss": 0.504,
+      "step": 88870
+    },
+    {
+      "epoch": 4.414423363464786,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0004468501043011821,
+      "loss": 0.5204,
+      "step": 88880
+    },
+    {
+      "epoch": 4.414920035760405,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004468103705175326,
+      "loss": 0.4997,
+      "step": 88890
+    },
+    {
+      "epoch": 4.415416708056025,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00044677063673388305,
+      "loss": 0.5183,
+      "step": 88900
+    },
+    {
+      "epoch": 4.415913380351644,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0004467309029502334,
+      "loss": 0.5336,
+      "step": 88910
+    },
+    {
+      "epoch": 4.416410052647263,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00044669116916658394,
+      "loss": 0.4998,
+      "step": 88920
+    },
+    {
+      "epoch": 4.416906724942883,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0004466514353829344,
+      "loss": 0.4927,
+      "step": 88930
+    },
+    {
+      "epoch": 4.417403397238502,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00044661170159928477,
+      "loss": 0.5184,
+      "step": 88940
+    },
+    {
+      "epoch": 4.417900069534121,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00044657196781563524,
+      "loss": 0.5013,
+      "step": 88950
+    },
+    {
+      "epoch": 4.41839674182974,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.00044653223403198577,
+      "loss": 0.4943,
+      "step": 88960
+    },
+    {
+      "epoch": 4.41889341412536,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00044649250024833613,
+      "loss": 0.5071,
+      "step": 88970
+    },
+    {
+      "epoch": 4.41939008642098,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0004464527664646866,
+      "loss": 0.5252,
+      "step": 88980
+    },
+    {
+      "epoch": 4.419886758716599,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00044641303268103707,
+      "loss": 0.5049,
+      "step": 88990
+    },
+    {
+      "epoch": 4.420383431012218,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0004463732988973875,
+      "loss": 0.507,
+      "step": 89000
+    },
+    {
+      "epoch": 4.420880103307837,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00044633356511373796,
+      "loss": 0.4965,
+      "step": 89010
+    },
+    {
+      "epoch": 4.4213767756034565,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00044629383133008843,
+      "loss": 0.5358,
+      "step": 89020
+    },
+    {
+      "epoch": 4.421873447899076,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0004462540975464389,
+      "loss": 0.5142,
+      "step": 89030
+    },
+    {
+      "epoch": 4.422370120194696,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004462143637627893,
+      "loss": 0.5253,
+      "step": 89040
+    },
+    {
+      "epoch": 4.422866792490315,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0004461746299791398,
+      "loss": 0.512,
+      "step": 89050
+    },
+    {
+      "epoch": 4.423363464785934,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00044613489619549026,
+      "loss": 0.5219,
+      "step": 89060
+    },
+    {
+      "epoch": 4.423860137081554,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0004460951624118407,
+      "loss": 0.5303,
+      "step": 89070
+    },
+    {
+      "epoch": 4.424356809377173,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00044605542862819115,
+      "loss": 0.5194,
+      "step": 89080
+    },
+    {
+      "epoch": 4.424853481672792,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004460156948445416,
+      "loss": 0.5159,
+      "step": 89090
+    },
+    {
+      "epoch": 4.425350153968411,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.000445975961060892,
+      "loss": 0.5096,
+      "step": 89100
+    },
+    {
+      "epoch": 4.425846826264031,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0004459362272772425,
+      "loss": 0.5184,
+      "step": 89110
+    },
+    {
+      "epoch": 4.426343498559651,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.000445896493493593,
+      "loss": 0.541,
+      "step": 89120
+    },
+    {
+      "epoch": 4.42684017085527,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00044585675970994335,
+      "loss": 0.5351,
+      "step": 89130
+    },
+    {
+      "epoch": 4.427336843150889,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0004458170259262938,
+      "loss": 0.4929,
+      "step": 89140
+    },
+    {
+      "epoch": 4.427833515446508,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00044577729214264434,
+      "loss": 0.4825,
+      "step": 89150
+    },
+    {
+      "epoch": 4.4283301877421275,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004457375583589948,
+      "loss": 0.5049,
+      "step": 89160
+    },
+    {
+      "epoch": 4.428826860037747,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0004456978245753452,
+      "loss": 0.5024,
+      "step": 89170
+    },
+    {
+      "epoch": 4.429323532333367,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00044565809079169565,
+      "loss": 0.5139,
+      "step": 89180
+    },
+    {
+      "epoch": 4.429820204628986,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0004456183570080462,
+      "loss": 0.516,
+      "step": 89190
+    },
+    {
+      "epoch": 4.430316876924605,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00044557862322439654,
+      "loss": 0.5034,
+      "step": 89200
+    },
+    {
+      "epoch": 4.4308135492202245,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.000445538889440747,
+      "loss": 0.5133,
+      "step": 89210
+    },
+    {
+      "epoch": 4.431310221515844,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0004454991556570975,
+      "loss": 0.4981,
+      "step": 89220
+    },
+    {
+      "epoch": 4.431806893811463,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004454594218734479,
+      "loss": 0.5328,
+      "step": 89230
+    },
+    {
+      "epoch": 4.432303566107082,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00044541968808979837,
+      "loss": 0.5556,
+      "step": 89240
+    },
+    {
+      "epoch": 4.432800238402702,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00044537995430614884,
+      "loss": 0.5132,
+      "step": 89250
+    },
+    {
+      "epoch": 4.4332969106983215,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00044534022052249926,
+      "loss": 0.519,
+      "step": 89260
+    },
+    {
+      "epoch": 4.433793582993941,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00044530048673884973,
+      "loss": 0.5153,
+      "step": 89270
+    },
+    {
+      "epoch": 4.43429025528956,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0004452607529552002,
+      "loss": 0.5048,
+      "step": 89280
+    },
+    {
+      "epoch": 4.434786927585179,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0004452210191715506,
+      "loss": 0.5422,
+      "step": 89290
+    },
+    {
+      "epoch": 4.4352835998807985,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0004451812853879011,
+      "loss": 0.5017,
+      "step": 89300
+    },
+    {
+      "epoch": 4.435780272176418,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00044514155160425156,
+      "loss": 0.516,
+      "step": 89310
+    },
+    {
+      "epoch": 4.436276944472038,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.00044510181782060203,
+      "loss": 0.5387,
+      "step": 89320
+    },
+    {
+      "epoch": 4.436773616767657,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004450620840369524,
+      "loss": 0.5209,
+      "step": 89330
+    },
+    {
+      "epoch": 4.437270289063276,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0004450223502533029,
+      "loss": 0.5306,
+      "step": 89340
+    },
+    {
+      "epoch": 4.4377669613588955,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0004449826164696534,
+      "loss": 0.4843,
+      "step": 89350
+    },
+    {
+      "epoch": 4.438263633654515,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00044494288268600375,
+      "loss": 0.5431,
+      "step": 89360
+    },
+    {
+      "epoch": 4.438760305950134,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0004449031489023542,
+      "loss": 0.538,
+      "step": 89370
+    },
+    {
+      "epoch": 4.439256978245753,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00044486341511870475,
+      "loss": 0.5177,
+      "step": 89380
+    },
+    {
+      "epoch": 4.439753650541372,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0004448236813350551,
+      "loss": 0.5403,
+      "step": 89390
+    },
+    {
+      "epoch": 4.4402503228369925,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0004447839475514056,
+      "loss": 0.5181,
+      "step": 89400
+    },
+    {
+      "epoch": 4.440746995132612,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00044474421376775606,
+      "loss": 0.5358,
+      "step": 89410
+    },
+    {
+      "epoch": 4.441243667428231,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0004447044799841065,
+      "loss": 0.5469,
+      "step": 89420
+    },
+    {
+      "epoch": 4.44174033972385,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00044466474620045694,
+      "loss": 0.4976,
+      "step": 89430
+    },
+    {
+      "epoch": 4.442237012019469,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004446250124168074,
+      "loss": 0.5204,
+      "step": 89440
+    },
+    {
+      "epoch": 4.442733684315089,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00044458527863315783,
+      "loss": 0.4834,
+      "step": 89450
+    },
+    {
+      "epoch": 4.443230356610708,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0004445455448495083,
+      "loss": 0.5473,
+      "step": 89460
+    },
+    {
+      "epoch": 4.443727028906328,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0004445058110658588,
+      "loss": 0.522,
+      "step": 89470
+    },
+    {
+      "epoch": 4.444223701201947,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.00044446607728220925,
+      "loss": 0.517,
+      "step": 89480
+    },
+    {
+      "epoch": 4.444720373497566,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00044442634349855966,
+      "loss": 0.5171,
+      "step": 89490
+    },
+    {
+      "epoch": 4.445217045793186,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00044438660971491014,
+      "loss": 0.5147,
+      "step": 89500
+    },
+    {
+      "epoch": 4.445713718088805,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0004443468759312606,
+      "loss": 0.5131,
+      "step": 89510
+    },
+    {
+      "epoch": 4.446210390384424,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.000444307142147611,
+      "loss": 0.5493,
+      "step": 89520
+    },
+    {
+      "epoch": 4.446707062680043,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0004442674083639615,
+      "loss": 0.5042,
+      "step": 89530
+    },
+    {
+      "epoch": 4.4472037349756635,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00044422767458031197,
+      "loss": 0.5174,
+      "step": 89540
+    },
+    {
+      "epoch": 4.447700407271283,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00044418794079666233,
+      "loss": 0.5088,
+      "step": 89550
+    },
+    {
+      "epoch": 4.448197079566902,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00044414820701301286,
+      "loss": 0.5215,
+      "step": 89560
+    },
+    {
+      "epoch": 4.448693751862521,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0004441084732293633,
+      "loss": 0.5181,
+      "step": 89570
+    },
+    {
+      "epoch": 4.44919042415814,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0004440687394457137,
+      "loss": 0.5321,
+      "step": 89580
+    },
+    {
+      "epoch": 4.44968709645376,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00044402900566206416,
+      "loss": 0.5109,
+      "step": 89590
+    },
+    {
+      "epoch": 4.450183768749379,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.00044398927187841463,
+      "loss": 0.5235,
+      "step": 89600
+    },
+    {
+      "epoch": 4.450680441044998,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00044394953809476516,
+      "loss": 0.5096,
+      "step": 89610
+    },
+    {
+      "epoch": 4.451177113340618,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0004439098043111155,
+      "loss": 0.5184,
+      "step": 89620
+    },
+    {
+      "epoch": 4.451673785636237,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.000443870070527466,
+      "loss": 0.5522,
+      "step": 89630
+    },
+    {
+      "epoch": 4.452170457931857,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00044383033674381646,
+      "loss": 0.5298,
+      "step": 89640
+    },
+    {
+      "epoch": 4.452667130227476,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0004437906029601669,
+      "loss": 0.5082,
+      "step": 89650
+    },
+    {
+      "epoch": 4.453163802523095,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00044375086917651735,
+      "loss": 0.5256,
+      "step": 89660
+    },
+    {
+      "epoch": 4.453660474818714,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0004437111353928678,
+      "loss": 0.4979,
+      "step": 89670
+    },
+    {
+      "epoch": 4.4541571471143335,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00044367140160921824,
+      "loss": 0.5271,
+      "step": 89680
+    },
+    {
+      "epoch": 4.454653819409954,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0004436316678255687,
+      "loss": 0.5414,
+      "step": 89690
+    },
+    {
+      "epoch": 4.455150491705573,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0004435919340419192,
+      "loss": 0.5237,
+      "step": 89700
+    },
+    {
+      "epoch": 4.455647164001192,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0004435522002582696,
+      "loss": 0.4943,
+      "step": 89710
+    },
+    {
+      "epoch": 4.456143836296811,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00044351246647462007,
+      "loss": 0.5289,
+      "step": 89720
+    },
+    {
+      "epoch": 4.4566405085924305,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00044347273269097054,
+      "loss": 0.5164,
+      "step": 89730
+    },
+    {
+      "epoch": 4.45713718088805,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0004434329989073209,
+      "loss": 0.4989,
+      "step": 89740
+    },
+    {
+      "epoch": 4.457633853183669,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00044339326512367143,
+      "loss": 0.5271,
+      "step": 89750
+    },
+    {
+      "epoch": 4.458130525479289,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0004433535313400219,
+      "loss": 0.504,
+      "step": 89760
+    },
+    {
+      "epoch": 4.458627197774908,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0004433137975563724,
+      "loss": 0.5256,
+      "step": 89770
+    },
+    {
+      "epoch": 4.459123870070528,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00044327406377272274,
+      "loss": 0.5199,
+      "step": 89780
+    },
+    {
+      "epoch": 4.459620542366147,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00044323432998907326,
+      "loss": 0.5165,
+      "step": 89790
+    },
+    {
+      "epoch": 4.460117214661766,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00044319459620542373,
+      "loss": 0.4881,
+      "step": 89800
+    },
+    {
+      "epoch": 4.460613886957385,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004431548624217741,
+      "loss": 0.5022,
+      "step": 89810
+    },
+    {
+      "epoch": 4.4611105592530045,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00044311512863812457,
+      "loss": 0.5424,
+      "step": 89820
+    },
+    {
+      "epoch": 4.461607231548625,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0004430753948544751,
+      "loss": 0.4923,
+      "step": 89830
+    },
+    {
+      "epoch": 4.462103903844244,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00044303566107082546,
+      "loss": 0.4948,
+      "step": 89840
+    },
+    {
+      "epoch": 4.462600576139863,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.00044299592728717593,
+      "loss": 0.5069,
+      "step": 89850
+    },
+    {
+      "epoch": 4.463097248435482,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0004429561935035264,
+      "loss": 0.5038,
+      "step": 89860
+    },
+    {
+      "epoch": 4.4635939207311015,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0004429164597198768,
+      "loss": 0.5163,
+      "step": 89870
+    },
+    {
+      "epoch": 4.464090593026721,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0004428767259362273,
+      "loss": 0.526,
+      "step": 89880
+    },
+    {
+      "epoch": 4.46458726532234,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00044283699215257776,
+      "loss": 0.5336,
+      "step": 89890
+    },
+    {
+      "epoch": 4.46508393761796,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0004427972583689282,
+      "loss": 0.5185,
+      "step": 89900
+    },
+    {
+      "epoch": 4.465580609913579,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00044275752458527865,
+      "loss": 0.494,
+      "step": 89910
+    },
+    {
+      "epoch": 4.4660772822091985,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0004427177908016291,
+      "loss": 0.5058,
+      "step": 89920
+    },
+    {
+      "epoch": 4.466573954504818,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0004426780570179796,
+      "loss": 0.5049,
+      "step": 89930
+    },
+    {
+      "epoch": 4.467070626800437,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00044263832323433,
+      "loss": 0.5247,
+      "step": 89940
+    },
+    {
+      "epoch": 4.467567299096056,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0004425985894506805,
+      "loss": 0.5206,
+      "step": 89950
+    },
+    {
+      "epoch": 4.468063971391675,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00044255885566703095,
+      "loss": 0.5468,
+      "step": 89960
+    },
+    {
+      "epoch": 4.4685606436872956,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0004425191218833813,
+      "loss": 0.5069,
+      "step": 89970
+    },
+    {
+      "epoch": 4.469057315982915,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00044247938809973184,
+      "loss": 0.5266,
+      "step": 89980
+    },
+    {
+      "epoch": 4.469553988278534,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0004424396543160823,
+      "loss": 0.4976,
+      "step": 89990
+    },
+    {
+      "epoch": 4.470050660574153,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0004423999205324327,
+      "loss": 0.5109,
+      "step": 90000
+    },
+    {
+      "epoch": 4.4705473328697725,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.00044236018674878314,
+      "loss": 0.5179,
+      "step": 90010
+    },
+    {
+      "epoch": 4.471044005165392,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.00044232045296513367,
+      "loss": 0.5047,
+      "step": 90020
+    },
+    {
+      "epoch": 4.471540677461011,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00044228071918148403,
+      "loss": 0.528,
+      "step": 90030
+    },
+    {
+      "epoch": 4.472037349756631,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0004422409853978345,
+      "loss": 0.5362,
+      "step": 90040
+    },
+    {
+      "epoch": 4.47253402205225,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.000442201251614185,
+      "loss": 0.5376,
+      "step": 90050
+    },
+    {
+      "epoch": 4.4730306943478695,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0004421615178305355,
+      "loss": 0.532,
+      "step": 90060
+    },
+    {
+      "epoch": 4.473527366643489,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.00044212178404688586,
+      "loss": 0.5163,
+      "step": 90070
+    },
+    {
+      "epoch": 4.474024038939108,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00044208205026323634,
+      "loss": 0.5136,
+      "step": 90080
+    },
+    {
+      "epoch": 4.474520711234727,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004420423164795868,
+      "loss": 0.5068,
+      "step": 90090
+    },
+    {
+      "epoch": 4.475017383530346,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0004420025826959372,
+      "loss": 0.5269,
+      "step": 90100
+    },
+    {
+      "epoch": 4.475514055825966,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004419628489122877,
+      "loss": 0.5006,
+      "step": 90110
+    },
+    {
+      "epoch": 4.476010728121586,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00044192311512863817,
+      "loss": 0.4995,
+      "step": 90120
+    },
+    {
+      "epoch": 4.476507400417205,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0004418833813449886,
+      "loss": 0.5093,
+      "step": 90130
+    },
+    {
+      "epoch": 4.477004072712824,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00044184364756133906,
+      "loss": 0.5181,
+      "step": 90140
+    },
+    {
+      "epoch": 4.477500745008443,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0004418039137776895,
+      "loss": 0.5198,
+      "step": 90150
+    },
+    {
+      "epoch": 4.477997417304063,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00044176417999403994,
+      "loss": 0.5444,
+      "step": 90160
+    },
+    {
+      "epoch": 4.478494089599682,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004417244462103904,
+      "loss": 0.5314,
+      "step": 90170
+    },
+    {
+      "epoch": 4.478990761895301,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0004416847124267409,
+      "loss": 0.4931,
+      "step": 90180
+    },
+    {
+      "epoch": 4.479487434190921,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00044164497864309125,
+      "loss": 0.536,
+      "step": 90190
+    },
+    {
+      "epoch": 4.47998410648654,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004416052448594417,
+      "loss": 0.4823,
+      "step": 90200
+    },
+    {
+      "epoch": 4.48048077878216,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00044156551107579225,
+      "loss": 0.5047,
+      "step": 90210
+    },
+    {
+      "epoch": 4.480977451077779,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0004415257772921427,
+      "loss": 0.531,
+      "step": 90220
+    },
+    {
+      "epoch": 4.481474123373398,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0004414860435084931,
+      "loss": 0.5304,
+      "step": 90230
+    },
+    {
+      "epoch": 4.481970795669017,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00044144630972484355,
+      "loss": 0.4967,
+      "step": 90240
+    },
+    {
+      "epoch": 4.482467467964637,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004414065759411941,
+      "loss": 0.531,
+      "step": 90250
+    },
+    {
+      "epoch": 4.482964140260257,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00044136684215754444,
+      "loss": 0.5198,
+      "step": 90260
+    },
+    {
+      "epoch": 4.483460812555876,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0004413271083738949,
+      "loss": 0.5248,
+      "step": 90270
+    },
+    {
+      "epoch": 4.483957484851495,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0004412873745902454,
+      "loss": 0.5253,
+      "step": 90280
+    },
+    {
+      "epoch": 4.484454157147114,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004412476408065958,
+      "loss": 0.5224,
+      "step": 90290
+    },
+    {
+      "epoch": 4.484950829442734,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00044120790702294627,
+      "loss": 0.4993,
+      "step": 90300
+    },
+    {
+      "epoch": 4.485447501738353,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00044116817323929674,
+      "loss": 0.543,
+      "step": 90310
+    },
+    {
+      "epoch": 4.485944174033972,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00044112843945564716,
+      "loss": 0.5356,
+      "step": 90320
+    },
+    {
+      "epoch": 4.486440846329591,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00044108870567199763,
+      "loss": 0.5123,
+      "step": 90330
+    },
+    {
+      "epoch": 4.486937518625211,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0004410489718883481,
+      "loss": 0.5289,
+      "step": 90340
+    },
+    {
+      "epoch": 4.487434190920831,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004410092381046985,
+      "loss": 0.5183,
+      "step": 90350
+    },
+    {
+      "epoch": 4.48793086321645,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.000440969504321049,
+      "loss": 0.5404,
+      "step": 90360
+    },
+    {
+      "epoch": 4.488427535512069,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00044092977053739946,
+      "loss": 0.5353,
+      "step": 90370
+    },
+    {
+      "epoch": 4.488924207807688,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00044089003675374993,
+      "loss": 0.5108,
+      "step": 90380
+    },
+    {
+      "epoch": 4.4894208801033075,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00044085030297010035,
+      "loss": 0.5424,
+      "step": 90390
+    },
+    {
+      "epoch": 4.489917552398927,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0004408105691864508,
+      "loss": 0.5061,
+      "step": 90400
+    },
+    {
+      "epoch": 4.490414224694547,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004407708354028013,
+      "loss": 0.5274,
+      "step": 90410
+    },
+    {
+      "epoch": 4.490910896990166,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00044073110161915166,
+      "loss": 0.5364,
+      "step": 90420
+    },
+    {
+      "epoch": 4.491407569285785,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0004406913678355022,
+      "loss": 0.5274,
+      "step": 90430
+    },
+    {
+      "epoch": 4.4919042415814046,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00044065163405185265,
+      "loss": 0.5184,
+      "step": 90440
+    },
+    {
+      "epoch": 4.492400913877024,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.000440611900268203,
+      "loss": 0.5215,
+      "step": 90450
+    },
+    {
+      "epoch": 4.492897586172643,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0004405721664845535,
+      "loss": 0.528,
+      "step": 90460
+    },
+    {
+      "epoch": 4.493394258468262,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00044053243270090396,
+      "loss": 0.5049,
+      "step": 90470
+    },
+    {
+      "epoch": 4.493890930763882,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0004404926989172544,
+      "loss": 0.5208,
+      "step": 90480
+    },
+    {
+      "epoch": 4.494387603059502,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.00044045296513360485,
+      "loss": 0.5114,
+      "step": 90490
+    },
+    {
+      "epoch": 4.494884275355121,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004404132313499553,
+      "loss": 0.5216,
+      "step": 90500
+    },
+    {
+      "epoch": 4.49538094765074,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0004403734975663058,
+      "loss": 0.5159,
+      "step": 90510
+    },
+    {
+      "epoch": 4.495877619946359,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0004403337637826562,
+      "loss": 0.5308,
+      "step": 90520
+    },
+    {
+      "epoch": 4.4963742922419785,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0004402940299990067,
+      "loss": 0.5129,
+      "step": 90530
+    },
+    {
+      "epoch": 4.496870964537598,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00044025429621535715,
+      "loss": 0.5012,
+      "step": 90540
+    },
+    {
+      "epoch": 4.497367636833218,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00044021456243170757,
+      "loss": 0.5065,
+      "step": 90550
+    },
+    {
+      "epoch": 4.497864309128837,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00044017482864805804,
+      "loss": 0.5324,
+      "step": 90560
+    },
+    {
+      "epoch": 4.498360981424456,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0004401350948644085,
+      "loss": 0.5212,
+      "step": 90570
+    },
+    {
+      "epoch": 4.4988576537200755,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00044009536108075893,
+      "loss": 0.4977,
+      "step": 90580
+    },
+    {
+      "epoch": 4.499354326015695,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0004400556272971094,
+      "loss": 0.5069,
+      "step": 90590
+    },
+    {
+      "epoch": 4.499850998311314,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00044001589351345987,
+      "loss": 0.5481,
+      "step": 90600
+    },
+    {
+      "epoch": 4.500347670606933,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00043997615972981023,
+      "loss": 0.5016,
+      "step": 90610
+    },
+    {
+      "epoch": 4.500844342902553,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00043993642594616076,
+      "loss": 0.5354,
+      "step": 90620
+    },
+    {
+      "epoch": 4.5013410151981725,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00043989669216251123,
+      "loss": 0.5329,
+      "step": 90630
+    },
+    {
+      "epoch": 4.501837687493792,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004398569583788616,
+      "loss": 0.523,
+      "step": 90640
+    },
+    {
+      "epoch": 4.502334359789411,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00043981722459521206,
+      "loss": 0.5129,
+      "step": 90650
+    },
+    {
+      "epoch": 4.50283103208503,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0004397774908115626,
+      "loss": 0.4894,
+      "step": 90660
+    },
+    {
+      "epoch": 4.503327704380649,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00043973775702791306,
+      "loss": 0.5067,
+      "step": 90670
+    },
+    {
+      "epoch": 4.503824376676269,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0004396980232442634,
+      "loss": 0.5217,
+      "step": 90680
+    },
+    {
+      "epoch": 4.504321048971889,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0004396582894606139,
+      "loss": 0.519,
+      "step": 90690
+    },
+    {
+      "epoch": 4.504817721267508,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00043961855567696437,
+      "loss": 0.4938,
+      "step": 90700
+    },
+    {
+      "epoch": 4.505314393563127,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004395788218933148,
+      "loss": 0.5179,
+      "step": 90710
+    },
+    {
+      "epoch": 4.5058110658587465,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00043953908810966526,
+      "loss": 0.4878,
+      "step": 90720
+    },
+    {
+      "epoch": 4.506307738154366,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0004394993543260157,
+      "loss": 0.5139,
+      "step": 90730
+    },
+    {
+      "epoch": 4.506804410449985,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00043945962054236614,
+      "loss": 0.5121,
+      "step": 90740
+    },
+    {
+      "epoch": 4.507301082745604,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0004394198867587166,
+      "loss": 0.4962,
+      "step": 90750
+    },
+    {
+      "epoch": 4.507797755041224,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0004393801529750671,
+      "loss": 0.4949,
+      "step": 90760
+    },
+    {
+      "epoch": 4.5082944273368435,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0004393404191914175,
+      "loss": 0.5008,
+      "step": 90770
+    },
+    {
+      "epoch": 4.508791099632463,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.000439300685407768,
+      "loss": 0.5548,
+      "step": 90780
+    },
+    {
+      "epoch": 4.509287771928082,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00043926095162411845,
+      "loss": 0.5179,
+      "step": 90790
+    },
+    {
+      "epoch": 4.509784444223701,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004392212178404689,
+      "loss": 0.518,
+      "step": 90800
+    },
+    {
+      "epoch": 4.51028111651932,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00043918148405681934,
+      "loss": 0.53,
+      "step": 90810
+    },
+    {
+      "epoch": 4.51077778881494,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0004391417502731698,
+      "loss": 0.4934,
+      "step": 90820
+    },
+    {
+      "epoch": 4.51127446111056,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0004391020164895203,
+      "loss": 0.5173,
+      "step": 90830
+    },
+    {
+      "epoch": 4.511771133406179,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00043906228270587064,
+      "loss": 0.5262,
+      "step": 90840
+    },
+    {
+      "epoch": 4.512267805701798,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.00043902254892222117,
+      "loss": 0.4987,
+      "step": 90850
+    },
+    {
+      "epoch": 4.512764477997417,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00043898281513857164,
+      "loss": 0.5014,
+      "step": 90860
+    },
+    {
+      "epoch": 4.513261150293037,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.000438943081354922,
+      "loss": 0.5175,
+      "step": 90870
+    },
+    {
+      "epoch": 4.513757822588656,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00043890334757127247,
+      "loss": 0.5199,
+      "step": 90880
+    },
+    {
+      "epoch": 4.514254494884275,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.000438863613787623,
+      "loss": 0.5094,
+      "step": 90890
+    },
+    {
+      "epoch": 4.514751167179894,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00043882388000397336,
+      "loss": 0.4907,
+      "step": 90900
+    },
+    {
+      "epoch": 4.515247839475514,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00043878414622032383,
+      "loss": 0.5088,
+      "step": 90910
+    },
+    {
+      "epoch": 4.515744511771134,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0004387444124366743,
+      "loss": 0.506,
+      "step": 90920
+    },
+    {
+      "epoch": 4.516241184066753,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0004387046786530247,
+      "loss": 0.504,
+      "step": 90930
+    },
+    {
+      "epoch": 4.516737856362372,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0004386649448693752,
+      "loss": 0.5338,
+      "step": 90940
+    },
+    {
+      "epoch": 4.517234528657991,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.00043862521108572566,
+      "loss": 0.5308,
+      "step": 90950
+    },
+    {
+      "epoch": 4.517731200953611,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00043858547730207613,
+      "loss": 0.521,
+      "step": 90960
+    },
+    {
+      "epoch": 4.51822787324923,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00043854574351842655,
+      "loss": 0.5086,
+      "step": 90970
+    },
+    {
+      "epoch": 4.518724545544849,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.000438506009734777,
+      "loss": 0.5313,
+      "step": 90980
+    },
+    {
+      "epoch": 4.519221217840469,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0004384662759511275,
+      "loss": 0.5356,
+      "step": 90990
+    },
+    {
+      "epoch": 4.519717890136088,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0004384265421674779,
+      "loss": 0.5374,
+      "step": 91000
+    },
+    {
+      "epoch": 4.520214562431708,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0004383868083838284,
+      "loss": 0.5043,
+      "step": 91010
+    },
+    {
+      "epoch": 4.520711234727327,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00043834707460017885,
+      "loss": 0.5101,
+      "step": 91020
+    },
+    {
+      "epoch": 4.521207907022946,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004383073408165292,
+      "loss": 0.5411,
+      "step": 91030
+    },
+    {
+      "epoch": 4.521704579318565,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00043826760703287974,
+      "loss": 0.5164,
+      "step": 91040
+    },
+    {
+      "epoch": 4.5222012516141845,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0004382278732492302,
+      "loss": 0.5594,
+      "step": 91050
+    },
+    {
+      "epoch": 4.522697923909805,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0004381881394655806,
+      "loss": 0.4987,
+      "step": 91060
+    },
+    {
+      "epoch": 4.523194596205424,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00043814840568193105,
+      "loss": 0.5175,
+      "step": 91070
+    },
+    {
+      "epoch": 4.523691268501043,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004381086718982816,
+      "loss": 0.5346,
+      "step": 91080
+    },
+    {
+      "epoch": 4.524187940796662,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00043806893811463194,
+      "loss": 0.5303,
+      "step": 91090
+    },
+    {
+      "epoch": 4.5246846130922815,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004380292043309824,
+      "loss": 0.5066,
+      "step": 91100
+    },
+    {
+      "epoch": 4.525181285387901,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0004379894705473329,
+      "loss": 0.5155,
+      "step": 91110
+    },
+    {
+      "epoch": 4.52567795768352,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004379497367636834,
+      "loss": 0.5041,
+      "step": 91120
+    },
+    {
+      "epoch": 4.52617462997914,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.00043791000298003377,
+      "loss": 0.4968,
+      "step": 91130
+    },
+    {
+      "epoch": 4.526671302274759,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00043787026919638424,
+      "loss": 0.5027,
+      "step": 91140
+    },
+    {
+      "epoch": 4.527167974570379,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0004378305354127347,
+      "loss": 0.5014,
+      "step": 91150
+    },
+    {
+      "epoch": 4.527664646865998,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00043779080162908513,
+      "loss": 0.5132,
+      "step": 91160
+    },
+    {
+      "epoch": 4.528161319161617,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004377510678454356,
+      "loss": 0.5287,
+      "step": 91170
+    },
+    {
+      "epoch": 4.528657991457236,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00043771133406178607,
+      "loss": 0.5218,
+      "step": 91180
+    },
+    {
+      "epoch": 4.5291546637528555,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0004376716002781365,
+      "loss": 0.532,
+      "step": 91190
+    },
+    {
+      "epoch": 4.529651336048476,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00043763186649448696,
+      "loss": 0.5352,
+      "step": 91200
+    },
+    {
+      "epoch": 4.530148008344095,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.00043759213271083743,
+      "loss": 0.5376,
+      "step": 91210
+    },
+    {
+      "epoch": 4.530644680639714,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00043755239892718785,
+      "loss": 0.5245,
+      "step": 91220
+    },
+    {
+      "epoch": 4.531141352935333,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0004375126651435383,
+      "loss": 0.535,
+      "step": 91230
+    },
+    {
+      "epoch": 4.5316380252309525,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0004374729313598888,
+      "loss": 0.5091,
+      "step": 91240
+    },
+    {
+      "epoch": 4.532134697526572,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00043743319757623926,
+      "loss": 0.5041,
+      "step": 91250
+    },
+    {
+      "epoch": 4.532631369822191,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0004373934637925897,
+      "loss": 0.5173,
+      "step": 91260
+    },
+    {
+      "epoch": 4.533128042117811,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00043735373000894015,
+      "loss": 0.5157,
+      "step": 91270
+    },
+    {
+      "epoch": 4.53362471441343,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0004373139962252906,
+      "loss": 0.5,
+      "step": 91280
+    },
+    {
+      "epoch": 4.5341213867090495,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.000437274262441641,
+      "loss": 0.532,
+      "step": 91290
+    },
+    {
+      "epoch": 4.534618059004669,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00043723452865799146,
+      "loss": 0.5365,
+      "step": 91300
+    },
+    {
+      "epoch": 4.535114731300288,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.000437194794874342,
+      "loss": 0.5145,
+      "step": 91310
+    },
+    {
+      "epoch": 4.535611403595907,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00043715506109069234,
+      "loss": 0.5214,
+      "step": 91320
+    },
+    {
+      "epoch": 4.536108075891526,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004371153273070428,
+      "loss": 0.512,
+      "step": 91330
+    },
+    {
+      "epoch": 4.5366047481871465,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0004370755935233933,
+      "loss": 0.5114,
+      "step": 91340
+    },
+    {
+      "epoch": 4.537101420482766,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004370358597397437,
+      "loss": 0.5316,
+      "step": 91350
+    },
+    {
+      "epoch": 4.537598092778385,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0004369961259560942,
+      "loss": 0.5055,
+      "step": 91360
+    },
+    {
+      "epoch": 4.538094765074004,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00043695639217244465,
+      "loss": 0.5538,
+      "step": 91370
+    },
+    {
+      "epoch": 4.5385914373696234,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00043691665838879506,
+      "loss": 0.5058,
+      "step": 91380
+    },
+    {
+      "epoch": 4.539088109665243,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00043687692460514554,
+      "loss": 0.52,
+      "step": 91390
+    },
+    {
+      "epoch": 4.539584781960862,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.000436837190821496,
+      "loss": 0.5364,
+      "step": 91400
+    },
+    {
+      "epoch": 4.540081454256482,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004367974570378465,
+      "loss": 0.5169,
+      "step": 91410
+    },
+    {
+      "epoch": 4.540578126552101,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004367577232541969,
+      "loss": 0.5479,
+      "step": 91420
+    },
+    {
+      "epoch": 4.5410747988477205,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00043671798947054737,
+      "loss": 0.5127,
+      "step": 91430
+    },
+    {
+      "epoch": 4.54157147114334,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00043667825568689784,
+      "loss": 0.5051,
+      "step": 91440
+    },
+    {
+      "epoch": 4.542068143438959,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00043663852190324826,
+      "loss": 0.5273,
+      "step": 91450
+    },
+    {
+      "epoch": 4.542564815734578,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004365987881195987,
+      "loss": 0.549,
+      "step": 91460
+    },
+    {
+      "epoch": 4.543061488030197,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0004365590543359492,
+      "loss": 0.5372,
+      "step": 91470
+    },
+    {
+      "epoch": 4.5435581603258175,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00043651932055229956,
+      "loss": 0.4916,
+      "step": 91480
+    },
+    {
+      "epoch": 4.544054832621437,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0004364795867686501,
+      "loss": 0.5001,
+      "step": 91490
+    },
+    {
+      "epoch": 4.544551504917056,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00043643985298500056,
+      "loss": 0.5066,
+      "step": 91500
+    },
+    {
+      "epoch": 4.545048177212675,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0004364001192013509,
+      "loss": 0.481,
+      "step": 91510
+    },
+    {
+      "epoch": 4.545544849508294,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0004363603854177014,
+      "loss": 0.5047,
+      "step": 91520
+    },
+    {
+      "epoch": 4.546041521803914,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0004363206516340519,
+      "loss": 0.5525,
+      "step": 91530
+    },
+    {
+      "epoch": 4.546538194099533,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0004362809178504023,
+      "loss": 0.52,
+      "step": 91540
+    },
+    {
+      "epoch": 4.547034866395153,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.00043624118406675275,
+      "loss": 0.4791,
+      "step": 91550
+    },
+    {
+      "epoch": 4.547531538690772,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0004362014502831032,
+      "loss": 0.5197,
+      "step": 91560
+    },
+    {
+      "epoch": 4.548028210986391,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0004361617164994537,
+      "loss": 0.4996,
+      "step": 91570
+    },
+    {
+      "epoch": 4.548524883282011,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0004361219827158041,
+      "loss": 0.5043,
+      "step": 91580
+    },
+    {
+      "epoch": 4.54902155557763,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0004360822489321546,
+      "loss": 0.5015,
+      "step": 91590
+    },
+    {
+      "epoch": 4.549518227873249,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00043604251514850505,
+      "loss": 0.5149,
+      "step": 91600
+    },
+    {
+      "epoch": 4.550014900168868,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.00043600278136485547,
+      "loss": 0.5205,
+      "step": 91610
+    },
+    {
+      "epoch": 4.550511572464488,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00043596304758120594,
+      "loss": 0.5549,
+      "step": 91620
+    },
+    {
+      "epoch": 4.551008244760107,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0004359233137975564,
+      "loss": 0.4934,
+      "step": 91630
+    },
+    {
+      "epoch": 4.551504917055727,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00043588358001390683,
+      "loss": 0.5018,
+      "step": 91640
+    },
+    {
+      "epoch": 4.552001589351346,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004358438462302573,
+      "loss": 0.5088,
+      "step": 91650
+    },
+    {
+      "epoch": 4.552498261646965,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0004358041124466078,
+      "loss": 0.5304,
+      "step": 91660
+    },
+    {
+      "epoch": 4.552994933942585,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00043576437866295814,
+      "loss": 0.5237,
+      "step": 91670
+    },
+    {
+      "epoch": 4.553491606238204,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00043572464487930866,
+      "loss": 0.4961,
+      "step": 91680
+    },
+    {
+      "epoch": 4.553988278533823,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00043568491109565913,
+      "loss": 0.507,
+      "step": 91690
+    },
+    {
+      "epoch": 4.554484950829442,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0004356451773120096,
+      "loss": 0.493,
+      "step": 91700
+    },
+    {
+      "epoch": 4.554981623125062,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00043560544352835997,
+      "loss": 0.5198,
+      "step": 91710
+    },
+    {
+      "epoch": 4.555478295420682,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0004355657097447105,
+      "loss": 0.4914,
+      "step": 91720
+    },
+    {
+      "epoch": 4.555974967716301,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00043552597596106097,
+      "loss": 0.5304,
+      "step": 91730
+    },
+    {
+      "epoch": 4.55647164001192,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.00043548624217741133,
+      "loss": 0.5141,
+      "step": 91740
+    },
+    {
+      "epoch": 4.556968312307539,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0004354465083937618,
+      "loss": 0.5095,
+      "step": 91750
+    },
+    {
+      "epoch": 4.5574649846031585,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0004354067746101123,
+      "loss": 0.5133,
+      "step": 91760
+    },
+    {
+      "epoch": 4.557961656898778,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004353670408264627,
+      "loss": 0.4957,
+      "step": 91770
+    },
+    {
+      "epoch": 4.558458329194398,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00043532730704281316,
+      "loss": 0.5412,
+      "step": 91780
+    },
+    {
+      "epoch": 4.558955001490017,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.00043528757325916363,
+      "loss": 0.5023,
+      "step": 91790
+    },
+    {
+      "epoch": 4.559451673785636,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00043524783947551405,
+      "loss": 0.5153,
+      "step": 91800
+    },
+    {
+      "epoch": 4.5599483460812555,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004352081056918645,
+      "loss": 0.5302,
+      "step": 91810
+    },
+    {
+      "epoch": 4.560445018376875,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.000435168371908215,
+      "loss": 0.5443,
+      "step": 91820
+    },
+    {
+      "epoch": 4.560941690672494,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0004351286381245654,
+      "loss": 0.5454,
+      "step": 91830
+    },
+    {
+      "epoch": 4.561438362968113,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0004350889043409159,
+      "loss": 0.5493,
+      "step": 91840
+    },
+    {
+      "epoch": 4.561935035263733,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00043504917055726635,
+      "loss": 0.5348,
+      "step": 91850
+    },
+    {
+      "epoch": 4.562431707559353,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0004350094367736168,
+      "loss": 0.5078,
+      "step": 91860
+    },
+    {
+      "epoch": 4.562928379854972,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00043496970298996724,
+      "loss": 0.5058,
+      "step": 91870
+    },
+    {
+      "epoch": 4.563425052150591,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0004349299692063177,
+      "loss": 0.509,
+      "step": 91880
+    },
+    {
+      "epoch": 4.56392172444621,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0004348902354226682,
+      "loss": 0.5261,
+      "step": 91890
+    },
+    {
+      "epoch": 4.5644183967418295,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00043485050163901854,
+      "loss": 0.5089,
+      "step": 91900
+    },
+    {
+      "epoch": 4.564915069037449,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00043481076785536907,
+      "loss": 0.5048,
+      "step": 91910
+    },
+    {
+      "epoch": 4.565411741333069,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00043477103407171954,
+      "loss": 0.5006,
+      "step": 91920
+    },
+    {
+      "epoch": 4.565908413628688,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004347313002880699,
+      "loss": 0.4904,
+      "step": 91930
+    },
+    {
+      "epoch": 4.566405085924307,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0004346915665044204,
+      "loss": 0.5116,
+      "step": 91940
+    },
+    {
+      "epoch": 4.5669017582199265,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0004346518327207709,
+      "loss": 0.5094,
+      "step": 91950
+    },
+    {
+      "epoch": 4.567398430515546,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00043461209893712126,
+      "loss": 0.4969,
+      "step": 91960
+    },
+    {
+      "epoch": 4.567895102811165,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00043457236515347174,
+      "loss": 0.512,
+      "step": 91970
+    },
+    {
+      "epoch": 4.568391775106784,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004345326313698222,
+      "loss": 0.5013,
+      "step": 91980
+    },
+    {
+      "epoch": 4.568888447402404,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0004344928975861726,
+      "loss": 0.504,
+      "step": 91990
+    },
+    {
+      "epoch": 4.5693851196980235,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004344531638025231,
+      "loss": 0.5093,
+      "step": 92000
+    },
+    {
+      "epoch": 4.569881791993643,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00043441343001887357,
+      "loss": 0.5301,
+      "step": 92010
+    },
+    {
+      "epoch": 4.570378464289262,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00043437369623522404,
+      "loss": 0.5294,
+      "step": 92020
+    },
+    {
+      "epoch": 4.570875136584881,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00043433396245157446,
+      "loss": 0.501,
+      "step": 92030
+    },
+    {
+      "epoch": 4.5713718088805,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0004342942286679249,
+      "loss": 0.5225,
+      "step": 92040
+    },
+    {
+      "epoch": 4.57186848117612,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0004342544948842754,
+      "loss": 0.5365,
+      "step": 92050
+    },
+    {
+      "epoch": 4.57236515347174,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004342147611006258,
+      "loss": 0.5126,
+      "step": 92060
+    },
+    {
+      "epoch": 4.572861825767359,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0004341750273169763,
+      "loss": 0.5127,
+      "step": 92070
+    },
+    {
+      "epoch": 4.573358498062978,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00043413529353332676,
+      "loss": 0.5216,
+      "step": 92080
+    },
+    {
+      "epoch": 4.5738551703585975,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0004340955597496772,
+      "loss": 0.4977,
+      "step": 92090
+    },
+    {
+      "epoch": 4.574351842654217,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00043405582596602765,
+      "loss": 0.5137,
+      "step": 92100
+    },
+    {
+      "epoch": 4.574848514949836,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004340160921823781,
+      "loss": 0.5194,
+      "step": 92110
+    },
+    {
+      "epoch": 4.575345187245455,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0004339763583987285,
+      "loss": 0.5342,
+      "step": 92120
+    },
+    {
+      "epoch": 4.575841859541075,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.000433936624615079,
+      "loss": 0.5041,
+      "step": 92130
+    },
+    {
+      "epoch": 4.5763385318366945,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0004338968908314295,
+      "loss": 0.5216,
+      "step": 92140
+    },
+    {
+      "epoch": 4.576835204132314,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00043385715704777995,
+      "loss": 0.494,
+      "step": 92150
+    },
+    {
+      "epoch": 4.577331876427933,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0004338174232641303,
+      "loss": 0.5261,
+      "step": 92160
+    },
+    {
+      "epoch": 4.577828548723552,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0004337776894804808,
+      "loss": 0.5084,
+      "step": 92170
+    },
+    {
+      "epoch": 4.578325221019171,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0004337379556968313,
+      "loss": 0.5247,
+      "step": 92180
+    },
+    {
+      "epoch": 4.578821893314791,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.00043369822191318167,
+      "loss": 0.5207,
+      "step": 92190
+    },
+    {
+      "epoch": 4.579318565610411,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00043365848812953214,
+      "loss": 0.5027,
+      "step": 92200
+    },
+    {
+      "epoch": 4.57981523790603,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004336187543458826,
+      "loss": 0.5522,
+      "step": 92210
+    },
+    {
+      "epoch": 4.580311910201649,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00043357902056223303,
+      "loss": 0.5035,
+      "step": 92220
+    },
+    {
+      "epoch": 4.580808582497268,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0004335392867785835,
+      "loss": 0.5145,
+      "step": 92230
+    },
+    {
+      "epoch": 4.581305254792888,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.000433499552994934,
+      "loss": 0.5112,
+      "step": 92240
+    },
+    {
+      "epoch": 4.581801927088507,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0004334598192112844,
+      "loss": 0.5067,
+      "step": 92250
+    },
+    {
+      "epoch": 4.582298599384126,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00043342008542763486,
+      "loss": 0.5239,
+      "step": 92260
+    },
+    {
+      "epoch": 4.582795271679745,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00043338035164398533,
+      "loss": 0.5027,
+      "step": 92270
+    },
+    {
+      "epoch": 4.583291943975365,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00043334061786033575,
+      "loss": 0.48,
+      "step": 92280
+    },
+    {
+      "epoch": 4.583788616270985,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0004333008840766862,
+      "loss": 0.5186,
+      "step": 92290
+    },
+    {
+      "epoch": 4.584285288566604,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0004332611502930367,
+      "loss": 0.5166,
+      "step": 92300
+    },
+    {
+      "epoch": 4.584781960862223,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00043322141650938717,
+      "loss": 0.5059,
+      "step": 92310
+    },
+    {
+      "epoch": 4.585278633157842,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0004331816827257376,
+      "loss": 0.5462,
+      "step": 92320
+    },
+    {
+      "epoch": 4.585775305453462,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00043314194894208805,
+      "loss": 0.4904,
+      "step": 92330
+    },
+    {
+      "epoch": 4.586271977749081,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004331022151584385,
+      "loss": 0.5082,
+      "step": 92340
+    },
+    {
+      "epoch": 4.5867686500447,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0004330624813747889,
+      "loss": 0.5404,
+      "step": 92350
+    },
+    {
+      "epoch": 4.58726532234032,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0004330227475911394,
+      "loss": 0.5415,
+      "step": 92360
+    },
+    {
+      "epoch": 4.587761994635939,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0004329830138074899,
+      "loss": 0.5209,
+      "step": 92370
+    },
+    {
+      "epoch": 4.588258666931559,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00043294328002384025,
+      "loss": 0.5144,
+      "step": 92380
+    },
+    {
+      "epoch": 4.588755339227178,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0004329035462401907,
+      "loss": 0.4951,
+      "step": 92390
+    },
+    {
+      "epoch": 4.589252011522797,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004328638124565412,
+      "loss": 0.506,
+      "step": 92400
+    },
+    {
+      "epoch": 4.589748683818416,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0004328240786728916,
+      "loss": 0.4986,
+      "step": 92410
+    },
+    {
+      "epoch": 4.5902453561140355,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0004327843448892421,
+      "loss": 0.5077,
+      "step": 92420
+    },
+    {
+      "epoch": 4.590742028409656,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00043274461110559255,
+      "loss": 0.5016,
+      "step": 92430
+    },
+    {
+      "epoch": 4.591238700705275,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.00043270487732194297,
+      "loss": 0.5008,
+      "step": 92440
+    },
+    {
+      "epoch": 4.591735373000894,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00043266514353829344,
+      "loss": 0.5352,
+      "step": 92450
+    },
+    {
+      "epoch": 4.592232045296513,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0004326254097546439,
+      "loss": 0.5004,
+      "step": 92460
+    },
+    {
+      "epoch": 4.5927287175921325,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004325856759709944,
+      "loss": 0.5296,
+      "step": 92470
+    },
+    {
+      "epoch": 4.593225389887752,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0004325459421873448,
+      "loss": 0.5047,
+      "step": 92480
+    },
+    {
+      "epoch": 4.593722062183371,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00043250620840369527,
+      "loss": 0.5274,
+      "step": 92490
+    },
+    {
+      "epoch": 4.594218734478991,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00043246647462004574,
+      "loss": 0.5117,
+      "step": 92500
+    },
+    {
+      "epoch": 4.59471540677461,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00043242674083639616,
+      "loss": 0.5121,
+      "step": 92510
+    },
+    {
+      "epoch": 4.5952120790702295,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00043238700705274663,
+      "loss": 0.5202,
+      "step": 92520
+    },
+    {
+      "epoch": 4.595708751365849,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0004323472732690971,
+      "loss": 0.5194,
+      "step": 92530
+    },
+    {
+      "epoch": 4.596205423661468,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00043230753948544746,
+      "loss": 0.5228,
+      "step": 92540
+    },
+    {
+      "epoch": 4.596702095957087,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.000432267805701798,
+      "loss": 0.5171,
+      "step": 92550
+    },
+    {
+      "epoch": 4.5971987682527065,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00043222807191814846,
+      "loss": 0.4623,
+      "step": 92560
+    },
+    {
+      "epoch": 4.597695440548327,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0004321883381344988,
+      "loss": 0.5129,
+      "step": 92570
+    },
+    {
+      "epoch": 4.598192112843946,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0004321486043508493,
+      "loss": 0.5342,
+      "step": 92580
+    },
+    {
+      "epoch": 4.598688785139565,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0004321088705671998,
+      "loss": 0.523,
+      "step": 92590
+    },
+    {
+      "epoch": 4.599185457435184,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004320691367835503,
+      "loss": 0.5319,
+      "step": 92600
+    },
+    {
+      "epoch": 4.5996821297308035,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00043202940299990066,
+      "loss": 0.5162,
+      "step": 92610
+    },
+    {
+      "epoch": 4.600178802026423,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0004319896692162511,
+      "loss": 0.486,
+      "step": 92620
+    },
+    {
+      "epoch": 4.600675474322042,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00043194993543260165,
+      "loss": 0.5066,
+      "step": 92630
+    },
+    {
+      "epoch": 4.601172146617662,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.000431910201648952,
+      "loss": 0.5231,
+      "step": 92640
+    },
+    {
+      "epoch": 4.601668818913281,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0004318704678653025,
+      "loss": 0.5092,
+      "step": 92650
+    },
+    {
+      "epoch": 4.6021654912089005,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00043183073408165296,
+      "loss": 0.5104,
+      "step": 92660
+    },
+    {
+      "epoch": 4.60266216350452,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0004317910002980034,
+      "loss": 0.4988,
+      "step": 92670
+    },
+    {
+      "epoch": 4.603158835800139,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00043175126651435385,
+      "loss": 0.5284,
+      "step": 92680
+    },
+    {
+      "epoch": 4.603655508095758,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0004317115327307043,
+      "loss": 0.5235,
+      "step": 92690
+    },
+    {
+      "epoch": 4.604152180391377,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00043167179894705474,
+      "loss": 0.5051,
+      "step": 92700
+    },
+    {
+      "epoch": 4.6046488526869975,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0004316320651634052,
+      "loss": 0.4786,
+      "step": 92710
+    },
+    {
+      "epoch": 4.605145524982617,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0004315923313797557,
+      "loss": 0.4958,
+      "step": 92720
+    },
+    {
+      "epoch": 4.605642197278236,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00043155259759610604,
+      "loss": 0.5089,
+      "step": 92730
+    },
+    {
+      "epoch": 4.606138869573855,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00043151286381245657,
+      "loss": 0.5285,
+      "step": 92740
+    },
+    {
+      "epoch": 4.606635541869474,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00043147313002880704,
+      "loss": 0.5246,
+      "step": 92750
+    },
+    {
+      "epoch": 4.607132214165094,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004314333962451575,
+      "loss": 0.5399,
+      "step": 92760
+    },
+    {
+      "epoch": 4.607628886460713,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00043139366246150787,
+      "loss": 0.5304,
+      "step": 92770
+    },
+    {
+      "epoch": 4.608125558756333,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0004313539286778584,
+      "loss": 0.5089,
+      "step": 92780
+    },
+    {
+      "epoch": 4.608622231051952,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00043131419489420887,
+      "loss": 0.5307,
+      "step": 92790
+    },
+    {
+      "epoch": 4.6091189033475715,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00043127446111055923,
+      "loss": 0.5138,
+      "step": 92800
+    },
+    {
+      "epoch": 4.609615575643191,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0004312347273269097,
+      "loss": 0.5108,
+      "step": 92810
+    },
+    {
+      "epoch": 4.61011224793881,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00043119499354326023,
+      "loss": 0.4865,
+      "step": 92820
+    },
+    {
+      "epoch": 4.610608920234429,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0004311552597596106,
+      "loss": 0.5165,
+      "step": 92830
+    },
+    {
+      "epoch": 4.611105592530048,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.00043111552597596106,
+      "loss": 0.491,
+      "step": 92840
+    },
+    {
+      "epoch": 4.6116022648256685,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00043107579219231153,
+      "loss": 0.5152,
+      "step": 92850
+    },
+    {
+      "epoch": 4.612098937121288,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00043103605840866195,
+      "loss": 0.5158,
+      "step": 92860
+    },
+    {
+      "epoch": 4.612595609416907,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0004309963246250124,
+      "loss": 0.5234,
+      "step": 92870
+    },
+    {
+      "epoch": 4.613092281712526,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0004309565908413629,
+      "loss": 0.5254,
+      "step": 92880
+    },
+    {
+      "epoch": 4.613588954008145,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00043091685705771337,
+      "loss": 0.497,
+      "step": 92890
+    },
+    {
+      "epoch": 4.614085626303765,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004308771232740638,
+      "loss": 0.5365,
+      "step": 92900
+    },
+    {
+      "epoch": 4.614582298599384,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00043083738949041425,
+      "loss": 0.5452,
+      "step": 92910
+    },
+    {
+      "epoch": 4.615078970895004,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0004307976557067647,
+      "loss": 0.5492,
+      "step": 92920
+    },
+    {
+      "epoch": 4.615575643190623,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00043075792192311514,
+      "loss": 0.5096,
+      "step": 92930
+    },
+    {
+      "epoch": 4.616072315486242,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0004307181881394656,
+      "loss": 0.5493,
+      "step": 92940
+    },
+    {
+      "epoch": 4.616568987781862,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0004306784543558161,
+      "loss": 0.5185,
+      "step": 92950
+    },
+    {
+      "epoch": 4.617065660077481,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0004306387205721665,
+      "loss": 0.5305,
+      "step": 92960
+    },
+    {
+      "epoch": 4.6175623323731,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.000430598986788517,
+      "loss": 0.5127,
+      "step": 92970
+    },
+    {
+      "epoch": 4.618059004668719,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00043055925300486745,
+      "loss": 0.528,
+      "step": 92980
+    },
+    {
+      "epoch": 4.6185556769643386,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004305195192212178,
+      "loss": 0.5339,
+      "step": 92990
+    },
+    {
+      "epoch": 4.619052349259959,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0004304797854375683,
+      "loss": 0.5293,
+      "step": 93000
+    },
+    {
+      "epoch": 4.619549021555578,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004304400516539188,
+      "loss": 0.5201,
+      "step": 93010
+    },
+    {
+      "epoch": 4.620045693851197,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00043040031787026917,
+      "loss": 0.505,
+      "step": 93020
+    },
+    {
+      "epoch": 4.620542366146816,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00043036058408661964,
+      "loss": 0.5059,
+      "step": 93030
+    },
+    {
+      "epoch": 4.621039038442436,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0004303208503029701,
+      "loss": 0.5123,
+      "step": 93040
+    },
+    {
+      "epoch": 4.621535710738055,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00043028111651932064,
+      "loss": 0.5062,
+      "step": 93050
+    },
+    {
+      "epoch": 4.622032383033674,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.000430241382735671,
+      "loss": 0.5308,
+      "step": 93060
+    },
+    {
+      "epoch": 4.622529055329293,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00043020164895202147,
+      "loss": 0.517,
+      "step": 93070
+    },
+    {
+      "epoch": 4.623025727624913,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00043016191516837194,
+      "loss": 0.5132,
+      "step": 93080
+    },
+    {
+      "epoch": 4.623522399920533,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00043012218138472236,
+      "loss": 0.5055,
+      "step": 93090
+    },
+    {
+      "epoch": 4.624019072216152,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00043008244760107283,
+      "loss": 0.4954,
+      "step": 93100
+    },
+    {
+      "epoch": 4.624515744511771,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0004300427138174233,
+      "loss": 0.4991,
+      "step": 93110
+    },
+    {
+      "epoch": 4.62501241680739,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0004300029800337737,
+      "loss": 0.5172,
+      "step": 93120
+    },
+    {
+      "epoch": 4.6255090891030095,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004299632462501242,
+      "loss": 0.5386,
+      "step": 93130
+    },
+    {
+      "epoch": 4.626005761398629,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00042992351246647466,
+      "loss": 0.5271,
+      "step": 93140
+    },
+    {
+      "epoch": 4.626502433694249,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0004298837786828251,
+      "loss": 0.5245,
+      "step": 93150
+    },
+    {
+      "epoch": 4.626999105989868,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.00042984404489917555,
+      "loss": 0.541,
+      "step": 93160
+    },
+    {
+      "epoch": 4.627495778285487,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.000429804311115526,
+      "loss": 0.4988,
+      "step": 93170
+    },
+    {
+      "epoch": 4.6279924505811065,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004297645773318764,
+      "loss": 0.5171,
+      "step": 93180
+    },
+    {
+      "epoch": 4.628489122876726,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0004297248435482269,
+      "loss": 0.5038,
+      "step": 93190
+    },
+    {
+      "epoch": 4.628985795172345,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004296851097645774,
+      "loss": 0.4955,
+      "step": 93200
+    },
+    {
+      "epoch": 4.629482467467964,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00042964537598092785,
+      "loss": 0.5067,
+      "step": 93210
+    },
+    {
+      "epoch": 4.629979139763584,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004296056421972782,
+      "loss": 0.5226,
+      "step": 93220
+    },
+    {
+      "epoch": 4.6304758120592036,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00042956590841362874,
+      "loss": 0.5224,
+      "step": 93230
+    },
+    {
+      "epoch": 4.630972484354823,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004295261746299792,
+      "loss": 0.5154,
+      "step": 93240
+    },
+    {
+      "epoch": 4.631469156650442,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0004294864408463296,
+      "loss": 0.5042,
+      "step": 93250
+    },
+    {
+      "epoch": 4.631965828946061,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00042944670706268005,
+      "loss": 0.5015,
+      "step": 93260
+    },
+    {
+      "epoch": 4.6324625012416805,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0004294069732790305,
+      "loss": 0.4962,
+      "step": 93270
+    },
+    {
+      "epoch": 4.6329591735373,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00042936723949538094,
+      "loss": 0.5325,
+      "step": 93280
+    },
+    {
+      "epoch": 4.63345584583292,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0004293275057117314,
+      "loss": 0.5356,
+      "step": 93290
+    },
+    {
+      "epoch": 4.633952518128539,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0004292877719280819,
+      "loss": 0.5387,
+      "step": 93300
+    },
+    {
+      "epoch": 4.634449190424158,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0004292480381444323,
+      "loss": 0.5287,
+      "step": 93310
+    },
+    {
+      "epoch": 4.6349458627197775,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00042920830436078277,
+      "loss": 0.4945,
+      "step": 93320
+    },
+    {
+      "epoch": 4.635442535015397,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00042916857057713324,
+      "loss": 0.5056,
+      "step": 93330
+    },
+    {
+      "epoch": 4.635939207311016,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0004291288367934837,
+      "loss": 0.5392,
+      "step": 93340
+    },
+    {
+      "epoch": 4.636435879606635,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0004290891030098341,
+      "loss": 0.5309,
+      "step": 93350
+    },
+    {
+      "epoch": 4.636932551902255,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004290493692261846,
+      "loss": 0.5161,
+      "step": 93360
+    },
+    {
+      "epoch": 4.6374292241978745,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00042900963544253507,
+      "loss": 0.5358,
+      "step": 93370
+    },
+    {
+      "epoch": 4.637925896493494,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0004289699016588855,
+      "loss": 0.5189,
+      "step": 93380
+    },
+    {
+      "epoch": 4.638422568789113,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.00042893016787523596,
+      "loss": 0.5122,
+      "step": 93390
+    },
+    {
+      "epoch": 4.638919241084732,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00042889043409158643,
+      "loss": 0.4765,
+      "step": 93400
+    },
+    {
+      "epoch": 4.639415913380351,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0004288507003079368,
+      "loss": 0.4987,
+      "step": 93410
+    },
+    {
+      "epoch": 4.639912585675971,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0004288109665242873,
+      "loss": 0.5058,
+      "step": 93420
+    },
+    {
+      "epoch": 4.640409257971591,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0004287712327406378,
+      "loss": 0.4966,
+      "step": 93430
+    },
+    {
+      "epoch": 4.64090593026721,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00042873149895698815,
+      "loss": 0.5476,
+      "step": 93440
+    },
+    {
+      "epoch": 4.641402602562829,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0004286917651733386,
+      "loss": 0.5086,
+      "step": 93450
+    },
+    {
+      "epoch": 4.641899274858448,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00042865203138968915,
+      "loss": 0.515,
+      "step": 93460
+    },
+    {
+      "epoch": 4.642395947154068,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004286122976060395,
+      "loss": 0.4866,
+      "step": 93470
+    },
+    {
+      "epoch": 4.642892619449687,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00042857256382239,
+      "loss": 0.4974,
+      "step": 93480
+    },
+    {
+      "epoch": 4.643389291745306,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00042853283003874045,
+      "loss": 0.5037,
+      "step": 93490
+    },
+    {
+      "epoch": 4.643885964040926,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.000428493096255091,
+      "loss": 0.5221,
+      "step": 93500
+    },
+    {
+      "epoch": 4.6443826363365455,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00042845336247144134,
+      "loss": 0.4928,
+      "step": 93510
+    },
+    {
+      "epoch": 4.644879308632165,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0004284136286877918,
+      "loss": 0.5028,
+      "step": 93520
+    },
+    {
+      "epoch": 4.645375980927784,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0004283738949041423,
+      "loss": 0.5423,
+      "step": 93530
+    },
+    {
+      "epoch": 4.645872653223403,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0004283341611204927,
+      "loss": 0.4969,
+      "step": 93540
+    },
+    {
+      "epoch": 4.646369325519022,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004282944273368432,
+      "loss": 0.5313,
+      "step": 93550
+    },
+    {
+      "epoch": 4.646865997814642,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00042825469355319365,
+      "loss": 0.5243,
+      "step": 93560
+    },
+    {
+      "epoch": 4.647362670110262,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00042821495976954406,
+      "loss": 0.5266,
+      "step": 93570
+    },
+    {
+      "epoch": 4.647859342405881,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00042817522598589453,
+      "loss": 0.4925,
+      "step": 93580
+    },
+    {
+      "epoch": 4.6483560147015,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.000428135492202245,
+      "loss": 0.5117,
+      "step": 93590
+    },
+    {
+      "epoch": 4.648852686997119,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00042809575841859537,
+      "loss": 0.502,
+      "step": 93600
+    },
+    {
+      "epoch": 4.649349359292739,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0004280560246349459,
+      "loss": 0.5174,
+      "step": 93610
+    },
+    {
+      "epoch": 4.649846031588358,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00042801629085129637,
+      "loss": 0.4945,
+      "step": 93620
+    },
+    {
+      "epoch": 4.650342703883977,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00042797655706764673,
+      "loss": 0.5148,
+      "step": 93630
+    },
+    {
+      "epoch": 4.650839376179597,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004279368232839972,
+      "loss": 0.5034,
+      "step": 93640
+    },
+    {
+      "epoch": 4.651336048475216,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0004278970895003477,
+      "loss": 0.5312,
+      "step": 93650
+    },
+    {
+      "epoch": 4.651832720770836,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0004278573557166982,
+      "loss": 0.4967,
+      "step": 93660
+    },
+    {
+      "epoch": 4.652329393066455,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00042781762193304856,
+      "loss": 0.5173,
+      "step": 93670
+    },
+    {
+      "epoch": 4.652826065362074,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00042777788814939903,
+      "loss": 0.5002,
+      "step": 93680
+    },
+    {
+      "epoch": 4.653322737657693,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.00042773815436574956,
+      "loss": 0.5336,
+      "step": 93690
+    },
+    {
+      "epoch": 4.653819409953313,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004276984205820999,
+      "loss": 0.5114,
+      "step": 93700
+    },
+    {
+      "epoch": 4.654316082248932,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0004276586867984504,
+      "loss": 0.4867,
+      "step": 93710
+    },
+    {
+      "epoch": 4.654812754544551,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00042761895301480086,
+      "loss": 0.4993,
+      "step": 93720
+    },
+    {
+      "epoch": 4.655309426840171,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0004275792192311513,
+      "loss": 0.5041,
+      "step": 93730
+    },
+    {
+      "epoch": 4.65580609913579,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00042753948544750175,
+      "loss": 0.5209,
+      "step": 93740
+    },
+    {
+      "epoch": 4.65630277143141,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0004274997516638522,
+      "loss": 0.5342,
+      "step": 93750
+    },
+    {
+      "epoch": 4.656799443727029,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00042746001788020264,
+      "loss": 0.5215,
+      "step": 93760
+    },
+    {
+      "epoch": 4.657296116022648,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004274202840965531,
+      "loss": 0.5045,
+      "step": 93770
+    },
+    {
+      "epoch": 4.657792788318267,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0004273805503129036,
+      "loss": 0.5057,
+      "step": 93780
+    },
+    {
+      "epoch": 4.6582894606138865,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00042734081652925405,
+      "loss": 0.511,
+      "step": 93790
+    },
+    {
+      "epoch": 4.658786132909507,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.00042730108274560447,
+      "loss": 0.5015,
+      "step": 93800
+    },
+    {
+      "epoch": 4.659282805205126,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00042726134896195494,
+      "loss": 0.5054,
+      "step": 93810
+    },
+    {
+      "epoch": 4.659779477500745,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004272216151783054,
+      "loss": 0.5361,
+      "step": 93820
+    },
+    {
+      "epoch": 4.660276149796364,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00042718188139465583,
+      "loss": 0.5188,
+      "step": 93830
+    },
+    {
+      "epoch": 4.6607728220919835,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0004271421476110063,
+      "loss": 0.4968,
+      "step": 93840
+    },
+    {
+      "epoch": 4.661269494387603,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00042710241382735677,
+      "loss": 0.4918,
+      "step": 93850
+    },
+    {
+      "epoch": 4.661766166683222,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00042706268004370714,
+      "loss": 0.533,
+      "step": 93860
+    },
+    {
+      "epoch": 4.662262838978842,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0004270229462600576,
+      "loss": 0.4973,
+      "step": 93870
+    },
+    {
+      "epoch": 4.662759511274461,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00042698321247640813,
+      "loss": 0.4874,
+      "step": 93880
+    },
+    {
+      "epoch": 4.6632561835700805,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0004269434786927585,
+      "loss": 0.5093,
+      "step": 93890
+    },
+    {
+      "epoch": 4.6637528558657,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00042690374490910897,
+      "loss": 0.4838,
+      "step": 93900
+    },
+    {
+      "epoch": 4.664249528161319,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00042686401112545944,
+      "loss": 0.5036,
+      "step": 93910
+    },
+    {
+      "epoch": 4.664746200456938,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00042682427734180986,
+      "loss": 0.5148,
+      "step": 93920
+    },
+    {
+      "epoch": 4.6652428727525574,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004267845435581603,
+      "loss": 0.4976,
+      "step": 93930
+    },
+    {
+      "epoch": 4.665739545048178,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0004267448097745108,
+      "loss": 0.5301,
+      "step": 93940
+    },
+    {
+      "epoch": 4.666236217343797,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00042670507599086127,
+      "loss": 0.5311,
+      "step": 93950
+    },
+    {
+      "epoch": 4.666732889639416,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0004266653422072117,
+      "loss": 0.5232,
+      "step": 93960
+    },
+    {
+      "epoch": 4.667229561935035,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00042662560842356216,
+      "loss": 0.5074,
+      "step": 93970
+    },
+    {
+      "epoch": 4.6677262342306545,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00042658587463991263,
+      "loss": 0.5078,
+      "step": 93980
+    },
+    {
+      "epoch": 4.668222906526274,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.00042654614085626305,
+      "loss": 0.5122,
+      "step": 93990
+    },
+    {
+      "epoch": 4.668719578821893,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0004265064070726135,
+      "loss": 0.5162,
+      "step": 94000
+    },
+    {
+      "epoch": 4.669216251117513,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000426466673288964,
+      "loss": 0.5243,
+      "step": 94010
+    },
+    {
+      "epoch": 4.669712923413132,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0004264269395053144,
+      "loss": 0.5339,
+      "step": 94020
+    },
+    {
+      "epoch": 4.6702095957087515,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0004263872057216649,
+      "loss": 0.5147,
+      "step": 94030
+    },
+    {
+      "epoch": 4.670706268004371,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00042634747193801535,
+      "loss": 0.5363,
+      "step": 94040
+    },
+    {
+      "epoch": 4.67120294029999,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0004263077381543657,
+      "loss": 0.5158,
+      "step": 94050
+    },
+    {
+      "epoch": 4.671699612595609,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00042626800437071624,
+      "loss": 0.5312,
+      "step": 94060
+    },
+    {
+      "epoch": 4.672196284891228,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0004262282705870667,
+      "loss": 0.529,
+      "step": 94070
+    },
+    {
+      "epoch": 4.6726929571868485,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00042618853680341707,
+      "loss": 0.5017,
+      "step": 94080
+    },
+    {
+      "epoch": 4.673189629482468,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00042614880301976754,
+      "loss": 0.5209,
+      "step": 94090
+    },
+    {
+      "epoch": 4.673686301778087,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.000426109069236118,
+      "loss": 0.5108,
+      "step": 94100
+    },
+    {
+      "epoch": 4.674182974073706,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00042606933545246854,
+      "loss": 0.5438,
+      "step": 94110
+    },
+    {
+      "epoch": 4.674679646369325,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004260296016688189,
+      "loss": 0.5113,
+      "step": 94120
+    },
+    {
+      "epoch": 4.675176318664945,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0004259898678851694,
+      "loss": 0.4907,
+      "step": 94130
+    },
+    {
+      "epoch": 4.675672990960564,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00042595013410151985,
+      "loss": 0.528,
+      "step": 94140
+    },
+    {
+      "epoch": 4.676169663256184,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00042591040031787026,
+      "loss": 0.53,
+      "step": 94150
+    },
+    {
+      "epoch": 4.676666335551803,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00042587066653422073,
+      "loss": 0.5285,
+      "step": 94160
+    },
+    {
+      "epoch": 4.6771630078474224,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004258309327505712,
+      "loss": 0.5242,
+      "step": 94170
+    },
+    {
+      "epoch": 4.677659680143042,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0004257911989669216,
+      "loss": 0.5066,
+      "step": 94180
+    },
+    {
+      "epoch": 4.678156352438661,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0004257514651832721,
+      "loss": 0.5107,
+      "step": 94190
+    },
+    {
+      "epoch": 4.67865302473428,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00042571173139962257,
+      "loss": 0.5478,
+      "step": 94200
+    },
+    {
+      "epoch": 4.679149697029899,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.000425671997615973,
+      "loss": 0.5088,
+      "step": 94210
+    },
+    {
+      "epoch": 4.6796463693255195,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00042563226383232345,
+      "loss": 0.5017,
+      "step": 94220
+    },
+    {
+      "epoch": 4.680143041621139,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0004255925300486739,
+      "loss": 0.5478,
+      "step": 94230
+    },
+    {
+      "epoch": 4.680639713916758,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0004255527962650244,
+      "loss": 0.5269,
+      "step": 94240
+    },
+    {
+      "epoch": 4.681136386212377,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0004255130624813748,
+      "loss": 0.506,
+      "step": 94250
+    },
+    {
+      "epoch": 4.681633058507996,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0004254733286977253,
+      "loss": 0.533,
+      "step": 94260
+    },
+    {
+      "epoch": 4.682129730803616,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00042543359491407576,
+      "loss": 0.5428,
+      "step": 94270
+    },
+    {
+      "epoch": 4.682626403099235,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004253938611304261,
+      "loss": 0.51,
+      "step": 94280
+    },
+    {
+      "epoch": 4.683123075394855,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00042535412734677664,
+      "loss": 0.5113,
+      "step": 94290
+    },
+    {
+      "epoch": 4.683619747690474,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004253143935631271,
+      "loss": 0.5103,
+      "step": 94300
+    },
+    {
+      "epoch": 4.684116419986093,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0004252746597794775,
+      "loss": 0.4983,
+      "step": 94310
+    },
+    {
+      "epoch": 4.684613092281713,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00042523492599582795,
+      "loss": 0.506,
+      "step": 94320
+    },
+    {
+      "epoch": 4.685109764577332,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0004251951922121785,
+      "loss": 0.5123,
+      "step": 94330
+    },
+    {
+      "epoch": 4.685606436872951,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00042515545842852884,
+      "loss": 0.5111,
+      "step": 94340
+    },
+    {
+      "epoch": 4.68610310916857,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0004251157246448793,
+      "loss": 0.5294,
+      "step": 94350
+    },
+    {
+      "epoch": 4.6865997814641895,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0004250759908612298,
+      "loss": 0.5078,
+      "step": 94360
+    },
+    {
+      "epoch": 4.68709645375981,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004250362570775802,
+      "loss": 0.5218,
+      "step": 94370
+    },
+    {
+      "epoch": 4.687593126055429,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00042499652329393067,
+      "loss": 0.5118,
+      "step": 94380
+    },
+    {
+      "epoch": 4.688089798351048,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00042495678951028114,
+      "loss": 0.5267,
+      "step": 94390
+    },
+    {
+      "epoch": 4.688586470646667,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0004249170557266316,
+      "loss": 0.5034,
+      "step": 94400
+    },
+    {
+      "epoch": 4.689083142942287,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00042487732194298203,
+      "loss": 0.502,
+      "step": 94410
+    },
+    {
+      "epoch": 4.689579815237906,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004248375881593325,
+      "loss": 0.514,
+      "step": 94420
+    },
+    {
+      "epoch": 4.690076487533525,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.000424797854375683,
+      "loss": 0.5047,
+      "step": 94430
+    },
+    {
+      "epoch": 4.690573159829144,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0004247581205920334,
+      "loss": 0.5045,
+      "step": 94440
+    },
+    {
+      "epoch": 4.691069832124764,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00042471838680838386,
+      "loss": 0.5306,
+      "step": 94450
+    },
+    {
+      "epoch": 4.691566504420384,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.00042467865302473433,
+      "loss": 0.4941,
+      "step": 94460
+    },
+    {
+      "epoch": 4.692063176716003,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0004246389192410847,
+      "loss": 0.5001,
+      "step": 94470
+    },
+    {
+      "epoch": 4.692559849011622,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004245991854574352,
+      "loss": 0.5268,
+      "step": 94480
+    },
+    {
+      "epoch": 4.693056521307241,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0004245594516737857,
+      "loss": 0.5235,
+      "step": 94490
+    },
+    {
+      "epoch": 4.6935531936028605,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00042451971789013606,
+      "loss": 0.5035,
+      "step": 94500
+    },
+    {
+      "epoch": 4.69404986589848,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0004244799841064865,
+      "loss": 0.5141,
+      "step": 94510
+    },
+    {
+      "epoch": 4.6945465381941,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00042444025032283705,
+      "loss": 0.5439,
+      "step": 94520
+    },
+    {
+      "epoch": 4.695043210489719,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0004244005165391874,
+      "loss": 0.5259,
+      "step": 94530
+    },
+    {
+      "epoch": 4.695539882785338,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0004243607827555379,
+      "loss": 0.5355,
+      "step": 94540
+    },
+    {
+      "epoch": 4.6960365550809575,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.00042432104897188836,
+      "loss": 0.5214,
+      "step": 94550
+    },
+    {
+      "epoch": 4.696533227376577,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0004242813151882389,
+      "loss": 0.5157,
+      "step": 94560
+    },
+    {
+      "epoch": 4.697029899672196,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.00042424158140458925,
+      "loss": 0.5294,
+      "step": 94570
+    },
+    {
+      "epoch": 4.697526571967815,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0004242018476209397,
+      "loss": 0.535,
+      "step": 94580
+    },
+    {
+      "epoch": 4.698023244263435,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004241621138372902,
+      "loss": 0.5182,
+      "step": 94590
+    },
+    {
+      "epoch": 4.6985199165590545,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004241223800536406,
+      "loss": 0.5396,
+      "step": 94600
+    },
+    {
+      "epoch": 4.699016588854674,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0004240826462699911,
+      "loss": 0.4946,
+      "step": 94610
+    },
+    {
+      "epoch": 4.699513261150293,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00042404291248634155,
+      "loss": 0.5151,
+      "step": 94620
+    },
+    {
+      "epoch": 4.700009933445912,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00042400317870269197,
+      "loss": 0.514,
+      "step": 94630
+    },
+    {
+      "epoch": 4.7005066057415315,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00042396344491904244,
+      "loss": 0.5305,
+      "step": 94640
+    },
+    {
+      "epoch": 4.701003278037151,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0004239237111353929,
+      "loss": 0.5106,
+      "step": 94650
+    },
+    {
+      "epoch": 4.701499950332771,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0004238839773517433,
+      "loss": 0.5304,
+      "step": 94660
+    },
+    {
+      "epoch": 4.70199662262839,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0004238442435680938,
+      "loss": 0.5195,
+      "step": 94670
+    },
+    {
+      "epoch": 4.702493294924009,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00042380450978444427,
+      "loss": 0.5279,
+      "step": 94680
+    },
+    {
+      "epoch": 4.7029899672196285,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00042376477600079474,
+      "loss": 0.5219,
+      "step": 94690
+    },
+    {
+      "epoch": 4.703486639515248,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004237250422171451,
+      "loss": 0.5455,
+      "step": 94700
+    },
+    {
+      "epoch": 4.703983311810867,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00042368530843349563,
+      "loss": 0.5194,
+      "step": 94710
+    },
+    {
+      "epoch": 4.704479984106486,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004236455746498461,
+      "loss": 0.5275,
+      "step": 94720
+    },
+    {
+      "epoch": 4.704976656402106,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00042360584086619646,
+      "loss": 0.5154,
+      "step": 94730
+    },
+    {
+      "epoch": 4.7054733286977255,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00042356610708254693,
+      "loss": 0.5294,
+      "step": 94740
+    },
+    {
+      "epoch": 4.705970000993345,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00042352637329889746,
+      "loss": 0.5077,
+      "step": 94750
+    },
+    {
+      "epoch": 4.706466673288964,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0004234866395152478,
+      "loss": 0.5258,
+      "step": 94760
+    },
+    {
+      "epoch": 4.706963345584583,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0004234469057315983,
+      "loss": 0.5071,
+      "step": 94770
+    },
+    {
+      "epoch": 4.707460017880202,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00042340717194794877,
+      "loss": 0.5164,
+      "step": 94780
+    },
+    {
+      "epoch": 4.707956690175822,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0004233674381642992,
+      "loss": 0.4936,
+      "step": 94790
+    },
+    {
+      "epoch": 4.708453362471442,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00042332770438064965,
+      "loss": 0.5188,
+      "step": 94800
+    },
+    {
+      "epoch": 4.708950034767061,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0004232879705970001,
+      "loss": 0.4913,
+      "step": 94810
+    },
+    {
+      "epoch": 4.70944670706268,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00042324823681335054,
+      "loss": 0.543,
+      "step": 94820
+    },
+    {
+      "epoch": 4.709943379358299,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.000423208503029701,
+      "loss": 0.5066,
+      "step": 94830
+    },
+    {
+      "epoch": 4.710440051653919,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0004231687692460515,
+      "loss": 0.5207,
+      "step": 94840
+    },
+    {
+      "epoch": 4.710936723949538,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00042312903546240196,
+      "loss": 0.5341,
+      "step": 94850
+    },
+    {
+      "epoch": 4.711433396245157,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0004230893016787524,
+      "loss": 0.5013,
+      "step": 94860
+    },
+    {
+      "epoch": 4.711930068540777,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00042304956789510285,
+      "loss": 0.5352,
+      "step": 94870
+    },
+    {
+      "epoch": 4.7124267408363965,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004230098341114533,
+      "loss": 0.5075,
+      "step": 94880
+    },
+    {
+      "epoch": 4.712923413132016,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00042297010032780373,
+      "loss": 0.4926,
+      "step": 94890
+    },
+    {
+      "epoch": 4.713420085427635,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004229303665441542,
+      "loss": 0.5439,
+      "step": 94900
+    },
+    {
+      "epoch": 4.713916757723254,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0004228906327605047,
+      "loss": 0.5166,
+      "step": 94910
+    },
+    {
+      "epoch": 4.714413430018873,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00042285089897685504,
+      "loss": 0.5406,
+      "step": 94920
+    },
+    {
+      "epoch": 4.714910102314493,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00042281116519320556,
+      "loss": 0.5206,
+      "step": 94930
+    },
+    {
+      "epoch": 4.715406774610113,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00042277143140955604,
+      "loss": 0.4995,
+      "step": 94940
+    },
+    {
+      "epoch": 4.715903446905732,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0004227316976259064,
+      "loss": 0.4979,
+      "step": 94950
+    },
+    {
+      "epoch": 4.716400119201351,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00042269196384225687,
+      "loss": 0.5015,
+      "step": 94960
+    },
+    {
+      "epoch": 4.71689679149697,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00042265223005860734,
+      "loss": 0.5038,
+      "step": 94970
+    },
+    {
+      "epoch": 4.71739346379259,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00042261249627495776,
+      "loss": 0.5304,
+      "step": 94980
+    },
+    {
+      "epoch": 4.717890136088209,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00042257276249130823,
+      "loss": 0.5447,
+      "step": 94990
+    },
+    {
+      "epoch": 4.718386808383828,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004225330287076587,
+      "loss": 0.4964,
+      "step": 95000
+    },
+    {
+      "epoch": 4.718883480679448,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0004224932949240092,
+      "loss": 0.5399,
+      "step": 95010
+    },
+    {
+      "epoch": 4.719380152975067,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0004224535611403596,
+      "loss": 0.5241,
+      "step": 95020
+    },
+    {
+      "epoch": 4.719876825270687,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00042241382735671006,
+      "loss": 0.5011,
+      "step": 95030
+    },
+    {
+      "epoch": 4.720373497566306,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00042237409357306053,
+      "loss": 0.5227,
+      "step": 95040
+    },
+    {
+      "epoch": 4.720870169861925,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00042233435978941095,
+      "loss": 0.4831,
+      "step": 95050
+    },
+    {
+      "epoch": 4.721366842157544,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0004222946260057614,
+      "loss": 0.5311,
+      "step": 95060
+    },
+    {
+      "epoch": 4.7218635144531635,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0004222548922221119,
+      "loss": 0.5486,
+      "step": 95070
+    },
+    {
+      "epoch": 4.722360186748783,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0004222151584384623,
+      "loss": 0.4966,
+      "step": 95080
+    },
+    {
+      "epoch": 4.722856859044403,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0004221754246548128,
+      "loss": 0.5321,
+      "step": 95090
+    },
+    {
+      "epoch": 4.723353531340022,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00042213569087116325,
+      "loss": 0.5146,
+      "step": 95100
+    },
+    {
+      "epoch": 4.723850203635641,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0004220959570875136,
+      "loss": 0.5115,
+      "step": 95110
+    },
+    {
+      "epoch": 4.724346875931261,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00042205622330386414,
+      "loss": 0.5373,
+      "step": 95120
+    },
+    {
+      "epoch": 4.72484354822688,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0004220164895202146,
+      "loss": 0.5361,
+      "step": 95130
+    },
+    {
+      "epoch": 4.725340220522499,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0004219767557365651,
+      "loss": 0.5529,
+      "step": 95140
+    },
+    {
+      "epoch": 4.725836892818118,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00042193702195291545,
+      "loss": 0.5089,
+      "step": 95150
+    },
+    {
+      "epoch": 4.7263335651137375,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00042189728816926597,
+      "loss": 0.5577,
+      "step": 95160
+    },
+    {
+      "epoch": 4.726830237409358,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00042185755438561644,
+      "loss": 0.5262,
+      "step": 95170
+    },
+    {
+      "epoch": 4.727326909704977,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0004218178206019668,
+      "loss": 0.5136,
+      "step": 95180
+    },
+    {
+      "epoch": 4.727823582000596,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0004217780868183173,
+      "loss": 0.4981,
+      "step": 95190
+    },
+    {
+      "epoch": 4.728320254296215,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0004217383530346678,
+      "loss": 0.5209,
+      "step": 95200
+    },
+    {
+      "epoch": 4.7288169265918345,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00042169861925101817,
+      "loss": 0.5366,
+      "step": 95210
+    },
+    {
+      "epoch": 4.729313598887454,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00042165888546736864,
+      "loss": 0.4795,
+      "step": 95220
+    },
+    {
+      "epoch": 4.729810271183073,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0004216191516837191,
+      "loss": 0.5345,
+      "step": 95230
+    },
+    {
+      "epoch": 4.730306943478693,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0004215794179000695,
+      "loss": 0.5245,
+      "step": 95240
+    },
+    {
+      "epoch": 4.730803615774312,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00042153968411642,
+      "loss": 0.5295,
+      "step": 95250
+    },
+    {
+      "epoch": 4.7313002880699315,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00042149995033277047,
+      "loss": 0.5018,
+      "step": 95260
+    },
+    {
+      "epoch": 4.731796960365551,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004214602165491209,
+      "loss": 0.4959,
+      "step": 95270
+    },
+    {
+      "epoch": 4.73229363266117,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.00042142048276547136,
+      "loss": 0.517,
+      "step": 95280
+    },
+    {
+      "epoch": 4.732790304956789,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00042138074898182183,
+      "loss": 0.5057,
+      "step": 95290
+    },
+    {
+      "epoch": 4.733286977252408,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0004213410151981723,
+      "loss": 0.5063,
+      "step": 95300
+    },
+    {
+      "epoch": 4.7337836495480285,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0004213012814145227,
+      "loss": 0.5184,
+      "step": 95310
+    },
+    {
+      "epoch": 4.734280321843648,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0004212615476308732,
+      "loss": 0.5132,
+      "step": 95320
+    },
+    {
+      "epoch": 4.734776994139267,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00042122181384722366,
+      "loss": 0.5021,
+      "step": 95330
+    },
+    {
+      "epoch": 4.735273666434886,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.000421182080063574,
+      "loss": 0.4903,
+      "step": 95340
+    },
+    {
+      "epoch": 4.7357703387305055,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00042114234627992455,
+      "loss": 0.5328,
+      "step": 95350
+    },
+    {
+      "epoch": 4.736267011026125,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.000421102612496275,
+      "loss": 0.5182,
+      "step": 95360
+    },
+    {
+      "epoch": 4.736763683321744,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0004210628787126254,
+      "loss": 0.5032,
+      "step": 95370
+    },
+    {
+      "epoch": 4.737260355617364,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00042102314492897585,
+      "loss": 0.5062,
+      "step": 95380
+    },
+    {
+      "epoch": 4.737757027912983,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004209834111453264,
+      "loss": 0.5156,
+      "step": 95390
+    },
+    {
+      "epoch": 4.7382537002086025,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.00042094367736167674,
+      "loss": 0.5158,
+      "step": 95400
+    },
+    {
+      "epoch": 4.738750372504222,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0004209039435780272,
+      "loss": 0.5502,
+      "step": 95410
+    },
+    {
+      "epoch": 4.739247044799841,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0004208642097943777,
+      "loss": 0.5035,
+      "step": 95420
+    },
+    {
+      "epoch": 4.73974371709546,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004208244760107282,
+      "loss": 0.5091,
+      "step": 95430
+    },
+    {
+      "epoch": 4.740240389391079,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004207847422270786,
+      "loss": 0.5248,
+      "step": 95440
+    },
+    {
+      "epoch": 4.7407370616866995,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00042074500844342905,
+      "loss": 0.5515,
+      "step": 95450
+    },
+    {
+      "epoch": 4.741233733982319,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0004207052746597795,
+      "loss": 0.5064,
+      "step": 95460
+    },
+    {
+      "epoch": 4.741730406277938,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00042066554087612993,
+      "loss": 0.506,
+      "step": 95470
+    },
+    {
+      "epoch": 4.742227078573557,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0004206258070924804,
+      "loss": 0.4984,
+      "step": 95480
+    },
+    {
+      "epoch": 4.742723750869176,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0004205860733088309,
+      "loss": 0.5167,
+      "step": 95490
+    },
+    {
+      "epoch": 4.743220423164796,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0004205463395251813,
+      "loss": 0.5086,
+      "step": 95500
+    },
+    {
+      "epoch": 4.743717095460415,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00042050660574153177,
+      "loss": 0.5237,
+      "step": 95510
+    },
+    {
+      "epoch": 4.744213767756035,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00042046687195788224,
+      "loss": 0.5113,
+      "step": 95520
+    },
+    {
+      "epoch": 4.744710440051654,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0004204271381742326,
+      "loss": 0.5052,
+      "step": 95530
+    },
+    {
+      "epoch": 4.745207112347273,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0004203874043905831,
+      "loss": 0.5447,
+      "step": 95540
+    },
+    {
+      "epoch": 4.745703784642893,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0004203476706069336,
+      "loss": 0.5118,
+      "step": 95550
+    },
+    {
+      "epoch": 4.746200456938512,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.00042030793682328396,
+      "loss": 0.4928,
+      "step": 95560
+    },
+    {
+      "epoch": 4.746697129234131,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00042026820303963443,
+      "loss": 0.5002,
+      "step": 95570
+    },
+    {
+      "epoch": 4.74719380152975,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00042022846925598496,
+      "loss": 0.5171,
+      "step": 95580
+    },
+    {
+      "epoch": 4.7476904738253705,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00042018873547233543,
+      "loss": 0.5324,
+      "step": 95590
+    },
+    {
+      "epoch": 4.74818714612099,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004201490016886858,
+      "loss": 0.5299,
+      "step": 95600
+    },
+    {
+      "epoch": 4.748683818416609,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00042010926790503626,
+      "loss": 0.5255,
+      "step": 95610
+    },
+    {
+      "epoch": 4.749180490712228,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0004200695341213868,
+      "loss": 0.5217,
+      "step": 95620
+    },
+    {
+      "epoch": 4.749677163007847,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00042002980033773715,
+      "loss": 0.4949,
+      "step": 95630
+    },
+    {
+      "epoch": 4.750173835303467,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0004199900665540876,
+      "loss": 0.5195,
+      "step": 95640
+    },
+    {
+      "epoch": 4.750670507599086,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0004199503327704381,
+      "loss": 0.5092,
+      "step": 95650
+    },
+    {
+      "epoch": 4.751167179894706,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0004199105989867885,
+      "loss": 0.5452,
+      "step": 95660
+    },
+    {
+      "epoch": 4.751663852190325,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.000419870865203139,
+      "loss": 0.5041,
+      "step": 95670
+    },
+    {
+      "epoch": 4.752160524485944,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00041983113141948945,
+      "loss": 0.5251,
+      "step": 95680
+    },
+    {
+      "epoch": 4.752657196781564,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00041979139763583987,
+      "loss": 0.5251,
+      "step": 95690
+    },
+    {
+      "epoch": 4.753153869077183,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00041975166385219034,
+      "loss": 0.4777,
+      "step": 95700
+    },
+    {
+      "epoch": 4.753650541372802,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004197119300685408,
+      "loss": 0.5222,
+      "step": 95710
+    },
+    {
+      "epoch": 4.754147213668421,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00041967219628489123,
+      "loss": 0.5147,
+      "step": 95720
+    },
+    {
+      "epoch": 4.754643885964041,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0004196324625012417,
+      "loss": 0.5202,
+      "step": 95730
+    },
+    {
+      "epoch": 4.755140558259661,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00041959272871759217,
+      "loss": 0.4835,
+      "step": 95740
+    },
+    {
+      "epoch": 4.75563723055528,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00041955299493394264,
+      "loss": 0.4836,
+      "step": 95750
+    },
+    {
+      "epoch": 4.756133902850899,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00041951326115029306,
+      "loss": 0.5481,
+      "step": 95760
+    },
+    {
+      "epoch": 4.756630575146518,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00041947352736664353,
+      "loss": 0.5153,
+      "step": 95770
+    },
+    {
+      "epoch": 4.7571272474421376,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.000419433793582994,
+      "loss": 0.5352,
+      "step": 95780
+    },
+    {
+      "epoch": 4.757623919737757,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00041939405979934437,
+      "loss": 0.5043,
+      "step": 95790
+    },
+    {
+      "epoch": 4.758120592033376,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00041935432601569484,
+      "loss": 0.5119,
+      "step": 95800
+    },
+    {
+      "epoch": 4.758617264328995,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00041931459223204536,
+      "loss": 0.5162,
+      "step": 95810
+    },
+    {
+      "epoch": 4.759113936624615,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004192748584483957,
+      "loss": 0.5127,
+      "step": 95820
+    },
+    {
+      "epoch": 4.759610608920235,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0004192351246647462,
+      "loss": 0.5209,
+      "step": 95830
+    },
+    {
+      "epoch": 4.760107281215854,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00041919539088109667,
+      "loss": 0.5126,
+      "step": 95840
+    },
+    {
+      "epoch": 4.760603953511473,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004191556570974471,
+      "loss": 0.4896,
+      "step": 95850
+    },
+    {
+      "epoch": 4.761100625807092,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00041911592331379756,
+      "loss": 0.5022,
+      "step": 95860
+    },
+    {
+      "epoch": 4.7615972981027115,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00041907618953014803,
+      "loss": 0.5168,
+      "step": 95870
+    },
+    {
+      "epoch": 4.762093970398331,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0004190364557464985,
+      "loss": 0.5049,
+      "step": 95880
+    },
+    {
+      "epoch": 4.762590642693951,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0004189967219628489,
+      "loss": 0.4866,
+      "step": 95890
+    },
+    {
+      "epoch": 4.76308731498957,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0004189569881791994,
+      "loss": 0.5106,
+      "step": 95900
+    },
+    {
+      "epoch": 4.763583987285189,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00041891725439554986,
+      "loss": 0.4877,
+      "step": 95910
+    },
+    {
+      "epoch": 4.7640806595808085,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0004188775206119003,
+      "loss": 0.513,
+      "step": 95920
+    },
+    {
+      "epoch": 4.764577331876428,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00041883778682825075,
+      "loss": 0.4947,
+      "step": 95930
+    },
+    {
+      "epoch": 4.765074004172047,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0004187980530446012,
+      "loss": 0.5392,
+      "step": 95940
+    },
+    {
+      "epoch": 4.765570676467666,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00041875831926095164,
+      "loss": 0.5238,
+      "step": 95950
+    },
+    {
+      "epoch": 4.766067348763286,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0004187185854773021,
+      "loss": 0.5083,
+      "step": 95960
+    },
+    {
+      "epoch": 4.7665640210589055,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0004186788516936526,
+      "loss": 0.5545,
+      "step": 95970
+    },
+    {
+      "epoch": 4.767060693354525,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00041863911791000294,
+      "loss": 0.5336,
+      "step": 95980
+    },
+    {
+      "epoch": 4.767557365650144,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00041859938412635347,
+      "loss": 0.5086,
+      "step": 95990
+    },
+    {
+      "epoch": 4.768054037945763,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00041855965034270394,
+      "loss": 0.5403,
+      "step": 96000
+    },
+    {
+      "epoch": 4.768550710241382,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0004185199165590543,
+      "loss": 0.5288,
+      "step": 96010
+    },
+    {
+      "epoch": 4.769047382537002,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0004184801827754048,
+      "loss": 0.5142,
+      "step": 96020
+    },
+    {
+      "epoch": 4.769544054832622,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0004184404489917553,
+      "loss": 0.509,
+      "step": 96030
+    },
+    {
+      "epoch": 4.770040727128241,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00041840071520810577,
+      "loss": 0.5234,
+      "step": 96040
+    },
+    {
+      "epoch": 4.77053739942386,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00041836098142445613,
+      "loss": 0.522,
+      "step": 96050
+    },
+    {
+      "epoch": 4.7710340717194795,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004183212476408066,
+      "loss": 0.5065,
+      "step": 96060
+    },
+    {
+      "epoch": 4.771530744015099,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004182815138571571,
+      "loss": 0.519,
+      "step": 96070
+    },
+    {
+      "epoch": 4.772027416310718,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004182417800735075,
+      "loss": 0.5165,
+      "step": 96080
+    },
+    {
+      "epoch": 4.772524088606337,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00041820204628985797,
+      "loss": 0.5168,
+      "step": 96090
+    },
+    {
+      "epoch": 4.773020760901957,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.00041816231250620844,
+      "loss": 0.4887,
+      "step": 96100
+    },
+    {
+      "epoch": 4.7735174331975765,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00041812257872255885,
+      "loss": 0.5135,
+      "step": 96110
+    },
+    {
+      "epoch": 4.774014105493196,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0004180828449389093,
+      "loss": 0.4996,
+      "step": 96120
+    },
+    {
+      "epoch": 4.774510777788815,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0004180431111552598,
+      "loss": 0.5189,
+      "step": 96130
+    },
+    {
+      "epoch": 4.775007450084434,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004180033773716102,
+      "loss": 0.5156,
+      "step": 96140
+    },
+    {
+      "epoch": 4.775504122380053,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004179636435879607,
+      "loss": 0.5192,
+      "step": 96150
+    },
+    {
+      "epoch": 4.776000794675673,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00041792390980431116,
+      "loss": 0.5164,
+      "step": 96160
+    },
+    {
+      "epoch": 4.776497466971293,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0004178841760206615,
+      "loss": 0.5318,
+      "step": 96170
+    },
+    {
+      "epoch": 4.776994139266912,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00041784444223701204,
+      "loss": 0.5487,
+      "step": 96180
+    },
+    {
+      "epoch": 4.777490811562531,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0004178047084533625,
+      "loss": 0.553,
+      "step": 96190
+    },
+    {
+      "epoch": 4.77798748385815,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.000417764974669713,
+      "loss": 0.5055,
+      "step": 96200
+    },
+    {
+      "epoch": 4.77848415615377,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00041772524088606335,
+      "loss": 0.4895,
+      "step": 96210
+    },
+    {
+      "epoch": 4.778980828449389,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0004176855071024139,
+      "loss": 0.5172,
+      "step": 96220
+    },
+    {
+      "epoch": 4.779477500745008,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00041764577331876435,
+      "loss": 0.4994,
+      "step": 96230
+    },
+    {
+      "epoch": 4.779974173040628,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004176060395351147,
+      "loss": 0.5072,
+      "step": 96240
+    },
+    {
+      "epoch": 4.780470845336247,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004175663057514652,
+      "loss": 0.5237,
+      "step": 96250
+    },
+    {
+      "epoch": 4.780967517631867,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0004175265719678157,
+      "loss": 0.5339,
+      "step": 96260
+    },
+    {
+      "epoch": 4.781464189927486,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00041748683818416607,
+      "loss": 0.5089,
+      "step": 96270
+    },
+    {
+      "epoch": 4.781960862223105,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00041744710440051654,
+      "loss": 0.4891,
+      "step": 96280
+    },
+    {
+      "epoch": 4.782457534518724,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.000417407370616867,
+      "loss": 0.5031,
+      "step": 96290
+    },
+    {
+      "epoch": 4.782954206814344,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00041736763683321743,
+      "loss": 0.5108,
+      "step": 96300
+    },
+    {
+      "epoch": 4.783450879109964,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0004173279030495679,
+      "loss": 0.5089,
+      "step": 96310
+    },
+    {
+      "epoch": 4.783947551405583,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00041728816926591837,
+      "loss": 0.5291,
+      "step": 96320
+    },
+    {
+      "epoch": 4.784444223701202,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00041724843548226884,
+      "loss": 0.51,
+      "step": 96330
+    },
+    {
+      "epoch": 4.784940895996821,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00041720870169861926,
+      "loss": 0.4878,
+      "step": 96340
+    },
+    {
+      "epoch": 4.785437568292441,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00041716896791496973,
+      "loss": 0.5025,
+      "step": 96350
+    },
+    {
+      "epoch": 4.78593424058806,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0004171292341313202,
+      "loss": 0.5162,
+      "step": 96360
+    },
+    {
+      "epoch": 4.786430912883679,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0004170895003476706,
+      "loss": 0.5003,
+      "step": 96370
+    },
+    {
+      "epoch": 4.786927585179299,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0004170497665640211,
+      "loss": 0.5095,
+      "step": 96380
+    },
+    {
+      "epoch": 4.787424257474918,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00041701003278037156,
+      "loss": 0.4919,
+      "step": 96390
+    },
+    {
+      "epoch": 4.787920929770538,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0004169702989967219,
+      "loss": 0.5275,
+      "step": 96400
+    },
+    {
+      "epoch": 4.788417602066157,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00041693056521307245,
+      "loss": 0.5021,
+      "step": 96410
+    },
+    {
+      "epoch": 4.788914274361776,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0004168908314294229,
+      "loss": 0.5032,
+      "step": 96420
+    },
+    {
+      "epoch": 4.789410946657395,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0004168510976457733,
+      "loss": 0.519,
+      "step": 96430
+    },
+    {
+      "epoch": 4.7899076189530145,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.00041681136386212376,
+      "loss": 0.5152,
+      "step": 96440
+    },
+    {
+      "epoch": 4.790404291248635,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004167716300784743,
+      "loss": 0.5233,
+      "step": 96450
+    },
+    {
+      "epoch": 4.790900963544254,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00041673189629482465,
+      "loss": 0.4867,
+      "step": 96460
+    },
+    {
+      "epoch": 4.791397635839873,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0004166921625111751,
+      "loss": 0.5125,
+      "step": 96470
+    },
+    {
+      "epoch": 4.791894308135492,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0004166524287275256,
+      "loss": 0.5196,
+      "step": 96480
+    },
+    {
+      "epoch": 4.792390980431112,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0004166126949438761,
+      "loss": 0.5539,
+      "step": 96490
+    },
+    {
+      "epoch": 4.792887652726731,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0004165729611602265,
+      "loss": 0.5388,
+      "step": 96500
+    },
+    {
+      "epoch": 4.79338432502235,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00041653322737657695,
+      "loss": 0.4887,
+      "step": 96510
+    },
+    {
+      "epoch": 4.793880997317969,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0004164934935929274,
+      "loss": 0.495,
+      "step": 96520
+    },
+    {
+      "epoch": 4.7943776696135885,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00041645375980927784,
+      "loss": 0.4969,
+      "step": 96530
+    },
+    {
+      "epoch": 4.794874341909209,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0004164140260256283,
+      "loss": 0.4896,
+      "step": 96540
+    },
+    {
+      "epoch": 4.795371014204828,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0004163742922419788,
+      "loss": 0.503,
+      "step": 96550
+    },
+    {
+      "epoch": 4.795867686500447,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0004163345584583292,
+      "loss": 0.5325,
+      "step": 96560
+    },
+    {
+      "epoch": 4.796364358796066,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00041629482467467967,
+      "loss": 0.5173,
+      "step": 96570
+    },
+    {
+      "epoch": 4.7968610310916855,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00041625509089103014,
+      "loss": 0.5358,
+      "step": 96580
+    },
+    {
+      "epoch": 4.797357703387305,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00041621535710738056,
+      "loss": 0.535,
+      "step": 96590
+    },
+    {
+      "epoch": 4.797854375682924,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00041617562332373103,
+      "loss": 0.522,
+      "step": 96600
+    },
+    {
+      "epoch": 4.798351047978544,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0004161358895400815,
+      "loss": 0.5228,
+      "step": 96610
+    },
+    {
+      "epoch": 4.798847720274163,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00041609615575643186,
+      "loss": 0.507,
+      "step": 96620
+    },
+    {
+      "epoch": 4.7993443925697825,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004160564219727824,
+      "loss": 0.4998,
+      "step": 96630
+    },
+    {
+      "epoch": 4.799841064865402,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00041601668818913286,
+      "loss": 0.5215,
+      "step": 96640
+    },
+    {
+      "epoch": 4.800337737161021,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00041597695440548333,
+      "loss": 0.5134,
+      "step": 96650
+    },
+    {
+      "epoch": 4.80083440945664,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004159372206218337,
+      "loss": 0.5027,
+      "step": 96660
+    },
+    {
+      "epoch": 4.801331081752259,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00041589748683818417,
+      "loss": 0.4889,
+      "step": 96670
+    },
+    {
+      "epoch": 4.8018277540478795,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0004158577530545347,
+      "loss": 0.5308,
+      "step": 96680
+    },
+    {
+      "epoch": 4.802324426343499,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00041581801927088505,
+      "loss": 0.5025,
+      "step": 96690
+    },
+    {
+      "epoch": 4.802821098639118,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0004157782854872355,
+      "loss": 0.4896,
+      "step": 96700
+    },
+    {
+      "epoch": 4.803317770934737,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.000415738551703586,
+      "loss": 0.4873,
+      "step": 96710
+    },
+    {
+      "epoch": 4.8038144432303564,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0004156988179199364,
+      "loss": 0.5143,
+      "step": 96720
+    },
+    {
+      "epoch": 4.804311115525976,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004156590841362869,
+      "loss": 0.5448,
+      "step": 96730
+    },
+    {
+      "epoch": 4.804807787821595,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00041561935035263736,
+      "loss": 0.4952,
+      "step": 96740
+    },
+    {
+      "epoch": 4.805304460117215,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0004155796165689878,
+      "loss": 0.5376,
+      "step": 96750
+    },
+    {
+      "epoch": 4.805801132412834,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00041553988278533825,
+      "loss": 0.5119,
+      "step": 96760
+    },
+    {
+      "epoch": 4.8062978047084535,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0004155001490016887,
+      "loss": 0.5141,
+      "step": 96770
+    },
+    {
+      "epoch": 4.806794477004073,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0004154604152180392,
+      "loss": 0.5169,
+      "step": 96780
+    },
+    {
+      "epoch": 4.807291149299692,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004154206814343896,
+      "loss": 0.5102,
+      "step": 96790
+    },
+    {
+      "epoch": 4.807787821595311,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004153809476507401,
+      "loss": 0.5279,
+      "step": 96800
+    },
+    {
+      "epoch": 4.80828449389093,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00041534121386709055,
+      "loss": 0.5039,
+      "step": 96810
+    },
+    {
+      "epoch": 4.8087811661865505,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.00041530148008344096,
+      "loss": 0.497,
+      "step": 96820
+    },
+    {
+      "epoch": 4.80927783848217,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00041526174629979144,
+      "loss": 0.4898,
+      "step": 96830
+    },
+    {
+      "epoch": 4.809774510777789,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004152220125161419,
+      "loss": 0.5293,
+      "step": 96840
+    },
+    {
+      "epoch": 4.810271183073408,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00041518227873249227,
+      "loss": 0.5282,
+      "step": 96850
+    },
+    {
+      "epoch": 4.810767855369027,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0004151425449488428,
+      "loss": 0.5407,
+      "step": 96860
+    },
+    {
+      "epoch": 4.811264527664647,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.00041510281116519327,
+      "loss": 0.5026,
+      "step": 96870
+    },
+    {
+      "epoch": 4.811761199960266,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00041506307738154363,
+      "loss": 0.4928,
+      "step": 96880
+    },
+    {
+      "epoch": 4.812257872255886,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0004150233435978941,
+      "loss": 0.4927,
+      "step": 96890
+    },
+    {
+      "epoch": 4.812754544551505,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00041498360981424463,
+      "loss": 0.5254,
+      "step": 96900
+    },
+    {
+      "epoch": 4.813251216847124,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.000414943876030595,
+      "loss": 0.5341,
+      "step": 96910
+    },
+    {
+      "epoch": 4.813747889142744,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00041490414224694546,
+      "loss": 0.4761,
+      "step": 96920
+    },
+    {
+      "epoch": 4.814244561438363,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.00041486440846329593,
+      "loss": 0.4771,
+      "step": 96930
+    },
+    {
+      "epoch": 4.814741233733982,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0004148246746796464,
+      "loss": 0.5111,
+      "step": 96940
+    },
+    {
+      "epoch": 4.815237906029601,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004147849408959968,
+      "loss": 0.5093,
+      "step": 96950
+    },
+    {
+      "epoch": 4.8157345783252214,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0004147452071123473,
+      "loss": 0.5041,
+      "step": 96960
+    },
+    {
+      "epoch": 4.816231250620841,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00041470547332869776,
+      "loss": 0.4918,
+      "step": 96970
+    },
+    {
+      "epoch": 4.81672792291646,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0004146657395450482,
+      "loss": 0.5107,
+      "step": 96980
+    },
+    {
+      "epoch": 4.817224595212079,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00041462600576139865,
+      "loss": 0.501,
+      "step": 96990
+    },
+    {
+      "epoch": 4.817721267507698,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004145862719777491,
+      "loss": 0.5194,
+      "step": 97000
+    },
+    {
+      "epoch": 4.818217939803318,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00041454653819409954,
+      "loss": 0.5162,
+      "step": 97010
+    },
+    {
+      "epoch": 4.818714612098937,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00041450680441045,
+      "loss": 0.5299,
+      "step": 97020
+    },
+    {
+      "epoch": 4.819211284394557,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0004144670706268005,
+      "loss": 0.4721,
+      "step": 97030
+    },
+    {
+      "epoch": 4.819707956690176,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00041442733684315085,
+      "loss": 0.5487,
+      "step": 97040
+    },
+    {
+      "epoch": 4.820204628985795,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00041438760305950137,
+      "loss": 0.5348,
+      "step": 97050
+    },
+    {
+      "epoch": 4.820701301281415,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00041434786927585184,
+      "loss": 0.5042,
+      "step": 97060
+    },
+    {
+      "epoch": 4.821197973577034,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0004143081354922022,
+      "loss": 0.4872,
+      "step": 97070
+    },
+    {
+      "epoch": 4.821694645872653,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0004142684017085527,
+      "loss": 0.4811,
+      "step": 97080
+    },
+    {
+      "epoch": 4.822191318168272,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004142286679249032,
+      "loss": 0.4971,
+      "step": 97090
+    },
+    {
+      "epoch": 4.822687990463892,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0004141889341412537,
+      "loss": 0.5381,
+      "step": 97100
+    },
+    {
+      "epoch": 4.823184662759512,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00041414920035760404,
+      "loss": 0.5092,
+      "step": 97110
+    },
+    {
+      "epoch": 4.823681335055131,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0004141094665739545,
+      "loss": 0.5213,
+      "step": 97120
+    },
+    {
+      "epoch": 4.82417800735075,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00041406973279030503,
+      "loss": 0.5406,
+      "step": 97130
+    },
+    {
+      "epoch": 4.824674679646369,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0004140299990066554,
+      "loss": 0.5084,
+      "step": 97140
+    },
+    {
+      "epoch": 4.8251713519419885,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00041399026522300587,
+      "loss": 0.5019,
+      "step": 97150
+    },
+    {
+      "epoch": 4.825668024237608,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00041395053143935634,
+      "loss": 0.5396,
+      "step": 97160
+    },
+    {
+      "epoch": 4.826164696533227,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00041391079765570676,
+      "loss": 0.5518,
+      "step": 97170
+    },
+    {
+      "epoch": 4.826661368828847,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00041387106387205723,
+      "loss": 0.4836,
+      "step": 97180
+    },
+    {
+      "epoch": 4.827158041124466,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0004138313300884077,
+      "loss": 0.5422,
+      "step": 97190
+    },
+    {
+      "epoch": 4.827654713420086,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0004137915963047581,
+      "loss": 0.4921,
+      "step": 97200
+    },
+    {
+      "epoch": 4.828151385715705,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0004137518625211086,
+      "loss": 0.5346,
+      "step": 97210
+    },
+    {
+      "epoch": 4.828648058011324,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00041371212873745906,
+      "loss": 0.4824,
+      "step": 97220
+    },
+    {
+      "epoch": 4.829144730306943,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00041367239495380953,
+      "loss": 0.4936,
+      "step": 97230
+    },
+    {
+      "epoch": 4.8296414026025625,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00041363266117015995,
+      "loss": 0.5038,
+      "step": 97240
+    },
+    {
+      "epoch": 4.830138074898182,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0004135929273865104,
+      "loss": 0.5332,
+      "step": 97250
+    },
+    {
+      "epoch": 4.830634747193802,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004135531936028609,
+      "loss": 0.5166,
+      "step": 97260
+    },
+    {
+      "epoch": 4.831131419489421,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00041351345981921125,
+      "loss": 0.5166,
+      "step": 97270
+    },
+    {
+      "epoch": 4.83162809178504,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004134737260355618,
+      "loss": 0.495,
+      "step": 97280
+    },
+    {
+      "epoch": 4.8321247640806595,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00041343399225191225,
+      "loss": 0.5042,
+      "step": 97290
+    },
+    {
+      "epoch": 4.832621436376279,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004133942584682626,
+      "loss": 0.5037,
+      "step": 97300
+    },
+    {
+      "epoch": 4.833118108671898,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0004133545246846131,
+      "loss": 0.5008,
+      "step": 97310
+    },
+    {
+      "epoch": 4.833614780967517,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0004133147909009636,
+      "loss": 0.5325,
+      "step": 97320
+    },
+    {
+      "epoch": 4.834111453263137,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.000413275057117314,
+      "loss": 0.4981,
+      "step": 97330
+    },
+    {
+      "epoch": 4.8346081255587565,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00041323532333366445,
+      "loss": 0.5227,
+      "step": 97340
+    },
+    {
+      "epoch": 4.835104797854376,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0004131955895500149,
+      "loss": 0.5002,
+      "step": 97350
+    },
+    {
+      "epoch": 4.835601470149995,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00041315585576636533,
+      "loss": 0.5431,
+      "step": 97360
+    },
+    {
+      "epoch": 4.836098142445614,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004131161219827158,
+      "loss": 0.5112,
+      "step": 97370
+    },
+    {
+      "epoch": 4.836594814741233,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004130763881990663,
+      "loss": 0.5155,
+      "step": 97380
+    },
+    {
+      "epoch": 4.837091487036853,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00041303665441541675,
+      "loss": 0.5001,
+      "step": 97390
+    },
+    {
+      "epoch": 4.837588159332473,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00041299692063176717,
+      "loss": 0.499,
+      "step": 97400
+    },
+    {
+      "epoch": 4.838084831628092,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00041295718684811764,
+      "loss": 0.5312,
+      "step": 97410
+    },
+    {
+      "epoch": 4.838581503923711,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0004129174530644681,
+      "loss": 0.5056,
+      "step": 97420
+    },
+    {
+      "epoch": 4.8390781762193305,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0004128777192808185,
+      "loss": 0.4915,
+      "step": 97430
+    },
+    {
+      "epoch": 4.83957484851495,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.000412837985497169,
+      "loss": 0.504,
+      "step": 97440
+    },
+    {
+      "epoch": 4.840071520810569,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00041279825171351947,
+      "loss": 0.5052,
+      "step": 97450
+    },
+    {
+      "epoch": 4.840568193106188,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0004127585179298699,
+      "loss": 0.5274,
+      "step": 97460
+    },
+    {
+      "epoch": 4.841064865401808,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00041271878414622036,
+      "loss": 0.4997,
+      "step": 97470
+    },
+    {
+      "epoch": 4.8415615376974275,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00041267905036257083,
+      "loss": 0.4722,
+      "step": 97480
+    },
+    {
+      "epoch": 4.842058209993047,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0004126393165789212,
+      "loss": 0.5154,
+      "step": 97490
+    },
+    {
+      "epoch": 4.842554882288666,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00041259958279527166,
+      "loss": 0.4914,
+      "step": 97500
+    },
+    {
+      "epoch": 4.843051554584285,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0004125598490116222,
+      "loss": 0.5148,
+      "step": 97510
+    },
+    {
+      "epoch": 4.843548226879904,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00041252011522797266,
+      "loss": 0.5141,
+      "step": 97520
+    },
+    {
+      "epoch": 4.844044899175524,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.000412480381444323,
+      "loss": 0.5172,
+      "step": 97530
+    },
+    {
+      "epoch": 4.844541571471144,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004124406476606735,
+      "loss": 0.5127,
+      "step": 97540
+    },
+    {
+      "epoch": 4.845038243766763,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.000412400913877024,
+      "loss": 0.4933,
+      "step": 97550
+    },
+    {
+      "epoch": 4.845534916062382,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004123611800933744,
+      "loss": 0.503,
+      "step": 97560
+    },
+    {
+      "epoch": 4.846031588358001,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.00041232144630972485,
+      "loss": 0.5205,
+      "step": 97570
+    },
+    {
+      "epoch": 4.846528260653621,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0004122817125260753,
+      "loss": 0.5166,
+      "step": 97580
+    },
+    {
+      "epoch": 4.84702493294924,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00041224197874242574,
+      "loss": 0.5364,
+      "step": 97590
+    },
+    {
+      "epoch": 4.847521605244859,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004122022449587762,
+      "loss": 0.5433,
+      "step": 97600
+    },
+    {
+      "epoch": 4.848018277540479,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0004121625111751267,
+      "loss": 0.5206,
+      "step": 97610
+    },
+    {
+      "epoch": 4.848514949836098,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004121227773914771,
+      "loss": 0.5149,
+      "step": 97620
+    },
+    {
+      "epoch": 4.849011622131718,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00041208304360782757,
+      "loss": 0.5024,
+      "step": 97630
+    },
+    {
+      "epoch": 4.849508294427337,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00041204330982417804,
+      "loss": 0.539,
+      "step": 97640
+    },
+    {
+      "epoch": 4.850004966722956,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00041200357604052846,
+      "loss": 0.5152,
+      "step": 97650
+    },
+    {
+      "epoch": 4.850501639018575,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00041196384225687893,
+      "loss": 0.5044,
+      "step": 97660
+    },
+    {
+      "epoch": 4.850998311314195,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0004119241084732294,
+      "loss": 0.4817,
+      "step": 97670
+    },
+    {
+      "epoch": 4.851494983609815,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0004118843746895799,
+      "loss": 0.5401,
+      "step": 97680
+    },
+    {
+      "epoch": 4.851991655905434,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004118446409059303,
+      "loss": 0.5415,
+      "step": 97690
+    },
+    {
+      "epoch": 4.852488328201053,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00041180490712228076,
+      "loss": 0.5361,
+      "step": 97700
+    },
+    {
+      "epoch": 4.852985000496672,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00041176517333863123,
+      "loss": 0.5263,
+      "step": 97710
+    },
+    {
+      "epoch": 4.853481672792292,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004117254395549816,
+      "loss": 0.5104,
+      "step": 97720
+    },
+    {
+      "epoch": 4.853978345087911,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0004116857057713321,
+      "loss": 0.5278,
+      "step": 97730
+    },
+    {
+      "epoch": 4.85447501738353,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0004116459719876826,
+      "loss": 0.5389,
+      "step": 97740
+    },
+    {
+      "epoch": 4.85497168967915,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00041160623820403296,
+      "loss": 0.5303,
+      "step": 97750
+    },
+    {
+      "epoch": 4.855468361974769,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00041156650442038343,
+      "loss": 0.5347,
+      "step": 97760
+    },
+    {
+      "epoch": 4.855965034270389,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0004115267706367339,
+      "loss": 0.5075,
+      "step": 97770
+    },
+    {
+      "epoch": 4.856461706566008,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0004114870368530843,
+      "loss": 0.5044,
+      "step": 97780
+    },
+    {
+      "epoch": 4.856958378861627,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004114473030694348,
+      "loss": 0.5152,
+      "step": 97790
+    },
+    {
+      "epoch": 4.857455051157246,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00041140756928578526,
+      "loss": 0.4994,
+      "step": 97800
+    },
+    {
+      "epoch": 4.8579517234528655,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0004113678355021357,
+      "loss": 0.5332,
+      "step": 97810
+    },
+    {
+      "epoch": 4.858448395748486,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00041132810171848615,
+      "loss": 0.5236,
+      "step": 97820
+    },
+    {
+      "epoch": 4.858945068044105,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004112883679348366,
+      "loss": 0.5387,
+      "step": 97830
+    },
+    {
+      "epoch": 4.859441740339724,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0004112486341511871,
+      "loss": 0.4905,
+      "step": 97840
+    },
+    {
+      "epoch": 4.859938412635343,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004112089003675375,
+      "loss": 0.4935,
+      "step": 97850
+    },
+    {
+      "epoch": 4.8604350849309625,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.000411169166583888,
+      "loss": 0.5209,
+      "step": 97860
+    },
+    {
+      "epoch": 4.860931757226582,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00041112943280023845,
+      "loss": 0.5148,
+      "step": 97870
+    },
+    {
+      "epoch": 4.861428429522201,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00041108969901658887,
+      "loss": 0.4997,
+      "step": 97880
+    },
+    {
+      "epoch": 4.86192510181782,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00041104996523293934,
+      "loss": 0.5335,
+      "step": 97890
+    },
+    {
+      "epoch": 4.8624217741134395,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0004110102314492898,
+      "loss": 0.5021,
+      "step": 97900
+    },
+    {
+      "epoch": 4.86291844640906,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0004109704976656402,
+      "loss": 0.521,
+      "step": 97910
+    },
+    {
+      "epoch": 4.863415118704679,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0004109307638819907,
+      "loss": 0.522,
+      "step": 97920
+    },
+    {
+      "epoch": 4.863911791000298,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00041089103009834117,
+      "loss": 0.5152,
+      "step": 97930
+    },
+    {
+      "epoch": 4.864408463295917,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00041085129631469153,
+      "loss": 0.4814,
+      "step": 97940
+    },
+    {
+      "epoch": 4.8649051355915365,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.000410811562531042,
+      "loss": 0.5029,
+      "step": 97950
+    },
+    {
+      "epoch": 4.865401807887156,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.00041077182874739253,
+      "loss": 0.4984,
+      "step": 97960
+    },
+    {
+      "epoch": 4.865898480182775,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.000410732094963743,
+      "loss": 0.5218,
+      "step": 97970
+    },
+    {
+      "epoch": 4.866395152478395,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00041069236118009337,
+      "loss": 0.5101,
+      "step": 97980
+    },
+    {
+      "epoch": 4.866891824774014,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00041065262739644384,
+      "loss": 0.5146,
+      "step": 97990
+    },
+    {
+      "epoch": 4.8673884970696335,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00041061289361279436,
+      "loss": 0.5067,
+      "step": 98000
+    },
+    {
+      "epoch": 4.867885169365253,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0004105731598291447,
+      "loss": 0.4766,
+      "step": 98010
+    },
+    {
+      "epoch": 4.868381841660872,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004105334260454952,
+      "loss": 0.4903,
+      "step": 98020
+    },
+    {
+      "epoch": 4.868878513956491,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00041049369226184567,
+      "loss": 0.5307,
+      "step": 98030
+    },
+    {
+      "epoch": 4.86937518625211,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0004104539584781961,
+      "loss": 0.5246,
+      "step": 98040
+    },
+    {
+      "epoch": 4.8698718585477305,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00041041422469454656,
+      "loss": 0.5032,
+      "step": 98050
+    },
+    {
+      "epoch": 4.87036853084335,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00041037449091089703,
+      "loss": 0.5249,
+      "step": 98060
+    },
+    {
+      "epoch": 4.870865203138969,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00041033475712724744,
+      "loss": 0.5255,
+      "step": 98070
+    },
+    {
+      "epoch": 4.871361875434588,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0004102950233435979,
+      "loss": 0.5082,
+      "step": 98080
+    },
+    {
+      "epoch": 4.871858547730207,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0004102552895599484,
+      "loss": 0.5224,
+      "step": 98090
+    },
+    {
+      "epoch": 4.872355220025827,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00041021555577629875,
+      "loss": 0.5227,
+      "step": 98100
+    },
+    {
+      "epoch": 4.872851892321446,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0004101758219926493,
+      "loss": 0.5188,
+      "step": 98110
+    },
+    {
+      "epoch": 4.873348564617066,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00041013608820899975,
+      "loss": 0.4935,
+      "step": 98120
+    },
+    {
+      "epoch": 4.873845236912685,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0004100963544253502,
+      "loss": 0.5272,
+      "step": 98130
+    },
+    {
+      "epoch": 4.8743419092083045,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0004100566206417006,
+      "loss": 0.5161,
+      "step": 98140
+    },
+    {
+      "epoch": 4.874838581503924,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0004100168868580511,
+      "loss": 0.4965,
+      "step": 98150
+    },
+    {
+      "epoch": 4.875335253799543,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0004099771530744016,
+      "loss": 0.4862,
+      "step": 98160
+    },
+    {
+      "epoch": 4.875831926095162,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00040993741929075194,
+      "loss": 0.5286,
+      "step": 98170
+    },
+    {
+      "epoch": 4.876328598390781,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0004098976855071024,
+      "loss": 0.5334,
+      "step": 98180
+    },
+    {
+      "epoch": 4.8768252706864015,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00040985795172345294,
+      "loss": 0.5077,
+      "step": 98190
+    },
+    {
+      "epoch": 4.877321942982021,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0004098182179398033,
+      "loss": 0.5194,
+      "step": 98200
+    },
+    {
+      "epoch": 4.87781861527764,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00040977848415615377,
+      "loss": 0.5228,
+      "step": 98210
+    },
+    {
+      "epoch": 4.878315287573259,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00040973875037250424,
+      "loss": 0.5258,
+      "step": 98220
+    },
+    {
+      "epoch": 4.878811959868878,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00040969901658885466,
+      "loss": 0.5157,
+      "step": 98230
+    },
+    {
+      "epoch": 4.879308632164498,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00040965928280520513,
+      "loss": 0.5134,
+      "step": 98240
+    },
+    {
+      "epoch": 4.879805304460117,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0004096195490215556,
+      "loss": 0.5122,
+      "step": 98250
+    },
+    {
+      "epoch": 4.880301976755737,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.000409579815237906,
+      "loss": 0.5279,
+      "step": 98260
+    },
+    {
+      "epoch": 4.880798649051356,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0004095400814542565,
+      "loss": 0.5175,
+      "step": 98270
+    },
+    {
+      "epoch": 4.881295321346975,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00040950034767060696,
+      "loss": 0.5035,
+      "step": 98280
+    },
+    {
+      "epoch": 4.881791993642595,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00040946061388695744,
+      "loss": 0.4978,
+      "step": 98290
+    },
+    {
+      "epoch": 4.882288665938214,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00040942088010330785,
+      "loss": 0.4894,
+      "step": 98300
+    },
+    {
+      "epoch": 4.882785338233833,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004093811463196583,
+      "loss": 0.5125,
+      "step": 98310
+    },
+    {
+      "epoch": 4.883282010529452,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004093414125360088,
+      "loss": 0.5366,
+      "step": 98320
+    },
+    {
+      "epoch": 4.883778682825072,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0004093016787523592,
+      "loss": 0.509,
+      "step": 98330
+    },
+    {
+      "epoch": 4.884275355120692,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0004092619449687097,
+      "loss": 0.5094,
+      "step": 98340
+    },
+    {
+      "epoch": 4.884772027416311,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00040922221118506015,
+      "loss": 0.4979,
+      "step": 98350
+    },
+    {
+      "epoch": 4.88526869971193,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0004091824774014105,
+      "loss": 0.4915,
+      "step": 98360
+    },
+    {
+      "epoch": 4.885765372007549,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.000409142743617761,
+      "loss": 0.5128,
+      "step": 98370
+    },
+    {
+      "epoch": 4.886262044303169,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0004091030098341115,
+      "loss": 0.5414,
+      "step": 98380
+    },
+    {
+      "epoch": 4.886758716598788,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0004090632760504619,
+      "loss": 0.4982,
+      "step": 98390
+    },
+    {
+      "epoch": 4.887255388894408,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00040902354226681235,
+      "loss": 0.4913,
+      "step": 98400
+    },
+    {
+      "epoch": 4.887752061190027,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0004089838084831628,
+      "loss": 0.5041,
+      "step": 98410
+    },
+    {
+      "epoch": 4.888248733485646,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00040894407469951335,
+      "loss": 0.5105,
+      "step": 98420
+    },
+    {
+      "epoch": 4.888745405781266,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0004089043409158637,
+      "loss": 0.5313,
+      "step": 98430
+    },
+    {
+      "epoch": 4.889242078076885,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0004088646071322142,
+      "loss": 0.5135,
+      "step": 98440
+    },
+    {
+      "epoch": 4.889738750372504,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00040882487334856465,
+      "loss": 0.5178,
+      "step": 98450
+    },
+    {
+      "epoch": 4.890235422668123,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00040878513956491507,
+      "loss": 0.5356,
+      "step": 98460
+    },
+    {
+      "epoch": 4.890732094963743,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00040874540578126554,
+      "loss": 0.5044,
+      "step": 98470
+    },
+    {
+      "epoch": 4.891228767259363,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.000408705671997616,
+      "loss": 0.495,
+      "step": 98480
+    },
+    {
+      "epoch": 4.891725439554982,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00040866593821396643,
+      "loss": 0.5322,
+      "step": 98490
+    },
+    {
+      "epoch": 4.892222111850601,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0004086262044303169,
+      "loss": 0.5052,
+      "step": 98500
+    },
+    {
+      "epoch": 4.89271878414622,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00040858647064666737,
+      "loss": 0.4894,
+      "step": 98510
+    },
+    {
+      "epoch": 4.8932154564418395,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0004085467368630178,
+      "loss": 0.5148,
+      "step": 98520
+    },
+    {
+      "epoch": 4.893712128737459,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00040850700307936826,
+      "loss": 0.5268,
+      "step": 98530
+    },
+    {
+      "epoch": 4.894208801033079,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00040846726929571873,
+      "loss": 0.4975,
+      "step": 98540
+    },
+    {
+      "epoch": 4.894705473328698,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004084275355120691,
+      "loss": 0.4836,
+      "step": 98550
+    },
+    {
+      "epoch": 4.895202145624317,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0004083878017284196,
+      "loss": 0.5274,
+      "step": 98560
+    },
+    {
+      "epoch": 4.8956988179199366,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0004083480679447701,
+      "loss": 0.52,
+      "step": 98570
+    },
+    {
+      "epoch": 4.896195490215556,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00040830833416112056,
+      "loss": 0.5171,
+      "step": 98580
+    },
+    {
+      "epoch": 4.896692162511175,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0004082686003774709,
+      "loss": 0.5075,
+      "step": 98590
+    },
+    {
+      "epoch": 4.897188834806794,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00040822886659382145,
+      "loss": 0.5407,
+      "step": 98600
+    },
+    {
+      "epoch": 4.8976855071024135,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0004081891328101719,
+      "loss": 0.5008,
+      "step": 98610
+    },
+    {
+      "epoch": 4.898182179398033,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0004081493990265223,
+      "loss": 0.5087,
+      "step": 98620
+    },
+    {
+      "epoch": 4.898678851693653,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00040810966524287276,
+      "loss": 0.52,
+      "step": 98630
+    },
+    {
+      "epoch": 4.899175523989272,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00040806993145922323,
+      "loss": 0.5035,
+      "step": 98640
+    },
+    {
+      "epoch": 4.899672196284891,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00040803019767557365,
+      "loss": 0.5082,
+      "step": 98650
+    },
+    {
+      "epoch": 4.9001688685805105,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0004079904638919241,
+      "loss": 0.4911,
+      "step": 98660
+    },
+    {
+      "epoch": 4.90066554087613,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0004079507301082746,
+      "loss": 0.5165,
+      "step": 98670
+    },
+    {
+      "epoch": 4.901162213171749,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.000407910996324625,
+      "loss": 0.5235,
+      "step": 98680
+    },
+    {
+      "epoch": 4.901658885467368,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0004078712625409755,
+      "loss": 0.5162,
+      "step": 98690
+    },
+    {
+      "epoch": 4.902155557762988,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00040783152875732595,
+      "loss": 0.5158,
+      "step": 98700
+    },
+    {
+      "epoch": 4.9026522300586075,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00040779179497367636,
+      "loss": 0.5105,
+      "step": 98710
+    },
+    {
+      "epoch": 4.903148902354227,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00040775206119002684,
+      "loss": 0.4942,
+      "step": 98720
+    },
+    {
+      "epoch": 4.903645574649846,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0004077123274063773,
+      "loss": 0.5166,
+      "step": 98730
+    },
+    {
+      "epoch": 4.904142246945465,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0004076725936227278,
+      "loss": 0.5246,
+      "step": 98740
+    },
+    {
+      "epoch": 4.904638919241084,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0004076328598390782,
+      "loss": 0.5428,
+      "step": 98750
+    },
+    {
+      "epoch": 4.905135591536704,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00040759312605542867,
+      "loss": 0.5142,
+      "step": 98760
+    },
+    {
+      "epoch": 4.905632263832324,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00040755339227177914,
+      "loss": 0.5362,
+      "step": 98770
+    },
+    {
+      "epoch": 4.906128936127943,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0004075136584881295,
+      "loss": 0.5124,
+      "step": 98780
+    },
+    {
+      "epoch": 4.906625608423562,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00040747392470448003,
+      "loss": 0.5158,
+      "step": 98790
+    },
+    {
+      "epoch": 4.907122280719181,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0004074341909208305,
+      "loss": 0.4925,
+      "step": 98800
+    },
+    {
+      "epoch": 4.907618953014801,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00040739445713718086,
+      "loss": 0.5157,
+      "step": 98810
+    },
+    {
+      "epoch": 4.90811562531042,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.00040735472335353133,
+      "loss": 0.4888,
+      "step": 98820
+    },
+    {
+      "epoch": 4.908612297606039,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00040731498956988186,
+      "loss": 0.5231,
+      "step": 98830
+    },
+    {
+      "epoch": 4.909108969901659,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004072752557862322,
+      "loss": 0.503,
+      "step": 98840
+    },
+    {
+      "epoch": 4.9096056421972785,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004072355220025827,
+      "loss": 0.4851,
+      "step": 98850
+    },
+    {
+      "epoch": 4.910102314492898,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00040719578821893316,
+      "loss": 0.5166,
+      "step": 98860
+    },
+    {
+      "epoch": 4.910598986788517,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0004071560544352837,
+      "loss": 0.51,
+      "step": 98870
+    },
+    {
+      "epoch": 4.911095659084136,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00040711632065163405,
+      "loss": 0.5411,
+      "step": 98880
+    },
+    {
+      "epoch": 4.911592331379755,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0004070765868679845,
+      "loss": 0.4917,
+      "step": 98890
+    },
+    {
+      "epoch": 4.912089003675375,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.000407036853084335,
+      "loss": 0.5179,
+      "step": 98900
+    },
+    {
+      "epoch": 4.912585675970995,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004069971193006854,
+      "loss": 0.5128,
+      "step": 98910
+    },
+    {
+      "epoch": 4.913082348266614,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004069573855170359,
+      "loss": 0.508,
+      "step": 98920
+    },
+    {
+      "epoch": 4.913579020562233,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00040691765173338636,
+      "loss": 0.5038,
+      "step": 98930
+    },
+    {
+      "epoch": 4.914075692857852,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00040687791794973677,
+      "loss": 0.5337,
+      "step": 98940
+    },
+    {
+      "epoch": 4.914572365153472,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.00040683818416608724,
+      "loss": 0.4889,
+      "step": 98950
+    },
+    {
+      "epoch": 4.915069037449091,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0004067984503824377,
+      "loss": 0.5286,
+      "step": 98960
+    },
+    {
+      "epoch": 4.91556570974471,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0004067587165987881,
+      "loss": 0.5046,
+      "step": 98970
+    },
+    {
+      "epoch": 4.91606238204033,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0004067189828151386,
+      "loss": 0.4965,
+      "step": 98980
+    },
+    {
+      "epoch": 4.916559054335949,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0004066792490314891,
+      "loss": 0.5457,
+      "step": 98990
+    },
+    {
+      "epoch": 4.917055726631569,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00040663951524783944,
+      "loss": 0.5051,
+      "step": 99000
+    },
+    {
+      "epoch": 4.917552398927188,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004065997814641899,
+      "loss": 0.4822,
+      "step": 99010
+    },
+    {
+      "epoch": 4.918049071222807,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00040656004768054043,
+      "loss": 0.5285,
+      "step": 99020
+    },
+    {
+      "epoch": 4.918545743518426,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0004065203138968909,
+      "loss": 0.4931,
+      "step": 99030
+    },
+    {
+      "epoch": 4.9190424158140456,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00040648058011324127,
+      "loss": 0.5481,
+      "step": 99040
+    },
+    {
+      "epoch": 4.919539088109666,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00040644084632959174,
+      "loss": 0.5057,
+      "step": 99050
+    },
+    {
+      "epoch": 4.920035760405285,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00040640111254594227,
+      "loss": 0.4832,
+      "step": 99060
+    },
+    {
+      "epoch": 4.920532432700904,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00040636137876229263,
+      "loss": 0.5402,
+      "step": 99070
+    },
+    {
+      "epoch": 4.921029104996523,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0004063216449786431,
+      "loss": 0.5001,
+      "step": 99080
+    },
+    {
+      "epoch": 4.921525777292143,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00040628191119499357,
+      "loss": 0.5159,
+      "step": 99090
+    },
+    {
+      "epoch": 4.922022449587762,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.000406242177411344,
+      "loss": 0.5343,
+      "step": 99100
+    },
+    {
+      "epoch": 4.922519121883381,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00040620244362769446,
+      "loss": 0.5164,
+      "step": 99110
+    },
+    {
+      "epoch": 4.923015794179001,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.00040616270984404493,
+      "loss": 0.498,
+      "step": 99120
+    },
+    {
+      "epoch": 4.92351246647462,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00040612297606039535,
+      "loss": 0.5175,
+      "step": 99130
+    },
+    {
+      "epoch": 4.92400913877024,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004060832422767458,
+      "loss": 0.5302,
+      "step": 99140
+    },
+    {
+      "epoch": 4.924505811065859,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0004060435084930963,
+      "loss": 0.5007,
+      "step": 99150
+    },
+    {
+      "epoch": 4.925002483361478,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0004060037747094467,
+      "loss": 0.5229,
+      "step": 99160
+    },
+    {
+      "epoch": 4.925499155657097,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0004059640409257972,
+      "loss": 0.519,
+      "step": 99170
+    },
+    {
+      "epoch": 4.9259958279527165,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00040592430714214765,
+      "loss": 0.5088,
+      "step": 99180
+    },
+    {
+      "epoch": 4.926492500248337,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004058845733584981,
+      "loss": 0.5215,
+      "step": 99190
+    },
+    {
+      "epoch": 4.926989172543956,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0004058448395748485,
+      "loss": 0.5158,
+      "step": 99200
+    },
+    {
+      "epoch": 4.927485844839575,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.000405805105791199,
+      "loss": 0.5514,
+      "step": 99210
+    },
+    {
+      "epoch": 4.927982517135194,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0004057653720075495,
+      "loss": 0.5251,
+      "step": 99220
+    },
+    {
+      "epoch": 4.9284791894308135,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00040572563822389985,
+      "loss": 0.4987,
+      "step": 99230
+    },
+    {
+      "epoch": 4.928975861726433,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0004056859044402503,
+      "loss": 0.495,
+      "step": 99240
+    },
+    {
+      "epoch": 4.929472534022052,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00040564617065660084,
+      "loss": 0.5459,
+      "step": 99250
+    },
+    {
+      "epoch": 4.929969206317671,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004056064368729512,
+      "loss": 0.5065,
+      "step": 99260
+    },
+    {
+      "epoch": 4.930465878613291,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0004055667030893017,
+      "loss": 0.5229,
+      "step": 99270
+    },
+    {
+      "epoch": 4.930962550908911,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00040552696930565215,
+      "loss": 0.4875,
+      "step": 99280
+    },
+    {
+      "epoch": 4.93145922320453,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00040548723552200256,
+      "loss": 0.5139,
+      "step": 99290
+    },
+    {
+      "epoch": 4.931955895500149,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00040544750173835304,
+      "loss": 0.528,
+      "step": 99300
+    },
+    {
+      "epoch": 4.932452567795768,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0004054077679547035,
+      "loss": 0.5157,
+      "step": 99310
+    },
+    {
+      "epoch": 4.9329492400913875,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.000405368034171054,
+      "loss": 0.5011,
+      "step": 99320
+    },
+    {
+      "epoch": 4.933445912387007,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004053283003874044,
+      "loss": 0.4929,
+      "step": 99330
+    },
+    {
+      "epoch": 4.933942584682626,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00040528856660375487,
+      "loss": 0.4835,
+      "step": 99340
+    },
+    {
+      "epoch": 4.934439256978246,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00040524883282010534,
+      "loss": 0.5361,
+      "step": 99350
+    },
+    {
+      "epoch": 4.934935929273865,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00040520909903645576,
+      "loss": 0.5291,
+      "step": 99360
+    },
+    {
+      "epoch": 4.9354326015694845,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00040516936525280623,
+      "loss": 0.5118,
+      "step": 99370
+    },
+    {
+      "epoch": 4.935929273865104,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0004051296314691567,
+      "loss": 0.4971,
+      "step": 99380
+    },
+    {
+      "epoch": 4.936425946160723,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0004050898976855071,
+      "loss": 0.4903,
+      "step": 99390
+    },
+    {
+      "epoch": 4.936922618456342,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0004050501639018576,
+      "loss": 0.5122,
+      "step": 99400
+    },
+    {
+      "epoch": 4.937419290751961,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00040501043011820806,
+      "loss": 0.4923,
+      "step": 99410
+    },
+    {
+      "epoch": 4.9379159630475815,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0004049706963345584,
+      "loss": 0.5291,
+      "step": 99420
+    },
+    {
+      "epoch": 4.938412635343201,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00040493096255090895,
+      "loss": 0.5176,
+      "step": 99430
+    },
+    {
+      "epoch": 4.93890930763882,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0004048912287672594,
+      "loss": 0.5011,
+      "step": 99440
+    },
+    {
+      "epoch": 4.939405979934439,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0004048514949836098,
+      "loss": 0.501,
+      "step": 99450
+    },
+    {
+      "epoch": 4.939902652230058,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00040481176119996025,
+      "loss": 0.5367,
+      "step": 99460
+    },
+    {
+      "epoch": 4.940399324525678,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0004047720274163107,
+      "loss": 0.5023,
+      "step": 99470
+    },
+    {
+      "epoch": 4.940895996821297,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00040473229363266125,
+      "loss": 0.5122,
+      "step": 99480
+    },
+    {
+      "epoch": 4.941392669116917,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0004046925598490116,
+      "loss": 0.5038,
+      "step": 99490
+    },
+    {
+      "epoch": 4.941889341412536,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0004046528260653621,
+      "loss": 0.5243,
+      "step": 99500
+    },
+    {
+      "epoch": 4.9423860137081554,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00040461309228171256,
+      "loss": 0.5047,
+      "step": 99510
+    },
+    {
+      "epoch": 4.942882686003775,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00040457335849806297,
+      "loss": 0.5128,
+      "step": 99520
+    },
+    {
+      "epoch": 4.943379358299394,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00040453362471441344,
+      "loss": 0.5099,
+      "step": 99530
+    },
+    {
+      "epoch": 4.943876030595013,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0004044938909307639,
+      "loss": 0.5317,
+      "step": 99540
+    },
+    {
+      "epoch": 4.944372702890632,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00040445415714711433,
+      "loss": 0.5239,
+      "step": 99550
+    },
+    {
+      "epoch": 4.9448693751862525,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0004044144233634648,
+      "loss": 0.5218,
+      "step": 99560
+    },
+    {
+      "epoch": 4.945366047481872,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0004043746895798153,
+      "loss": 0.5068,
+      "step": 99570
+    },
+    {
+      "epoch": 4.945862719777491,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0004043349557961657,
+      "loss": 0.5081,
+      "step": 99580
+    },
+    {
+      "epoch": 4.94635939207311,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00040429522201251616,
+      "loss": 0.4895,
+      "step": 99590
+    },
+    {
+      "epoch": 4.946856064368729,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00040425548822886663,
+      "loss": 0.5348,
+      "step": 99600
+    },
+    {
+      "epoch": 4.947352736664349,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.000404215754445217,
+      "loss": 0.5263,
+      "step": 99610
+    },
+    {
+      "epoch": 4.947849408959968,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0004041760206615675,
+      "loss": 0.5113,
+      "step": 99620
+    },
+    {
+      "epoch": 4.948346081255588,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.000404136286877918,
+      "loss": 0.4971,
+      "step": 99630
+    },
+    {
+      "epoch": 4.948842753551207,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00040409655309426847,
+      "loss": 0.5496,
+      "step": 99640
+    },
+    {
+      "epoch": 4.949339425846826,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00040405681931061883,
+      "loss": 0.4924,
+      "step": 99650
+    },
+    {
+      "epoch": 4.949836098142446,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.00040401708552696935,
+      "loss": 0.4995,
+      "step": 99660
+    },
+    {
+      "epoch": 4.950332770438065,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0004039773517433198,
+      "loss": 0.5248,
+      "step": 99670
+    },
+    {
+      "epoch": 4.950829442733684,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0004039376179596702,
+      "loss": 0.5132,
+      "step": 99680
+    },
+    {
+      "epoch": 4.951326115029303,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00040389788417602066,
+      "loss": 0.5077,
+      "step": 99690
+    },
+    {
+      "epoch": 4.951822787324923,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0004038581503923712,
+      "loss": 0.4941,
+      "step": 99700
+    },
+    {
+      "epoch": 4.952319459620543,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00040381841660872155,
+      "loss": 0.5123,
+      "step": 99710
+    },
+    {
+      "epoch": 4.952816131916162,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.000403778682825072,
+      "loss": 0.4875,
+      "step": 99720
+    },
+    {
+      "epoch": 4.953312804211781,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0004037389490414225,
+      "loss": 0.5222,
+      "step": 99730
+    },
+    {
+      "epoch": 4.9538094765074,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0004036992152577729,
+      "loss": 0.487,
+      "step": 99740
+    },
+    {
+      "epoch": 4.95430614880302,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0004036594814741234,
+      "loss": 0.512,
+      "step": 99750
+    },
+    {
+      "epoch": 4.954802821098639,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00040361974769047385,
+      "loss": 0.5148,
+      "step": 99760
+    },
+    {
+      "epoch": 4.955299493394259,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0004035800139068243,
+      "loss": 0.4891,
+      "step": 99770
+    },
+    {
+      "epoch": 4.955796165689878,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00040354028012317474,
+      "loss": 0.5063,
+      "step": 99780
+    },
+    {
+      "epoch": 4.956292837985497,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0004035005463395252,
+      "loss": 0.5276,
+      "step": 99790
+    },
+    {
+      "epoch": 4.956789510281117,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004034608125558757,
+      "loss": 0.5103,
+      "step": 99800
+    },
+    {
+      "epoch": 4.957286182576736,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0004034210787722261,
+      "loss": 0.5315,
+      "step": 99810
+    },
+    {
+      "epoch": 4.957782854872355,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00040338134498857657,
+      "loss": 0.4878,
+      "step": 99820
+    },
+    {
+      "epoch": 4.958279527167974,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00040334161120492704,
+      "loss": 0.4885,
+      "step": 99830
+    },
+    {
+      "epoch": 4.958776199463594,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0004033018774212774,
+      "loss": 0.5389,
+      "step": 99840
+    },
+    {
+      "epoch": 4.959272871759214,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00040326214363762793,
+      "loss": 0.5157,
+      "step": 99850
+    },
+    {
+      "epoch": 4.959769544054833,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0004032224098539784,
+      "loss": 0.4998,
+      "step": 99860
+    },
+    {
+      "epoch": 4.960266216350452,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00040318267607032877,
+      "loss": 0.4896,
+      "step": 99870
+    },
+    {
+      "epoch": 4.960762888646071,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.00040314294228667924,
+      "loss": 0.5059,
+      "step": 99880
+    },
+    {
+      "epoch": 4.9612595609416905,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00040310320850302976,
+      "loss": 0.5358,
+      "step": 99890
+    },
+    {
+      "epoch": 4.96175623323731,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0004030634747193801,
+      "loss": 0.5355,
+      "step": 99900
+    },
+    {
+      "epoch": 4.96225290553293,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004030237409357306,
+      "loss": 0.5232,
+      "step": 99910
+    },
+    {
+      "epoch": 4.962749577828549,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00040298400715208107,
+      "loss": 0.5263,
+      "step": 99920
+    },
+    {
+      "epoch": 4.963246250124168,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0004029442733684316,
+      "loss": 0.5206,
+      "step": 99930
+    },
+    {
+      "epoch": 4.9637429224197875,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00040290453958478196,
+      "loss": 0.546,
+      "step": 99940
+    },
+    {
+      "epoch": 4.964239594715407,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.00040286480580113243,
+      "loss": 0.5026,
+      "step": 99950
+    },
+    {
+      "epoch": 4.964736267011026,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0004028250720174829,
+      "loss": 0.5343,
+      "step": 99960
+    },
+    {
+      "epoch": 4.965232939306645,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0004027853382338333,
+      "loss": 0.498,
+      "step": 99970
+    },
+    {
+      "epoch": 4.9657296116022644,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004027456044501838,
+      "loss": 0.498,
+      "step": 99980
+    },
+    {
+      "epoch": 4.966226283897885,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00040270587066653426,
+      "loss": 0.5245,
+      "step": 99990
+    },
+    {
+      "epoch": 4.966722956193504,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0004026661368828847,
+      "loss": 0.5091,
+      "step": 100000
+    },
+    {
+      "epoch": 4.967219628489123,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00040262640309923515,
+      "loss": 0.4981,
+      "step": 100010
+    },
+    {
+      "epoch": 4.967716300784742,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0004025866693155856,
+      "loss": 0.5341,
+      "step": 100020
+    },
+    {
+      "epoch": 4.9682129730803615,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00040254693553193604,
+      "loss": 0.5405,
+      "step": 100030
+    },
+    {
+      "epoch": 4.968709645375981,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0004025072017482865,
+      "loss": 0.4873,
+      "step": 100040
+    },
+    {
+      "epoch": 4.9692063176716,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.000402467467964637,
+      "loss": 0.5229,
+      "step": 100050
+    },
+    {
+      "epoch": 4.969702989967219,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00040242773418098745,
+      "loss": 0.5282,
+      "step": 100060
+    },
+    {
+      "epoch": 4.970199662262839,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0004023880003973378,
+      "loss": 0.5119,
+      "step": 100070
+    },
+    {
+      "epoch": 4.9706963345584585,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00040234826661368834,
+      "loss": 0.5049,
+      "step": 100080
+    },
+    {
+      "epoch": 4.971193006854078,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0004023085328300388,
+      "loss": 0.5449,
+      "step": 100090
+    },
+    {
+      "epoch": 4.971689679149697,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00040226879904638917,
+      "loss": 0.5154,
+      "step": 100100
+    },
+    {
+      "epoch": 4.972186351445316,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00040222906526273964,
+      "loss": 0.5116,
+      "step": 100110
+    },
+    {
+      "epoch": 4.972683023740935,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00040218933147909017,
+      "loss": 0.5156,
+      "step": 100120
+    },
+    {
+      "epoch": 4.973179696036555,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00040214959769544053,
+      "loss": 0.517,
+      "step": 100130
+    },
+    {
+      "epoch": 4.973676368332175,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.000402109863911791,
+      "loss": 0.4893,
+      "step": 100140
+    },
+    {
+      "epoch": 4.974173040627794,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0004020701301281415,
+      "loss": 0.5071,
+      "step": 100150
+    },
+    {
+      "epoch": 4.974669712923413,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0004020303963444919,
+      "loss": 0.5004,
+      "step": 100160
+    },
+    {
+      "epoch": 4.975166385219032,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00040199066256084236,
+      "loss": 0.538,
+      "step": 100170
+    },
+    {
+      "epoch": 4.975663057514652,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00040195092877719284,
+      "loss": 0.523,
+      "step": 100180
+    },
+    {
+      "epoch": 4.976159729810271,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00040191119499354325,
+      "loss": 0.5077,
+      "step": 100190
+    },
+    {
+      "epoch": 4.97665640210589,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0004018714612098937,
+      "loss": 0.5341,
+      "step": 100200
+    },
+    {
+      "epoch": 4.97715307440151,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0004018317274262442,
+      "loss": 0.5223,
+      "step": 100210
+    },
+    {
+      "epoch": 4.9776497466971295,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00040179199364259467,
+      "loss": 0.4992,
+      "step": 100220
+    },
+    {
+      "epoch": 4.978146418992749,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0004017522598589451,
+      "loss": 0.5209,
+      "step": 100230
+    },
+    {
+      "epoch": 4.978643091288368,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00040171252607529555,
+      "loss": 0.5051,
+      "step": 100240
+    },
+    {
+      "epoch": 4.979139763583987,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.000401672792291646,
+      "loss": 0.5268,
+      "step": 100250
+    },
+    {
+      "epoch": 4.979636435879606,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00040163305850799644,
+      "loss": 0.5149,
+      "step": 100260
+    },
+    {
+      "epoch": 4.980133108175226,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0004015933247243469,
+      "loss": 0.5173,
+      "step": 100270
+    },
+    {
+      "epoch": 4.980629780470846,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0004015535909406974,
+      "loss": 0.4917,
+      "step": 100280
+    },
+    {
+      "epoch": 4.981126452766465,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00040151385715704775,
+      "loss": 0.5374,
+      "step": 100290
+    },
+    {
+      "epoch": 4.981623125062084,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0004014741233733983,
+      "loss": 0.5289,
+      "step": 100300
+    },
+    {
+      "epoch": 4.982119797357703,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00040143438958974875,
+      "loss": 0.5264,
+      "step": 100310
+    },
+    {
+      "epoch": 4.982616469653323,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0004013946558060991,
+      "loss": 0.5166,
+      "step": 100320
+    },
+    {
+      "epoch": 4.983113141948942,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0004013549220224496,
+      "loss": 0.5035,
+      "step": 100330
+    },
+    {
+      "epoch": 4.983609814244561,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00040131518823880005,
+      "loss": 0.5512,
+      "step": 100340
+    },
+    {
+      "epoch": 4.984106486540181,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00040127545445515047,
+      "loss": 0.5229,
+      "step": 100350
+    },
+    {
+      "epoch": 4.9846031588358,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00040123572067150094,
+      "loss": 0.5159,
+      "step": 100360
+    },
+    {
+      "epoch": 4.98509983113142,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0004011959868878514,
+      "loss": 0.5254,
+      "step": 100370
+    },
+    {
+      "epoch": 4.985596503427039,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0004011562531042019,
+      "loss": 0.5119,
+      "step": 100380
+    },
+    {
+      "epoch": 4.986093175722658,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0004011165193205523,
+      "loss": 0.5259,
+      "step": 100390
+    },
+    {
+      "epoch": 4.986589848018277,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00040107678553690277,
+      "loss": 0.5211,
+      "step": 100400
+    },
+    {
+      "epoch": 4.9870865203138965,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00040103705175325324,
+      "loss": 0.5141,
+      "step": 100410
+    },
+    {
+      "epoch": 4.987583192609517,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00040099731796960366,
+      "loss": 0.5036,
+      "step": 100420
+    },
+    {
+      "epoch": 4.988079864905136,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00040095758418595413,
+      "loss": 0.5237,
+      "step": 100430
+    },
+    {
+      "epoch": 4.988576537200755,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0004009178504023046,
+      "loss": 0.4908,
+      "step": 100440
+    },
+    {
+      "epoch": 4.989073209496374,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.000400878116618655,
+      "loss": 0.5049,
+      "step": 100450
+    },
+    {
+      "epoch": 4.989569881791994,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0004008383828350055,
+      "loss": 0.5167,
+      "step": 100460
+    },
+    {
+      "epoch": 4.990066554087613,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00040079864905135596,
+      "loss": 0.4888,
+      "step": 100470
+    },
+    {
+      "epoch": 4.990563226383232,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0004007589152677063,
+      "loss": 0.5351,
+      "step": 100480
+    },
+    {
+      "epoch": 4.991059898678852,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.00040071918148405685,
+      "loss": 0.5331,
+      "step": 100490
+    },
+    {
+      "epoch": 4.991556570974471,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0004006794477004073,
+      "loss": 0.5051,
+      "step": 100500
+    },
+    {
+      "epoch": 4.992053243270091,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0004006397139167578,
+      "loss": 0.5323,
+      "step": 100510
+    },
+    {
+      "epoch": 4.99254991556571,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00040059998013310816,
+      "loss": 0.5355,
+      "step": 100520
+    },
+    {
+      "epoch": 4.993046587861329,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0004005602463494587,
+      "loss": 0.5263,
+      "step": 100530
+    },
+    {
+      "epoch": 4.993543260156948,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00040052051256580915,
+      "loss": 0.5215,
+      "step": 100540
+    },
+    {
+      "epoch": 4.9940399324525675,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0004004807787821595,
+      "loss": 0.5193,
+      "step": 100550
+    },
+    {
+      "epoch": 4.994536604748188,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00040044104499851,
+      "loss": 0.4774,
+      "step": 100560
+    },
+    {
+      "epoch": 4.995033277043807,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00040040131121486046,
+      "loss": 0.5122,
+      "step": 100570
+    },
+    {
+      "epoch": 4.995529949339426,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0004003615774312109,
+      "loss": 0.5145,
+      "step": 100580
+    },
+    {
+      "epoch": 4.996026621635045,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00040032184364756135,
+      "loss": 0.5119,
+      "step": 100590
+    },
+    {
+      "epoch": 4.9965232939306645,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0004002821098639118,
+      "loss": 0.5173,
+      "step": 100600
+    },
+    {
+      "epoch": 4.997019966226284,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00040024237608026224,
+      "loss": 0.5329,
+      "step": 100610
+    },
+    {
+      "epoch": 4.997516638521903,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0004002026422966127,
+      "loss": 0.5277,
+      "step": 100620
+    },
+    {
+      "epoch": 4.998013310817523,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0004001629085129632,
+      "loss": 0.5278,
+      "step": 100630
+    },
+    {
+      "epoch": 4.998509983113142,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0004001231747293136,
+      "loss": 0.514,
+      "step": 100640
+    },
+    {
+      "epoch": 4.9990066554087615,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00040008344094566407,
+      "loss": 0.51,
+      "step": 100650
+    },
+    {
+      "epoch": 4.999503327704381,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00040004370716201454,
+      "loss": 0.5158,
+      "step": 100660
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.000400003973378365,
+      "loss": 0.5036,
+      "step": 100670
+    },
+    {
+      "epoch": 5.000496672295619,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00039996423959471543,
+      "loss": 0.5218,
+      "step": 100680
+    },
+    {
+      "epoch": 5.0009933445912385,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0003999245058110659,
+      "loss": 0.5116,
+      "step": 100690
+    },
+    {
+      "epoch": 5.001490016886858,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0003998847720274163,
+      "loss": 0.4877,
+      "step": 100700
+    },
+    {
+      "epoch": 5.001986689182478,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003998450382437668,
+      "loss": 0.5158,
+      "step": 100710
+    },
+    {
+      "epoch": 5.002483361478097,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00039980530446011726,
+      "loss": 0.5278,
+      "step": 100720
+    },
+    {
+      "epoch": 5.002980033773716,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0003997655706764677,
+      "loss": 0.5015,
+      "step": 100730
+    },
+    {
+      "epoch": 5.0034767060693355,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00039972583689281815,
+      "loss": 0.4608,
+      "step": 100740
+    },
+    {
+      "epoch": 5.003973378364955,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00039968610310916856,
+      "loss": 0.5155,
+      "step": 100750
+    },
+    {
+      "epoch": 5.004470050660574,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00039964636932551904,
+      "loss": 0.5058,
+      "step": 100760
+    },
+    {
+      "epoch": 5.004966722956193,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0003996066355418695,
+      "loss": 0.4725,
+      "step": 100770
+    },
+    {
+      "epoch": 5.005463395251813,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003995669017582199,
+      "loss": 0.4929,
+      "step": 100780
+    },
+    {
+      "epoch": 5.0059600675474325,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0003995271679745704,
+      "loss": 0.5012,
+      "step": 100790
+    },
+    {
+      "epoch": 5.006456739843052,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00039948743419092087,
+      "loss": 0.4765,
+      "step": 100800
+    },
+    {
+      "epoch": 5.006953412138671,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0003994477004072713,
+      "loss": 0.4979,
+      "step": 100810
+    },
+    {
+      "epoch": 5.00745008443429,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00039940796662362176,
+      "loss": 0.5149,
+      "step": 100820
+    },
+    {
+      "epoch": 5.007946756729909,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003993682328399722,
+      "loss": 0.4895,
+      "step": 100830
+    },
+    {
+      "epoch": 5.008443429025529,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00039932849905632264,
+      "loss": 0.5264,
+      "step": 100840
+    },
+    {
+      "epoch": 5.008940101321148,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0003992887652726731,
+      "loss": 0.4954,
+      "step": 100850
+    },
+    {
+      "epoch": 5.009436773616768,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00039924903148902353,
+      "loss": 0.5156,
+      "step": 100860
+    },
+    {
+      "epoch": 5.009933445912387,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.000399209297705374,
+      "loss": 0.4886,
+      "step": 100870
+    },
+    {
+      "epoch": 5.010430118208006,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0003991695639217245,
+      "loss": 0.4991,
+      "step": 100880
+    },
+    {
+      "epoch": 5.010926790503626,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0003991298301380749,
+      "loss": 0.4754,
+      "step": 100890
+    },
+    {
+      "epoch": 5.011423462799245,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00039909009635442536,
+      "loss": 0.5036,
+      "step": 100900
+    },
+    {
+      "epoch": 5.011920135094864,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00039905036257077583,
+      "loss": 0.5072,
+      "step": 100910
+    },
+    {
+      "epoch": 5.012416807390483,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0003990106287871263,
+      "loss": 0.5208,
+      "step": 100920
+    },
+    {
+      "epoch": 5.0129134796861035,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003989708950034767,
+      "loss": 0.5063,
+      "step": 100930
+    },
+    {
+      "epoch": 5.013410151981723,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00039893116121982714,
+      "loss": 0.5126,
+      "step": 100940
+    },
+    {
+      "epoch": 5.013906824277342,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00039889142743617767,
+      "loss": 0.501,
+      "step": 100950
+    },
+    {
+      "epoch": 5.014403496572961,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003988516936525281,
+      "loss": 0.5127,
+      "step": 100960
+    },
+    {
+      "epoch": 5.01490016886858,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0003988119598688785,
+      "loss": 0.4887,
+      "step": 100970
+    },
+    {
+      "epoch": 5.0153968411642,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00039877222608522897,
+      "loss": 0.5193,
+      "step": 100980
+    },
+    {
+      "epoch": 5.015893513459819,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00039873249230157944,
+      "loss": 0.5054,
+      "step": 100990
+    },
+    {
+      "epoch": 5.016390185755439,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0003986927585179299,
+      "loss": 0.4664,
+      "step": 101000
+    },
+    {
+      "epoch": 5.016886858051058,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00039865302473428033,
+      "loss": 0.4959,
+      "step": 101010
+    },
+    {
+      "epoch": 5.017383530346677,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0003986132909506308,
+      "loss": 0.5126,
+      "step": 101020
+    },
+    {
+      "epoch": 5.017880202642297,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0003985735571669813,
+      "loss": 0.4981,
+      "step": 101030
+    },
+    {
+      "epoch": 5.018376874937916,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0003985338233833317,
+      "loss": 0.5043,
+      "step": 101040
+    },
+    {
+      "epoch": 5.018873547233535,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003984940895996821,
+      "loss": 0.5113,
+      "step": 101050
+    },
+    {
+      "epoch": 5.019370219529154,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00039845435581603263,
+      "loss": 0.5359,
+      "step": 101060
+    },
+    {
+      "epoch": 5.019866891824774,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00039841462203238305,
+      "loss": 0.4985,
+      "step": 101070
+    },
+    {
+      "epoch": 5.020363564120394,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0003983748882487335,
+      "loss": 0.4846,
+      "step": 101080
+    },
+    {
+      "epoch": 5.020860236416013,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00039833515446508394,
+      "loss": 0.5227,
+      "step": 101090
+    },
+    {
+      "epoch": 5.021356908711632,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003982954206814344,
+      "loss": 0.5161,
+      "step": 101100
+    },
+    {
+      "epoch": 5.021853581007251,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003982556868977849,
+      "loss": 0.4944,
+      "step": 101110
+    },
+    {
+      "epoch": 5.0223502533028705,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0003982159531141353,
+      "loss": 0.5032,
+      "step": 101120
+    },
+    {
+      "epoch": 5.02284692559849,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00039817621933048577,
+      "loss": 0.5054,
+      "step": 101130
+    },
+    {
+      "epoch": 5.02334359789411,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00039813648554683624,
+      "loss": 0.5094,
+      "step": 101140
+    },
+    {
+      "epoch": 5.023840270189729,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00039809675176318666,
+      "loss": 0.4775,
+      "step": 101150
+    },
+    {
+      "epoch": 5.024336942485348,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00039805701797953713,
+      "loss": 0.4934,
+      "step": 101160
+    },
+    {
+      "epoch": 5.024833614780968,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00039801728419588755,
+      "loss": 0.5035,
+      "step": 101170
+    },
+    {
+      "epoch": 5.025330287076587,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.000397977550412238,
+      "loss": 0.5133,
+      "step": 101180
+    },
+    {
+      "epoch": 5.025826959372206,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0003979378166285885,
+      "loss": 0.5045,
+      "step": 101190
+    },
+    {
+      "epoch": 5.026323631667825,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0003978980828449389,
+      "loss": 0.495,
+      "step": 101200
+    },
+    {
+      "epoch": 5.0268203039634445,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003978583490612894,
+      "loss": 0.4998,
+      "step": 101210
+    },
+    {
+      "epoch": 5.027316976259065,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00039781861527763985,
+      "loss": 0.478,
+      "step": 101220
+    },
+    {
+      "epoch": 5.027813648554684,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00039777888149399027,
+      "loss": 0.4937,
+      "step": 101230
+    },
+    {
+      "epoch": 5.028310320850303,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00039773914771034074,
+      "loss": 0.4847,
+      "step": 101240
+    },
+    {
+      "epoch": 5.028806993145922,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0003976994139266912,
+      "loss": 0.5081,
+      "step": 101250
+    },
+    {
+      "epoch": 5.0293036654415415,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00039765968014304163,
+      "loss": 0.5119,
+      "step": 101260
+    },
+    {
+      "epoch": 5.029800337737161,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0003976199463593921,
+      "loss": 0.5127,
+      "step": 101270
+    },
+    {
+      "epoch": 5.03029701003278,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0003975802125757425,
+      "loss": 0.4919,
+      "step": 101280
+    },
+    {
+      "epoch": 5.0307936823284,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.000397540478792093,
+      "loss": 0.5063,
+      "step": 101290
+    },
+    {
+      "epoch": 5.031290354624019,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00039750074500844346,
+      "loss": 0.4855,
+      "step": 101300
+    },
+    {
+      "epoch": 5.0317870269196385,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0003974610112247939,
+      "loss": 0.4819,
+      "step": 101310
+    },
+    {
+      "epoch": 5.032283699215258,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00039742127744114435,
+      "loss": 0.4999,
+      "step": 101320
+    },
+    {
+      "epoch": 5.032780371510877,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0003973815436574948,
+      "loss": 0.4966,
+      "step": 101330
+    },
+    {
+      "epoch": 5.033277043806496,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00039734180987384524,
+      "loss": 0.529,
+      "step": 101340
+    },
+    {
+      "epoch": 5.033773716102115,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003973020760901957,
+      "loss": 0.4916,
+      "step": 101350
+    },
+    {
+      "epoch": 5.0342703883977356,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0003972623423065462,
+      "loss": 0.4746,
+      "step": 101360
+    },
+    {
+      "epoch": 5.034767060693355,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00039722260852289665,
+      "loss": 0.5153,
+      "step": 101370
+    },
+    {
+      "epoch": 5.035263732988974,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00039718287473924707,
+      "loss": 0.4996,
+      "step": 101380
+    },
+    {
+      "epoch": 5.035760405284593,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003971431409555975,
+      "loss": 0.5083,
+      "step": 101390
+    },
+    {
+      "epoch": 5.0362570775802125,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.000397103407171948,
+      "loss": 0.4963,
+      "step": 101400
+    },
+    {
+      "epoch": 5.036753749875832,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0003970636733882984,
+      "loss": 0.5013,
+      "step": 101410
+    },
+    {
+      "epoch": 5.037250422171451,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00039702393960464884,
+      "loss": 0.4922,
+      "step": 101420
+    },
+    {
+      "epoch": 5.037747094467071,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0003969842058209993,
+      "loss": 0.4955,
+      "step": 101430
+    },
+    {
+      "epoch": 5.03824376676269,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0003969444720373498,
+      "loss": 0.4976,
+      "step": 101440
+    },
+    {
+      "epoch": 5.0387404390583095,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00039690473825370026,
+      "loss": 0.5043,
+      "step": 101450
+    },
+    {
+      "epoch": 5.039237111353929,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0003968650044700507,
+      "loss": 0.4751,
+      "step": 101460
+    },
+    {
+      "epoch": 5.039733783649548,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0003968252706864011,
+      "loss": 0.4963,
+      "step": 101470
+    },
+    {
+      "epoch": 5.040230455945167,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0003967855369027516,
+      "loss": 0.5233,
+      "step": 101480
+    },
+    {
+      "epoch": 5.040727128240786,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00039674580311910203,
+      "loss": 0.5207,
+      "step": 101490
+    },
+    {
+      "epoch": 5.0412238005364065,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00039670606933545245,
+      "loss": 0.4913,
+      "step": 101500
+    },
+    {
+      "epoch": 5.041720472832026,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0003966663355518029,
+      "loss": 0.5108,
+      "step": 101510
+    },
+    {
+      "epoch": 5.042217145127645,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003966266017681534,
+      "loss": 0.5051,
+      "step": 101520
+    },
+    {
+      "epoch": 5.042713817423264,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00039658686798450387,
+      "loss": 0.5184,
+      "step": 101530
+    },
+    {
+      "epoch": 5.043210489718883,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0003965471342008543,
+      "loss": 0.4665,
+      "step": 101540
+    },
+    {
+      "epoch": 5.043707162014503,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00039650740041720475,
+      "loss": 0.522,
+      "step": 101550
+    },
+    {
+      "epoch": 5.044203834310122,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003964676666335552,
+      "loss": 0.4723,
+      "step": 101560
+    },
+    {
+      "epoch": 5.044700506605741,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00039642793284990564,
+      "loss": 0.4845,
+      "step": 101570
+    },
+    {
+      "epoch": 5.045197178901361,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00039638819906625606,
+      "loss": 0.4704,
+      "step": 101580
+    },
+    {
+      "epoch": 5.04569385119698,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003963484652826066,
+      "loss": 0.5048,
+      "step": 101590
+    },
+    {
+      "epoch": 5.0461905234926,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.000396308731498957,
+      "loss": 0.5225,
+      "step": 101600
+    },
+    {
+      "epoch": 5.046687195788219,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003962689977153075,
+      "loss": 0.5385,
+      "step": 101610
+    },
+    {
+      "epoch": 5.047183868083838,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003962292639316579,
+      "loss": 0.4829,
+      "step": 101620
+    },
+    {
+      "epoch": 5.047680540379457,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00039618953014800836,
+      "loss": 0.4772,
+      "step": 101630
+    },
+    {
+      "epoch": 5.048177212675077,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00039614979636435883,
+      "loss": 0.5052,
+      "step": 101640
+    },
+    {
+      "epoch": 5.048673884970697,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00039611006258070925,
+      "loss": 0.5336,
+      "step": 101650
+    },
+    {
+      "epoch": 5.049170557266316,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0003960703287970597,
+      "loss": 0.5237,
+      "step": 101660
+    },
+    {
+      "epoch": 5.049667229561935,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003960305950134102,
+      "loss": 0.5515,
+      "step": 101670
+    },
+    {
+      "epoch": 5.050163901857554,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003959908612297606,
+      "loss": 0.5257,
+      "step": 101680
+    },
+    {
+      "epoch": 5.050660574153174,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0003959511274461111,
+      "loss": 0.5001,
+      "step": 101690
+    },
+    {
+      "epoch": 5.051157246448793,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00039591139366246155,
+      "loss": 0.5143,
+      "step": 101700
+    },
+    {
+      "epoch": 5.051653918744412,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00039587165987881197,
+      "loss": 0.5013,
+      "step": 101710
+    },
+    {
+      "epoch": 5.052150591040032,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00039583192609516244,
+      "loss": 0.5004,
+      "step": 101720
+    },
+    {
+      "epoch": 5.052647263335651,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00039579219231151286,
+      "loss": 0.515,
+      "step": 101730
+    },
+    {
+      "epoch": 5.053143935631271,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00039575245852786333,
+      "loss": 0.4932,
+      "step": 101740
+    },
+    {
+      "epoch": 5.05364060792689,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0003957127247442138,
+      "loss": 0.5008,
+      "step": 101750
+    },
+    {
+      "epoch": 5.054137280222509,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003956729909605642,
+      "loss": 0.5068,
+      "step": 101760
+    },
+    {
+      "epoch": 5.054633952518128,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0003956332571769147,
+      "loss": 0.4884,
+      "step": 101770
+    },
+    {
+      "epoch": 5.0551306248137475,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00039559352339326516,
+      "loss": 0.5051,
+      "step": 101780
+    },
+    {
+      "epoch": 5.055627297109368,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0003955537896096156,
+      "loss": 0.4967,
+      "step": 101790
+    },
+    {
+      "epoch": 5.056123969404987,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00039551405582596605,
+      "loss": 0.5092,
+      "step": 101800
+    },
+    {
+      "epoch": 5.056620641700606,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00039547432204231647,
+      "loss": 0.52,
+      "step": 101810
+    },
+    {
+      "epoch": 5.057117313996225,
+      "grad_norm": 0.125,
+      "learning_rate": 0.000395434588258667,
+      "loss": 0.5318,
+      "step": 101820
+    },
+    {
+      "epoch": 5.0576139862918446,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0003953948544750174,
+      "loss": 0.4582,
+      "step": 101830
+    },
+    {
+      "epoch": 5.058110658587464,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00039535512069136783,
+      "loss": 0.4796,
+      "step": 101840
+    },
+    {
+      "epoch": 5.058607330883083,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003953153869077183,
+      "loss": 0.52,
+      "step": 101850
+    },
+    {
+      "epoch": 5.059104003178703,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00039527565312406877,
+      "loss": 0.4915,
+      "step": 101860
+    },
+    {
+      "epoch": 5.059600675474322,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0003952359193404192,
+      "loss": 0.4914,
+      "step": 101870
+    },
+    {
+      "epoch": 5.060097347769942,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00039519618555676966,
+      "loss": 0.5179,
+      "step": 101880
+    },
+    {
+      "epoch": 5.060594020065561,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00039515645177312013,
+      "loss": 0.5114,
+      "step": 101890
+    },
+    {
+      "epoch": 5.06109069236118,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003951167179894706,
+      "loss": 0.5279,
+      "step": 101900
+    },
+    {
+      "epoch": 5.061587364656799,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000395076984205821,
+      "loss": 0.4745,
+      "step": 101910
+    },
+    {
+      "epoch": 5.0620840369524185,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00039503725042217144,
+      "loss": 0.4826,
+      "step": 101920
+    },
+    {
+      "epoch": 5.062580709248038,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00039499751663852196,
+      "loss": 0.5154,
+      "step": 101930
+    },
+    {
+      "epoch": 5.063077381543658,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003949577828548724,
+      "loss": 0.4944,
+      "step": 101940
+    },
+    {
+      "epoch": 5.063574053839277,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0003949180490712228,
+      "loss": 0.5743,
+      "step": 101950
+    },
+    {
+      "epoch": 5.064070726134896,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00039487831528757327,
+      "loss": 0.5165,
+      "step": 101960
+    },
+    {
+      "epoch": 5.0645673984305155,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00039483858150392374,
+      "loss": 0.4993,
+      "step": 101970
+    },
+    {
+      "epoch": 5.065064070726135,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0003947988477202742,
+      "loss": 0.4998,
+      "step": 101980
+    },
+    {
+      "epoch": 5.065560743021754,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0003947591139366246,
+      "loss": 0.478,
+      "step": 101990
+    },
+    {
+      "epoch": 5.066057415317373,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003947193801529751,
+      "loss": 0.5334,
+      "step": 102000
+    },
+    {
+      "epoch": 5.066554087612993,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00039467964636932557,
+      "loss": 0.5197,
+      "step": 102010
+    },
+    {
+      "epoch": 5.0670507599086125,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.000394639912585676,
+      "loss": 0.5053,
+      "step": 102020
+    },
+    {
+      "epoch": 5.067547432204232,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0003946001788020264,
+      "loss": 0.4691,
+      "step": 102030
+    },
+    {
+      "epoch": 5.068044104499851,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0003945604450183769,
+      "loss": 0.4815,
+      "step": 102040
+    },
+    {
+      "epoch": 5.06854077679547,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.00039452071123472735,
+      "loss": 0.4819,
+      "step": 102050
+    },
+    {
+      "epoch": 5.069037449091089,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0003944809774510778,
+      "loss": 0.4896,
+      "step": 102060
+    },
+    {
+      "epoch": 5.069534121386709,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00039444124366742824,
+      "loss": 0.4975,
+      "step": 102070
+    },
+    {
+      "epoch": 5.070030793682329,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0003944015098837787,
+      "loss": 0.487,
+      "step": 102080
+    },
+    {
+      "epoch": 5.070527465977948,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003943617761001292,
+      "loss": 0.5121,
+      "step": 102090
+    },
+    {
+      "epoch": 5.071024138273567,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003943220423164796,
+      "loss": 0.5056,
+      "step": 102100
+    },
+    {
+      "epoch": 5.0715208105691865,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00039428230853283,
+      "loss": 0.5018,
+      "step": 102110
+    },
+    {
+      "epoch": 5.072017482864806,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00039424257474918054,
+      "loss": 0.5073,
+      "step": 102120
+    },
+    {
+      "epoch": 5.072514155160425,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00039420284096553095,
+      "loss": 0.5065,
+      "step": 102130
+    },
+    {
+      "epoch": 5.073010827456044,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0003941631071818814,
+      "loss": 0.5004,
+      "step": 102140
+    },
+    {
+      "epoch": 5.073507499751664,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00039412337339823184,
+      "loss": 0.5063,
+      "step": 102150
+    },
+    {
+      "epoch": 5.0740041720472835,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0003940836396145823,
+      "loss": 0.4852,
+      "step": 102160
+    },
+    {
+      "epoch": 5.074500844342903,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0003940439058309328,
+      "loss": 0.4731,
+      "step": 102170
+    },
+    {
+      "epoch": 5.074997516638522,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0003940041720472832,
+      "loss": 0.5157,
+      "step": 102180
+    },
+    {
+      "epoch": 5.075494188934141,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003939644382636337,
+      "loss": 0.5061,
+      "step": 102190
+    },
+    {
+      "epoch": 5.07599086122976,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.00039392470447998415,
+      "loss": 0.503,
+      "step": 102200
+    },
+    {
+      "epoch": 5.07648753352538,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00039388497069633456,
+      "loss": 0.4641,
+      "step": 102210
+    },
+    {
+      "epoch": 5.076984205821,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00039384523691268503,
+      "loss": 0.4968,
+      "step": 102220
+    },
+    {
+      "epoch": 5.077480878116619,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0003938055031290355,
+      "loss": 0.4932,
+      "step": 102230
+    },
+    {
+      "epoch": 5.077977550412238,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0003937657693453859,
+      "loss": 0.5269,
+      "step": 102240
+    },
+    {
+      "epoch": 5.078474222707857,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0003937260355617364,
+      "loss": 0.4944,
+      "step": 102250
+    },
+    {
+      "epoch": 5.078970895003477,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0003936863017780868,
+      "loss": 0.5135,
+      "step": 102260
+    },
+    {
+      "epoch": 5.079467567299096,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0003936465679944373,
+      "loss": 0.5092,
+      "step": 102270
+    },
+    {
+      "epoch": 5.079964239594715,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00039360683421078775,
+      "loss": 0.4748,
+      "step": 102280
+    },
+    {
+      "epoch": 5.080460911890334,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00039356710042713817,
+      "loss": 0.4965,
+      "step": 102290
+    },
+    {
+      "epoch": 5.0809575841859544,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00039352736664348864,
+      "loss": 0.523,
+      "step": 102300
+    },
+    {
+      "epoch": 5.081454256481574,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0003934876328598391,
+      "loss": 0.5026,
+      "step": 102310
+    },
+    {
+      "epoch": 5.081950928777193,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00039344789907618953,
+      "loss": 0.5191,
+      "step": 102320
+    },
+    {
+      "epoch": 5.082447601072812,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00039340816529254,
+      "loss": 0.5199,
+      "step": 102330
+    },
+    {
+      "epoch": 5.082944273368431,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003933684315088904,
+      "loss": 0.5079,
+      "step": 102340
+    },
+    {
+      "epoch": 5.083440945664051,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00039332869772524095,
+      "loss": 0.5021,
+      "step": 102350
+    },
+    {
+      "epoch": 5.08393761795967,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00039328896394159136,
+      "loss": 0.5217,
+      "step": 102360
+    },
+    {
+      "epoch": 5.08443429025529,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0003932492301579418,
+      "loss": 0.4928,
+      "step": 102370
+    },
+    {
+      "epoch": 5.084930962550909,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00039320949637429225,
+      "loss": 0.5058,
+      "step": 102380
+    },
+    {
+      "epoch": 5.085427634846528,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0003931697625906427,
+      "loss": 0.4907,
+      "step": 102390
+    },
+    {
+      "epoch": 5.085924307142148,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.00039313002880699314,
+      "loss": 0.5038,
+      "step": 102400
+    },
+    {
+      "epoch": 5.086420979437767,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0003930902950233436,
+      "loss": 0.4868,
+      "step": 102410
+    },
+    {
+      "epoch": 5.086917651733386,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003930505612396941,
+      "loss": 0.4926,
+      "step": 102420
+    },
+    {
+      "epoch": 5.087414324029005,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00039301082745604455,
+      "loss": 0.5046,
+      "step": 102430
+    },
+    {
+      "epoch": 5.087910996324625,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00039297109367239497,
+      "loss": 0.5194,
+      "step": 102440
+    },
+    {
+      "epoch": 5.088407668620245,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0003929313598887454,
+      "loss": 0.5021,
+      "step": 102450
+    },
+    {
+      "epoch": 5.088904340915864,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0003928916261050959,
+      "loss": 0.5071,
+      "step": 102460
+    },
+    {
+      "epoch": 5.089401013211483,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00039285189232144633,
+      "loss": 0.5201,
+      "step": 102470
+    },
+    {
+      "epoch": 5.089897685507102,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00039281215853779675,
+      "loss": 0.4912,
+      "step": 102480
+    },
+    {
+      "epoch": 5.0903943578027215,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0003927724247541472,
+      "loss": 0.5096,
+      "step": 102490
+    },
+    {
+      "epoch": 5.090891030098341,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003927326909704977,
+      "loss": 0.5001,
+      "step": 102500
+    },
+    {
+      "epoch": 5.091387702393961,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00039269295718684816,
+      "loss": 0.4984,
+      "step": 102510
+    },
+    {
+      "epoch": 5.09188437468958,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0003926532234031986,
+      "loss": 0.5196,
+      "step": 102520
+    },
+    {
+      "epoch": 5.092381046985199,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00039261348961954905,
+      "loss": 0.4902,
+      "step": 102530
+    },
+    {
+      "epoch": 5.092877719280819,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003925737558358995,
+      "loss": 0.4998,
+      "step": 102540
+    },
+    {
+      "epoch": 5.093374391576438,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.00039253402205224994,
+      "loss": 0.4996,
+      "step": 102550
+    },
+    {
+      "epoch": 5.093871063872057,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00039249428826860036,
+      "loss": 0.4727,
+      "step": 102560
+    },
+    {
+      "epoch": 5.094367736167676,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0003924545544849508,
+      "loss": 0.4791,
+      "step": 102570
+    },
+    {
+      "epoch": 5.094864408463296,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003924148207013013,
+      "loss": 0.5087,
+      "step": 102580
+    },
+    {
+      "epoch": 5.095361080758916,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00039237508691765177,
+      "loss": 0.4893,
+      "step": 102590
+    },
+    {
+      "epoch": 5.095857753054535,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003923353531340022,
+      "loss": 0.4865,
+      "step": 102600
+    },
+    {
+      "epoch": 5.096354425350154,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00039229561935035266,
+      "loss": 0.5286,
+      "step": 102610
+    },
+    {
+      "epoch": 5.096851097645773,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00039225588556670313,
+      "loss": 0.491,
+      "step": 102620
+    },
+    {
+      "epoch": 5.0973477699413925,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00039221615178305355,
+      "loss": 0.4991,
+      "step": 102630
+    },
+    {
+      "epoch": 5.097844442237012,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.000392176417999404,
+      "loss": 0.4782,
+      "step": 102640
+    },
+    {
+      "epoch": 5.098341114532631,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0003921366842157545,
+      "loss": 0.4842,
+      "step": 102650
+    },
+    {
+      "epoch": 5.098837786828251,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003920969504321049,
+      "loss": 0.5013,
+      "step": 102660
+    },
+    {
+      "epoch": 5.09933445912387,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0003920572166484554,
+      "loss": 0.5183,
+      "step": 102670
+    },
+    {
+      "epoch": 5.0998311314194895,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003920174828648058,
+      "loss": 0.5004,
+      "step": 102680
+    },
+    {
+      "epoch": 5.100327803715109,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00039197774908115627,
+      "loss": 0.5079,
+      "step": 102690
+    },
+    {
+      "epoch": 5.100824476010728,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00039193801529750674,
+      "loss": 0.5287,
+      "step": 102700
+    },
+    {
+      "epoch": 5.101321148306347,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00039189828151385715,
+      "loss": 0.4842,
+      "step": 102710
+    },
+    {
+      "epoch": 5.101817820601966,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003918585477302076,
+      "loss": 0.496,
+      "step": 102720
+    },
+    {
+      "epoch": 5.1023144928975865,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0003918188139465581,
+      "loss": 0.4851,
+      "step": 102730
+    },
+    {
+      "epoch": 5.102811165193206,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003917790801629085,
+      "loss": 0.5089,
+      "step": 102740
+    },
+    {
+      "epoch": 5.103307837488825,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.000391739346379259,
+      "loss": 0.5283,
+      "step": 102750
+    },
+    {
+      "epoch": 5.103804509784444,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00039169961259560946,
+      "loss": 0.5042,
+      "step": 102760
+    },
+    {
+      "epoch": 5.1043011820800634,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0003916598788119599,
+      "loss": 0.5159,
+      "step": 102770
+    },
+    {
+      "epoch": 5.104797854375683,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00039162014502831035,
+      "loss": 0.4816,
+      "step": 102780
+    },
+    {
+      "epoch": 5.105294526671302,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00039158041124466076,
+      "loss": 0.4739,
+      "step": 102790
+    },
+    {
+      "epoch": 5.105791198966922,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0003915406774610113,
+      "loss": 0.4951,
+      "step": 102800
+    },
+    {
+      "epoch": 5.106287871262541,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0003915009436773617,
+      "loss": 0.5007,
+      "step": 102810
+    },
+    {
+      "epoch": 5.1067845435581605,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0003914612098937121,
+      "loss": 0.531,
+      "step": 102820
+    },
+    {
+      "epoch": 5.10728121585378,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003914214761100626,
+      "loss": 0.4928,
+      "step": 102830
+    },
+    {
+      "epoch": 5.107777888149399,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00039138174232641307,
+      "loss": 0.5146,
+      "step": 102840
+    },
+    {
+      "epoch": 5.108274560445018,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003913420085427635,
+      "loss": 0.5104,
+      "step": 102850
+    },
+    {
+      "epoch": 5.108771232740637,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00039130227475911395,
+      "loss": 0.5028,
+      "step": 102860
+    },
+    {
+      "epoch": 5.1092679050362575,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00039126254097546437,
+      "loss": 0.5051,
+      "step": 102870
+    },
+    {
+      "epoch": 5.109764577331877,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003912228071918149,
+      "loss": 0.4709,
+      "step": 102880
+    },
+    {
+      "epoch": 5.110261249627496,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0003911830734081653,
+      "loss": 0.5024,
+      "step": 102890
+    },
+    {
+      "epoch": 5.110757921923115,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00039114333962451573,
+      "loss": 0.5313,
+      "step": 102900
+    },
+    {
+      "epoch": 5.111254594218734,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0003911036058408662,
+      "loss": 0.4929,
+      "step": 102910
+    },
+    {
+      "epoch": 5.111751266514354,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003910638720572167,
+      "loss": 0.515,
+      "step": 102920
+    },
+    {
+      "epoch": 5.112247938809973,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0003910241382735671,
+      "loss": 0.5006,
+      "step": 102930
+    },
+    {
+      "epoch": 5.112744611105592,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00039098440448991756,
+      "loss": 0.5125,
+      "step": 102940
+    },
+    {
+      "epoch": 5.113241283401212,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00039094467070626803,
+      "loss": 0.5217,
+      "step": 102950
+    },
+    {
+      "epoch": 5.113737955696831,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003909049369226185,
+      "loss": 0.4785,
+      "step": 102960
+    },
+    {
+      "epoch": 5.114234627992451,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003908652031389689,
+      "loss": 0.5008,
+      "step": 102970
+    },
+    {
+      "epoch": 5.11473130028807,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00039082546935531934,
+      "loss": 0.4956,
+      "step": 102980
+    },
+    {
+      "epoch": 5.115227972583689,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00039078573557166987,
+      "loss": 0.508,
+      "step": 102990
+    },
+    {
+      "epoch": 5.115724644879308,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0003907460017880203,
+      "loss": 0.4925,
+      "step": 103000
+    },
+    {
+      "epoch": 5.116221317174928,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00039070626800437075,
+      "loss": 0.4951,
+      "step": 103010
+    },
+    {
+      "epoch": 5.116717989470548,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00039066653422072117,
+      "loss": 0.4956,
+      "step": 103020
+    },
+    {
+      "epoch": 5.117214661766167,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00039062680043707164,
+      "loss": 0.478,
+      "step": 103030
+    },
+    {
+      "epoch": 5.117711334061786,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003905870666534221,
+      "loss": 0.5068,
+      "step": 103040
+    },
+    {
+      "epoch": 5.118208006357405,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00039054733286977253,
+      "loss": 0.4952,
+      "step": 103050
+    },
+    {
+      "epoch": 5.118704678653025,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.000390507599086123,
+      "loss": 0.5261,
+      "step": 103060
+    },
+    {
+      "epoch": 5.119201350948644,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0003904678653024735,
+      "loss": 0.5127,
+      "step": 103070
+    },
+    {
+      "epoch": 5.119698023244263,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003904281315188239,
+      "loss": 0.5266,
+      "step": 103080
+    },
+    {
+      "epoch": 5.120194695539883,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00039038839773517436,
+      "loss": 0.5181,
+      "step": 103090
+    },
+    {
+      "epoch": 5.120691367835502,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00039034866395152483,
+      "loss": 0.4815,
+      "step": 103100
+    },
+    {
+      "epoch": 5.121188040131122,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00039030893016787525,
+      "loss": 0.5016,
+      "step": 103110
+    },
+    {
+      "epoch": 5.121684712426741,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003902691963842257,
+      "loss": 0.5309,
+      "step": 103120
+    },
+    {
+      "epoch": 5.12218138472236,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00039022946260057614,
+      "loss": 0.4882,
+      "step": 103130
+    },
+    {
+      "epoch": 5.122678057017979,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0003901897288169266,
+      "loss": 0.4995,
+      "step": 103140
+    },
+    {
+      "epoch": 5.1231747293135985,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0003901499950332771,
+      "loss": 0.486,
+      "step": 103150
+    },
+    {
+      "epoch": 5.123671401609219,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0003901102612496275,
+      "loss": 0.5112,
+      "step": 103160
+    },
+    {
+      "epoch": 5.124168073904838,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00039007052746597797,
+      "loss": 0.5025,
+      "step": 103170
+    },
+    {
+      "epoch": 5.124664746200457,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00039003079368232844,
+      "loss": 0.4685,
+      "step": 103180
+    },
+    {
+      "epoch": 5.125161418496076,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00038999105989867886,
+      "loss": 0.5198,
+      "step": 103190
+    },
+    {
+      "epoch": 5.1256580907916955,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00038995132611502933,
+      "loss": 0.5119,
+      "step": 103200
+    },
+    {
+      "epoch": 5.126154763087315,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00038991159233137975,
+      "loss": 0.5381,
+      "step": 103210
+    },
+    {
+      "epoch": 5.126651435382934,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003898718585477302,
+      "loss": 0.5304,
+      "step": 103220
+    },
+    {
+      "epoch": 5.127148107678554,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003898321247640807,
+      "loss": 0.4936,
+      "step": 103230
+    },
+    {
+      "epoch": 5.127644779974173,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0003897923909804311,
+      "loss": 0.4661,
+      "step": 103240
+    },
+    {
+      "epoch": 5.128141452269793,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0003897526571967816,
+      "loss": 0.5276,
+      "step": 103250
+    },
+    {
+      "epoch": 5.128638124565412,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00038971292341313205,
+      "loss": 0.4815,
+      "step": 103260
+    },
+    {
+      "epoch": 5.129134796861031,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00038967318962948247,
+      "loss": 0.5073,
+      "step": 103270
+    },
+    {
+      "epoch": 5.12963146915665,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00038963345584583294,
+      "loss": 0.4962,
+      "step": 103280
+    },
+    {
+      "epoch": 5.1301281414522695,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003895937220621834,
+      "loss": 0.5054,
+      "step": 103290
+    },
+    {
+      "epoch": 5.13062481374789,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003895539882785338,
+      "loss": 0.4623,
+      "step": 103300
+    },
+    {
+      "epoch": 5.131121486043509,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0003895142544948843,
+      "loss": 0.5243,
+      "step": 103310
+    },
+    {
+      "epoch": 5.131618158339128,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003894745207112347,
+      "loss": 0.5063,
+      "step": 103320
+    },
+    {
+      "epoch": 5.132114830634747,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00038943478692758524,
+      "loss": 0.512,
+      "step": 103330
+    },
+    {
+      "epoch": 5.1326115029303665,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00038939505314393566,
+      "loss": 0.4908,
+      "step": 103340
+    },
+    {
+      "epoch": 5.133108175225986,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003893553193602861,
+      "loss": 0.4813,
+      "step": 103350
+    },
+    {
+      "epoch": 5.133604847521605,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00038931558557663655,
+      "loss": 0.5134,
+      "step": 103360
+    },
+    {
+      "epoch": 5.134101519817224,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.000389275851792987,
+      "loss": 0.4953,
+      "step": 103370
+    },
+    {
+      "epoch": 5.134598192112844,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00038923611800933743,
+      "loss": 0.479,
+      "step": 103380
+    },
+    {
+      "epoch": 5.1350948644084635,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0003891963842256879,
+      "loss": 0.4816,
+      "step": 103390
+    },
+    {
+      "epoch": 5.135591536704083,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0003891566504420384,
+      "loss": 0.4737,
+      "step": 103400
+    },
+    {
+      "epoch": 5.136088208999702,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00038911691665838885,
+      "loss": 0.4755,
+      "step": 103410
+    },
+    {
+      "epoch": 5.136584881295321,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00038907718287473927,
+      "loss": 0.5256,
+      "step": 103420
+    },
+    {
+      "epoch": 5.13708155359094,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0003890374490910897,
+      "loss": 0.5065,
+      "step": 103430
+    },
+    {
+      "epoch": 5.13757822588656,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00038899771530744015,
+      "loss": 0.5308,
+      "step": 103440
+    },
+    {
+      "epoch": 5.13807489818218,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003889579815237906,
+      "loss": 0.5085,
+      "step": 103450
+    },
+    {
+      "epoch": 5.138571570477799,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0003889182477401411,
+      "loss": 0.4964,
+      "step": 103460
+    },
+    {
+      "epoch": 5.139068242773418,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0003888785139564915,
+      "loss": 0.4791,
+      "step": 103470
+    },
+    {
+      "epoch": 5.1395649150690375,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.000388838780172842,
+      "loss": 0.5215,
+      "step": 103480
+    },
+    {
+      "epoch": 5.140061587364657,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00038879904638919246,
+      "loss": 0.4902,
+      "step": 103490
+    },
+    {
+      "epoch": 5.140558259660276,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0003887593126055429,
+      "loss": 0.5147,
+      "step": 103500
+    },
+    {
+      "epoch": 5.141054931955895,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003887195788218933,
+      "loss": 0.468,
+      "step": 103510
+    },
+    {
+      "epoch": 5.141551604251515,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0003886798450382438,
+      "loss": 0.4909,
+      "step": 103520
+    },
+    {
+      "epoch": 5.1420482765471345,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00038864011125459423,
+      "loss": 0.5078,
+      "step": 103530
+    },
+    {
+      "epoch": 5.142544948842754,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0003886003774709447,
+      "loss": 0.511,
+      "step": 103540
+    },
+    {
+      "epoch": 5.143041621138373,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003885606436872951,
+      "loss": 0.5178,
+      "step": 103550
+    },
+    {
+      "epoch": 5.143538293433992,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003885209099036456,
+      "loss": 0.4828,
+      "step": 103560
+    },
+    {
+      "epoch": 5.144034965729611,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00038848117611999607,
+      "loss": 0.4752,
+      "step": 103570
+    },
+    {
+      "epoch": 5.144531638025231,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003884414423363465,
+      "loss": 0.4961,
+      "step": 103580
+    },
+    {
+      "epoch": 5.145028310320851,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00038840170855269695,
+      "loss": 0.5233,
+      "step": 103590
+    },
+    {
+      "epoch": 5.14552498261647,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003883619747690474,
+      "loss": 0.5459,
+      "step": 103600
+    },
+    {
+      "epoch": 5.146021654912089,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00038832224098539784,
+      "loss": 0.5349,
+      "step": 103610
+    },
+    {
+      "epoch": 5.146518327207708,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0003882825072017483,
+      "loss": 0.5141,
+      "step": 103620
+    },
+    {
+      "epoch": 5.147014999503328,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0003882427734180988,
+      "loss": 0.5031,
+      "step": 103630
+    },
+    {
+      "epoch": 5.147511671798947,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0003882030396344492,
+      "loss": 0.5,
+      "step": 103640
+    },
+    {
+      "epoch": 5.148008344094566,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003881633058507997,
+      "loss": 0.4862,
+      "step": 103650
+    },
+    {
+      "epoch": 5.148505016390185,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0003881235720671501,
+      "loss": 0.5254,
+      "step": 103660
+    },
+    {
+      "epoch": 5.149001688685805,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00038808383828350056,
+      "loss": 0.5075,
+      "step": 103670
+    },
+    {
+      "epoch": 5.149498360981425,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00038804410449985103,
+      "loss": 0.5022,
+      "step": 103680
+    },
+    {
+      "epoch": 5.149995033277044,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00038800437071620145,
+      "loss": 0.5217,
+      "step": 103690
+    },
+    {
+      "epoch": 5.150491705572663,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0003879646369325519,
+      "loss": 0.4857,
+      "step": 103700
+    },
+    {
+      "epoch": 5.150988377868282,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0003879249031489024,
+      "loss": 0.5378,
+      "step": 103710
+    },
+    {
+      "epoch": 5.151485050163902,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0003878851693652528,
+      "loss": 0.4913,
+      "step": 103720
+    },
+    {
+      "epoch": 5.151981722459521,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0003878454355816033,
+      "loss": 0.4918,
+      "step": 103730
+    },
+    {
+      "epoch": 5.152478394755141,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0003878057017979537,
+      "loss": 0.5136,
+      "step": 103740
+    },
+    {
+      "epoch": 5.15297506705076,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00038776596801430417,
+      "loss": 0.4936,
+      "step": 103750
+    },
+    {
+      "epoch": 5.153471739346379,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00038772623423065464,
+      "loss": 0.4903,
+      "step": 103760
+    },
+    {
+      "epoch": 5.153968411641999,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00038768650044700506,
+      "loss": 0.5015,
+      "step": 103770
+    },
+    {
+      "epoch": 5.154465083937618,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00038764676666335553,
+      "loss": 0.486,
+      "step": 103780
+    },
+    {
+      "epoch": 5.154961756233237,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.000387607032879706,
+      "loss": 0.4983,
+      "step": 103790
+    },
+    {
+      "epoch": 5.155458428528856,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003875672990960564,
+      "loss": 0.5041,
+      "step": 103800
+    },
+    {
+      "epoch": 5.155955100824476,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003875275653124069,
+      "loss": 0.5199,
+      "step": 103810
+    },
+    {
+      "epoch": 5.156451773120096,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00038748783152875736,
+      "loss": 0.4944,
+      "step": 103820
+    },
+    {
+      "epoch": 5.156948445415715,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0003874480977451078,
+      "loss": 0.5237,
+      "step": 103830
+    },
+    {
+      "epoch": 5.157445117711334,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00038740836396145825,
+      "loss": 0.4921,
+      "step": 103840
+    },
+    {
+      "epoch": 5.157941790006953,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00038736863017780867,
+      "loss": 0.5085,
+      "step": 103850
+    },
+    {
+      "epoch": 5.1584384623025725,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003873288963941592,
+      "loss": 0.5085,
+      "step": 103860
+    },
+    {
+      "epoch": 5.158935134598192,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0003872891626105096,
+      "loss": 0.4656,
+      "step": 103870
+    },
+    {
+      "epoch": 5.159431806893812,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.00038724942882686,
+      "loss": 0.4783,
+      "step": 103880
+    },
+    {
+      "epoch": 5.159928479189431,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0003872096950432105,
+      "loss": 0.5244,
+      "step": 103890
+    },
+    {
+      "epoch": 5.16042515148505,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00038716996125956097,
+      "loss": 0.4822,
+      "step": 103900
+    },
+    {
+      "epoch": 5.1609218237806695,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00038713022747591144,
+      "loss": 0.4976,
+      "step": 103910
+    },
+    {
+      "epoch": 5.161418496076289,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00038709049369226186,
+      "loss": 0.52,
+      "step": 103920
+    },
+    {
+      "epoch": 5.161915168371908,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00038705075990861233,
+      "loss": 0.4899,
+      "step": 103930
+    },
+    {
+      "epoch": 5.162411840667527,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0003870110261249628,
+      "loss": 0.509,
+      "step": 103940
+    },
+    {
+      "epoch": 5.162908512963147,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0003869712923413132,
+      "loss": 0.5289,
+      "step": 103950
+    },
+    {
+      "epoch": 5.163405185258767,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00038693155855766363,
+      "loss": 0.516,
+      "step": 103960
+    },
+    {
+      "epoch": 5.163901857554386,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0003868918247740141,
+      "loss": 0.5035,
+      "step": 103970
+    },
+    {
+      "epoch": 5.164398529850005,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003868520909903646,
+      "loss": 0.4768,
+      "step": 103980
+    },
+    {
+      "epoch": 5.164895202145624,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00038681235720671505,
+      "loss": 0.5205,
+      "step": 103990
+    },
+    {
+      "epoch": 5.1653918744412435,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00038677262342306547,
+      "loss": 0.4778,
+      "step": 104000
+    },
+    {
+      "epoch": 5.165888546736863,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00038673288963941594,
+      "loss": 0.507,
+      "step": 104010
+    },
+    {
+      "epoch": 5.166385219032483,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003866931558557664,
+      "loss": 0.4865,
+      "step": 104020
+    },
+    {
+      "epoch": 5.166881891328102,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003866534220721168,
+      "loss": 0.513,
+      "step": 104030
+    },
+    {
+      "epoch": 5.167378563623721,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00038661368828846724,
+      "loss": 0.498,
+      "step": 104040
+    },
+    {
+      "epoch": 5.1678752359193405,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00038657395450481777,
+      "loss": 0.4811,
+      "step": 104050
+    },
+    {
+      "epoch": 5.16837190821496,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0003865342207211682,
+      "loss": 0.516,
+      "step": 104060
+    },
+    {
+      "epoch": 5.168868580510579,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00038649448693751866,
+      "loss": 0.4896,
+      "step": 104070
+    },
+    {
+      "epoch": 5.169365252806198,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0003864547531538691,
+      "loss": 0.4965,
+      "step": 104080
+    },
+    {
+      "epoch": 5.169861925101817,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00038641501937021955,
+      "loss": 0.5355,
+      "step": 104090
+    },
+    {
+      "epoch": 5.1703585973974375,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00038637528558657,
+      "loss": 0.4911,
+      "step": 104100
+    },
+    {
+      "epoch": 5.170855269693057,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00038633555180292043,
+      "loss": 0.4893,
+      "step": 104110
+    },
+    {
+      "epoch": 5.171351941988676,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0003862958180192709,
+      "loss": 0.5215,
+      "step": 104120
+    },
+    {
+      "epoch": 5.171848614284295,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003862560842356214,
+      "loss": 0.4862,
+      "step": 104130
+    },
+    {
+      "epoch": 5.172345286579914,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0003862163504519718,
+      "loss": 0.5223,
+      "step": 104140
+    },
+    {
+      "epoch": 5.172841958875534,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00038617661666832227,
+      "loss": 0.5009,
+      "step": 104150
+    },
+    {
+      "epoch": 5.173338631171153,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00038613688288467274,
+      "loss": 0.5038,
+      "step": 104160
+    },
+    {
+      "epoch": 5.173835303466773,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00038609714910102315,
+      "loss": 0.503,
+      "step": 104170
+    },
+    {
+      "epoch": 5.174331975762392,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0003860574153173736,
+      "loss": 0.5017,
+      "step": 104180
+    },
+    {
+      "epoch": 5.1748286480580115,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00038601768153372404,
+      "loss": 0.5268,
+      "step": 104190
+    },
+    {
+      "epoch": 5.175325320353631,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0003859779477500745,
+      "loss": 0.5443,
+      "step": 104200
+    },
+    {
+      "epoch": 5.17582199264925,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.000385938213966425,
+      "loss": 0.4868,
+      "step": 104210
+    },
+    {
+      "epoch": 5.176318664944869,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003858984801827754,
+      "loss": 0.5093,
+      "step": 104220
+    },
+    {
+      "epoch": 5.176815337240488,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003858587463991259,
+      "loss": 0.4801,
+      "step": 104230
+    },
+    {
+      "epoch": 5.1773120095361085,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00038581901261547635,
+      "loss": 0.5064,
+      "step": 104240
+    },
+    {
+      "epoch": 5.177808681831728,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.00038577927883182676,
+      "loss": 0.5123,
+      "step": 104250
+    },
+    {
+      "epoch": 5.178305354127347,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00038573954504817723,
+      "loss": 0.4905,
+      "step": 104260
+    },
+    {
+      "epoch": 5.178802026422966,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00038569981126452765,
+      "loss": 0.4858,
+      "step": 104270
+    },
+    {
+      "epoch": 5.179298698718585,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0003856600774808782,
+      "loss": 0.4955,
+      "step": 104280
+    },
+    {
+      "epoch": 5.179795371014205,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0003856203436972286,
+      "loss": 0.4959,
+      "step": 104290
+    },
+    {
+      "epoch": 5.180292043309824,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.000385580609913579,
+      "loss": 0.4797,
+      "step": 104300
+    },
+    {
+      "epoch": 5.180788715605443,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003855408761299295,
+      "loss": 0.5123,
+      "step": 104310
+    },
+    {
+      "epoch": 5.181285387901063,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00038550114234627995,
+      "loss": 0.5286,
+      "step": 104320
+    },
+    {
+      "epoch": 5.181782060196682,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00038546140856263037,
+      "loss": 0.5173,
+      "step": 104330
+    },
+    {
+      "epoch": 5.182278732492302,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00038542167477898084,
+      "loss": 0.5002,
+      "step": 104340
+    },
+    {
+      "epoch": 5.182775404787921,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0003853819409953313,
+      "loss": 0.5172,
+      "step": 104350
+    },
+    {
+      "epoch": 5.18327207708354,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003853422072116818,
+      "loss": 0.5017,
+      "step": 104360
+    },
+    {
+      "epoch": 5.183768749379159,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003853024734280322,
+      "loss": 0.5134,
+      "step": 104370
+    },
+    {
+      "epoch": 5.1842654216747786,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0003852627396443826,
+      "loss": 0.5253,
+      "step": 104380
+    },
+    {
+      "epoch": 5.184762093970399,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00038522300586073314,
+      "loss": 0.4886,
+      "step": 104390
+    },
+    {
+      "epoch": 5.185258766266018,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00038518327207708356,
+      "loss": 0.549,
+      "step": 104400
+    },
+    {
+      "epoch": 5.185755438561637,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.000385143538293434,
+      "loss": 0.5098,
+      "step": 104410
+    },
+    {
+      "epoch": 5.186252110857256,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00038510380450978445,
+      "loss": 0.5156,
+      "step": 104420
+    },
+    {
+      "epoch": 5.186748783152876,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003850640707261349,
+      "loss": 0.5012,
+      "step": 104430
+    },
+    {
+      "epoch": 5.187245455448495,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0003850243369424854,
+      "loss": 0.4997,
+      "step": 104440
+    },
+    {
+      "epoch": 5.187742127744114,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0003849846031588358,
+      "loss": 0.5118,
+      "step": 104450
+    },
+    {
+      "epoch": 5.188238800039734,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003849448693751863,
+      "loss": 0.4779,
+      "step": 104460
+    },
+    {
+      "epoch": 5.188735472335353,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00038490513559153675,
+      "loss": 0.5122,
+      "step": 104470
+    },
+    {
+      "epoch": 5.189232144630973,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00038486540180788717,
+      "loss": 0.5162,
+      "step": 104480
+    },
+    {
+      "epoch": 5.189728816926592,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003848256680242376,
+      "loss": 0.5082,
+      "step": 104490
+    },
+    {
+      "epoch": 5.190225489222211,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003847859342405881,
+      "loss": 0.5018,
+      "step": 104500
+    },
+    {
+      "epoch": 5.19072216151783,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00038474620045693853,
+      "loss": 0.5197,
+      "step": 104510
+    },
+    {
+      "epoch": 5.1912188338134495,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.000384706466673289,
+      "loss": 0.4853,
+      "step": 104520
+    },
+    {
+      "epoch": 5.19171550610907,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0003846667328896394,
+      "loss": 0.5236,
+      "step": 104530
+    },
+    {
+      "epoch": 5.192212178404689,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003846269991059899,
+      "loss": 0.5041,
+      "step": 104540
+    },
+    {
+      "epoch": 5.192708850700308,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00038458726532234036,
+      "loss": 0.5011,
+      "step": 104550
+    },
+    {
+      "epoch": 5.193205522995927,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0003845475315386908,
+      "loss": 0.4773,
+      "step": 104560
+    },
+    {
+      "epoch": 5.1937021952915465,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003845077977550412,
+      "loss": 0.5185,
+      "step": 104570
+    },
+    {
+      "epoch": 5.194198867587166,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003844680639713917,
+      "loss": 0.5093,
+      "step": 104580
+    },
+    {
+      "epoch": 5.194695539882785,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00038442833018774214,
+      "loss": 0.4653,
+      "step": 104590
+    },
+    {
+      "epoch": 5.195192212178405,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003843885964040926,
+      "loss": 0.5008,
+      "step": 104600
+    },
+    {
+      "epoch": 5.195688884474024,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.000384348862620443,
+      "loss": 0.4891,
+      "step": 104610
+    },
+    {
+      "epoch": 5.1961855567696436,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0003843091288367935,
+      "loss": 0.5258,
+      "step": 104620
+    },
+    {
+      "epoch": 5.196682229065263,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00038426939505314397,
+      "loss": 0.5075,
+      "step": 104630
+    },
+    {
+      "epoch": 5.197178901360882,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0003842296612694944,
+      "loss": 0.5251,
+      "step": 104640
+    },
+    {
+      "epoch": 5.197675573656501,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.00038418992748584486,
+      "loss": 0.5075,
+      "step": 104650
+    },
+    {
+      "epoch": 5.1981722459521205,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00038415019370219533,
+      "loss": 0.4976,
+      "step": 104660
+    },
+    {
+      "epoch": 5.198668918247741,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00038411045991854575,
+      "loss": 0.4958,
+      "step": 104670
+    },
+    {
+      "epoch": 5.19916559054336,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0003840707261348962,
+      "loss": 0.5061,
+      "step": 104680
+    },
+    {
+      "epoch": 5.199662262838979,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003840309923512467,
+      "loss": 0.5057,
+      "step": 104690
+    },
+    {
+      "epoch": 5.200158935134598,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003839912585675971,
+      "loss": 0.5427,
+      "step": 104700
+    },
+    {
+      "epoch": 5.2006556074302175,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0003839515247839476,
+      "loss": 0.4838,
+      "step": 104710
+    },
+    {
+      "epoch": 5.201152279725837,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.000383911791000298,
+      "loss": 0.5102,
+      "step": 104720
+    },
+    {
+      "epoch": 5.201648952021456,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003838720572166485,
+      "loss": 0.4762,
+      "step": 104730
+    },
+    {
+      "epoch": 5.202145624317075,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00038383232343299894,
+      "loss": 0.5001,
+      "step": 104740
+    },
+    {
+      "epoch": 5.202642296612695,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00038379258964934935,
+      "loss": 0.5035,
+      "step": 104750
+    },
+    {
+      "epoch": 5.2031389689083145,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0003837528558656998,
+      "loss": 0.5256,
+      "step": 104760
+    },
+    {
+      "epoch": 5.203635641203934,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0003837131220820503,
+      "loss": 0.4938,
+      "step": 104770
+    },
+    {
+      "epoch": 5.204132313499553,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003836733882984007,
+      "loss": 0.5047,
+      "step": 104780
+    },
+    {
+      "epoch": 5.204628985795172,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003836336545147512,
+      "loss": 0.4969,
+      "step": 104790
+    },
+    {
+      "epoch": 5.205125658090791,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00038359392073110166,
+      "loss": 0.5321,
+      "step": 104800
+    },
+    {
+      "epoch": 5.205622330386411,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00038355418694745213,
+      "loss": 0.49,
+      "step": 104810
+    },
+    {
+      "epoch": 5.206119002682031,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00038351445316380255,
+      "loss": 0.4994,
+      "step": 104820
+    },
+    {
+      "epoch": 5.20661567497765,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00038347471938015296,
+      "loss": 0.494,
+      "step": 104830
+    },
+    {
+      "epoch": 5.207112347273269,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00038343498559650343,
+      "loss": 0.5085,
+      "step": 104840
+    },
+    {
+      "epoch": 5.207609019568888,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0003833952518128539,
+      "loss": 0.4951,
+      "step": 104850
+    },
+    {
+      "epoch": 5.208105691864508,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0003833555180292043,
+      "loss": 0.4921,
+      "step": 104860
+    },
+    {
+      "epoch": 5.208602364160127,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0003833157842455548,
+      "loss": 0.4962,
+      "step": 104870
+    },
+    {
+      "epoch": 5.209099036455746,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00038327605046190526,
+      "loss": 0.4833,
+      "step": 104880
+    },
+    {
+      "epoch": 5.209595708751366,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00038323631667825574,
+      "loss": 0.5146,
+      "step": 104890
+    },
+    {
+      "epoch": 5.2100923810469855,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00038319658289460615,
+      "loss": 0.5261,
+      "step": 104900
+    },
+    {
+      "epoch": 5.210589053342605,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00038315684911095657,
+      "loss": 0.5051,
+      "step": 104910
+    },
+    {
+      "epoch": 5.211085725638224,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0003831171153273071,
+      "loss": 0.5243,
+      "step": 104920
+    },
+    {
+      "epoch": 5.211582397933843,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0003830773815436575,
+      "loss": 0.5141,
+      "step": 104930
+    },
+    {
+      "epoch": 5.212079070229462,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00038303764776000793,
+      "loss": 0.5209,
+      "step": 104940
+    },
+    {
+      "epoch": 5.212575742525082,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0003829979139763584,
+      "loss": 0.5151,
+      "step": 104950
+    },
+    {
+      "epoch": 5.213072414820702,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0003829581801927089,
+      "loss": 0.4639,
+      "step": 104960
+    },
+    {
+      "epoch": 5.213569087116321,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00038291844640905934,
+      "loss": 0.5022,
+      "step": 104970
+    },
+    {
+      "epoch": 5.21406575941194,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00038287871262540976,
+      "loss": 0.5297,
+      "step": 104980
+    },
+    {
+      "epoch": 5.214562431707559,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00038283897884176023,
+      "loss": 0.5268,
+      "step": 104990
+    },
+    {
+      "epoch": 5.215059104003179,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0003827992450581107,
+      "loss": 0.4914,
+      "step": 105000
+    },
+    {
+      "epoch": 5.215555776298798,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0003827595112744611,
+      "loss": 0.4876,
+      "step": 105010
+    },
+    {
+      "epoch": 5.216052448594417,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00038271977749081154,
+      "loss": 0.4791,
+      "step": 105020
+    },
+    {
+      "epoch": 5.216549120890036,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00038268004370716206,
+      "loss": 0.5153,
+      "step": 105030
+    },
+    {
+      "epoch": 5.217045793185656,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0003826403099235125,
+      "loss": 0.4919,
+      "step": 105040
+    },
+    {
+      "epoch": 5.217542465481276,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00038260057613986295,
+      "loss": 0.529,
+      "step": 105050
+    },
+    {
+      "epoch": 5.218039137776895,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00038256084235621337,
+      "loss": 0.5042,
+      "step": 105060
+    },
+    {
+      "epoch": 5.218535810072514,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00038252110857256384,
+      "loss": 0.4991,
+      "step": 105070
+    },
+    {
+      "epoch": 5.219032482368133,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003824813747889143,
+      "loss": 0.5019,
+      "step": 105080
+    },
+    {
+      "epoch": 5.219529154663753,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00038244164100526473,
+      "loss": 0.4988,
+      "step": 105090
+    },
+    {
+      "epoch": 5.220025826959372,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0003824019072216152,
+      "loss": 0.5088,
+      "step": 105100
+    },
+    {
+      "epoch": 5.220522499254992,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00038236217343796567,
+      "loss": 0.5021,
+      "step": 105110
+    },
+    {
+      "epoch": 5.221019171550611,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0003823224396543161,
+      "loss": 0.4859,
+      "step": 105120
+    },
+    {
+      "epoch": 5.22151584384623,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00038228270587066656,
+      "loss": 0.5081,
+      "step": 105130
+    },
+    {
+      "epoch": 5.22201251614185,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.000382242972087017,
+      "loss": 0.5057,
+      "step": 105140
+    },
+    {
+      "epoch": 5.222509188437469,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00038220323830336745,
+      "loss": 0.5026,
+      "step": 105150
+    },
+    {
+      "epoch": 5.223005860733088,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003821635045197179,
+      "loss": 0.5161,
+      "step": 105160
+    },
+    {
+      "epoch": 5.223502533028707,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00038212377073606834,
+      "loss": 0.5062,
+      "step": 105170
+    },
+    {
+      "epoch": 5.223999205324327,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003820840369524188,
+      "loss": 0.5017,
+      "step": 105180
+    },
+    {
+      "epoch": 5.224495877619947,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003820443031687693,
+      "loss": 0.5259,
+      "step": 105190
+    },
+    {
+      "epoch": 5.224992549915566,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003820045693851197,
+      "loss": 0.5123,
+      "step": 105200
+    },
+    {
+      "epoch": 5.225489222211185,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00038196483560147017,
+      "loss": 0.533,
+      "step": 105210
+    },
+    {
+      "epoch": 5.225985894506804,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00038192510181782064,
+      "loss": 0.4918,
+      "step": 105220
+    },
+    {
+      "epoch": 5.2264825668024235,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00038188536803417106,
+      "loss": 0.4995,
+      "step": 105230
+    },
+    {
+      "epoch": 5.226979239098043,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00038184563425052153,
+      "loss": 0.5495,
+      "step": 105240
+    },
+    {
+      "epoch": 5.227475911393663,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00038180590046687195,
+      "loss": 0.4786,
+      "step": 105250
+    },
+    {
+      "epoch": 5.227972583689282,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00038176616668322247,
+      "loss": 0.494,
+      "step": 105260
+    },
+    {
+      "epoch": 5.228469255984901,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0003817264328995729,
+      "loss": 0.4965,
+      "step": 105270
+    },
+    {
+      "epoch": 5.2289659282805205,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003816866991159233,
+      "loss": 0.4844,
+      "step": 105280
+    },
+    {
+      "epoch": 5.22946260057614,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0003816469653322738,
+      "loss": 0.4911,
+      "step": 105290
+    },
+    {
+      "epoch": 5.229959272871759,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00038160723154862425,
+      "loss": 0.5403,
+      "step": 105300
+    },
+    {
+      "epoch": 5.230455945167378,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00038156749776497467,
+      "loss": 0.503,
+      "step": 105310
+    },
+    {
+      "epoch": 5.230952617462998,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00038152776398132514,
+      "loss": 0.5058,
+      "step": 105320
+    },
+    {
+      "epoch": 5.231449289758618,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0003814880301976756,
+      "loss": 0.5087,
+      "step": 105330
+    },
+    {
+      "epoch": 5.231945962054237,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0003814482964140261,
+      "loss": 0.5184,
+      "step": 105340
+    },
+    {
+      "epoch": 5.232442634349856,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003814085626303765,
+      "loss": 0.4996,
+      "step": 105350
+    },
+    {
+      "epoch": 5.232939306645475,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003813688288467269,
+      "loss": 0.4916,
+      "step": 105360
+    },
+    {
+      "epoch": 5.2334359789410945,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00038132909506307744,
+      "loss": 0.5033,
+      "step": 105370
+    },
+    {
+      "epoch": 5.233932651236714,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00038128936127942786,
+      "loss": 0.5365,
+      "step": 105380
+    },
+    {
+      "epoch": 5.234429323532334,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0003812496274957783,
+      "loss": 0.4962,
+      "step": 105390
+    },
+    {
+      "epoch": 5.234925995827953,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00038120989371212875,
+      "loss": 0.4889,
+      "step": 105400
+    },
+    {
+      "epoch": 5.235422668123572,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003811701599284792,
+      "loss": 0.5129,
+      "step": 105410
+    },
+    {
+      "epoch": 5.2359193404191915,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003811304261448297,
+      "loss": 0.4706,
+      "step": 105420
+    },
+    {
+      "epoch": 5.236416012714811,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003810906923611801,
+      "loss": 0.4972,
+      "step": 105430
+    },
+    {
+      "epoch": 5.23691268501043,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003810509585775305,
+      "loss": 0.5081,
+      "step": 105440
+    },
+    {
+      "epoch": 5.237409357306049,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00038101122479388105,
+      "loss": 0.4888,
+      "step": 105450
+    },
+    {
+      "epoch": 5.237906029601668,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00038097149101023147,
+      "loss": 0.4939,
+      "step": 105460
+    },
+    {
+      "epoch": 5.2384027018972885,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003809317572265819,
+      "loss": 0.5051,
+      "step": 105470
+    },
+    {
+      "epoch": 5.238899374192908,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00038089202344293235,
+      "loss": 0.5286,
+      "step": 105480
+    },
+    {
+      "epoch": 5.239396046488527,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003808522896592828,
+      "loss": 0.5226,
+      "step": 105490
+    },
+    {
+      "epoch": 5.239892718784146,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0003808125558756333,
+      "loss": 0.5016,
+      "step": 105500
+    },
+    {
+      "epoch": 5.240389391079765,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003807728220919837,
+      "loss": 0.4833,
+      "step": 105510
+    },
+    {
+      "epoch": 5.240886063375385,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003807330883083342,
+      "loss": 0.5191,
+      "step": 105520
+    },
+    {
+      "epoch": 5.241382735671004,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00038069335452468466,
+      "loss": 0.4965,
+      "step": 105530
+    },
+    {
+      "epoch": 5.241879407966624,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003806536207410351,
+      "loss": 0.4785,
+      "step": 105540
+    },
+    {
+      "epoch": 5.242376080262243,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00038061388695738554,
+      "loss": 0.5122,
+      "step": 105550
+    },
+    {
+      "epoch": 5.2428727525578624,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.000380574153173736,
+      "loss": 0.4763,
+      "step": 105560
+    },
+    {
+      "epoch": 5.243369424853482,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00038053441939008643,
+      "loss": 0.5306,
+      "step": 105570
+    },
+    {
+      "epoch": 5.243866097149101,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0003804946856064369,
+      "loss": 0.5305,
+      "step": 105580
+    },
+    {
+      "epoch": 5.24436276944472,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0003804549518227873,
+      "loss": 0.4957,
+      "step": 105590
+    },
+    {
+      "epoch": 5.244859441740339,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003804152180391378,
+      "loss": 0.5046,
+      "step": 105600
+    },
+    {
+      "epoch": 5.2453561140359595,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00038037548425548826,
+      "loss": 0.4918,
+      "step": 105610
+    },
+    {
+      "epoch": 5.245852786331579,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003803357504718387,
+      "loss": 0.486,
+      "step": 105620
+    },
+    {
+      "epoch": 5.246349458627198,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00038029601668818915,
+      "loss": 0.4918,
+      "step": 105630
+    },
+    {
+      "epoch": 5.246846130922817,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0003802562829045396,
+      "loss": 0.4641,
+      "step": 105640
+    },
+    {
+      "epoch": 5.247342803218436,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00038021654912089004,
+      "loss": 0.5065,
+      "step": 105650
+    },
+    {
+      "epoch": 5.247839475514056,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003801768153372405,
+      "loss": 0.5032,
+      "step": 105660
+    },
+    {
+      "epoch": 5.248336147809675,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00038013708155359093,
+      "loss": 0.4881,
+      "step": 105670
+    },
+    {
+      "epoch": 5.248832820105295,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0003800973477699414,
+      "loss": 0.4983,
+      "step": 105680
+    },
+    {
+      "epoch": 5.249329492400914,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00038005761398629187,
+      "loss": 0.4834,
+      "step": 105690
+    },
+    {
+      "epoch": 5.249826164696533,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003800178802026423,
+      "loss": 0.4972,
+      "step": 105700
+    },
+    {
+      "epoch": 5.250322836992153,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00037997814641899276,
+      "loss": 0.5257,
+      "step": 105710
+    },
+    {
+      "epoch": 5.250819509287772,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00037993841263534323,
+      "loss": 0.5216,
+      "step": 105720
+    },
+    {
+      "epoch": 5.251316181583391,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00037989867885169365,
+      "loss": 0.4775,
+      "step": 105730
+    },
+    {
+      "epoch": 5.25181285387901,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0003798589450680441,
+      "loss": 0.4706,
+      "step": 105740
+    },
+    {
+      "epoch": 5.2523095261746295,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0003798192112843946,
+      "loss": 0.5176,
+      "step": 105750
+    },
+    {
+      "epoch": 5.25280619847025,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.000379779477500745,
+      "loss": 0.5154,
+      "step": 105760
+    },
+    {
+      "epoch": 5.253302870765869,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0003797397437170955,
+      "loss": 0.4837,
+      "step": 105770
+    },
+    {
+      "epoch": 5.253799543061488,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003797000099334459,
+      "loss": 0.5053,
+      "step": 105780
+    },
+    {
+      "epoch": 5.254296215357107,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0003796602761497964,
+      "loss": 0.5251,
+      "step": 105790
+    },
+    {
+      "epoch": 5.254792887652727,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00037962054236614684,
+      "loss": 0.4851,
+      "step": 105800
+    },
+    {
+      "epoch": 5.255289559948346,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00037958080858249726,
+      "loss": 0.502,
+      "step": 105810
+    },
+    {
+      "epoch": 5.255786232243965,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00037954107479884773,
+      "loss": 0.5086,
+      "step": 105820
+    },
+    {
+      "epoch": 5.256282904539585,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0003795013410151982,
+      "loss": 0.5406,
+      "step": 105830
+    },
+    {
+      "epoch": 5.256779576835204,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003794616072315486,
+      "loss": 0.5044,
+      "step": 105840
+    },
+    {
+      "epoch": 5.257276249130824,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003794218734478991,
+      "loss": 0.4961,
+      "step": 105850
+    },
+    {
+      "epoch": 5.257772921426443,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00037938213966424956,
+      "loss": 0.5446,
+      "step": 105860
+    },
+    {
+      "epoch": 5.258269593722062,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00037934240588060003,
+      "loss": 0.4895,
+      "step": 105870
+    },
+    {
+      "epoch": 5.258766266017681,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00037930267209695045,
+      "loss": 0.4977,
+      "step": 105880
+    },
+    {
+      "epoch": 5.2592629383133005,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00037926293831330087,
+      "loss": 0.5135,
+      "step": 105890
+    },
+    {
+      "epoch": 5.259759610608921,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003792232045296514,
+      "loss": 0.4763,
+      "step": 105900
+    },
+    {
+      "epoch": 5.26025628290454,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003791834707460018,
+      "loss": 0.4835,
+      "step": 105910
+    },
+    {
+      "epoch": 5.260752955200159,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003791437369623522,
+      "loss": 0.4958,
+      "step": 105920
+    },
+    {
+      "epoch": 5.261249627495778,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003791040031787027,
+      "loss": 0.4818,
+      "step": 105930
+    },
+    {
+      "epoch": 5.2617462997913975,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00037906426939505317,
+      "loss": 0.5103,
+      "step": 105940
+    },
+    {
+      "epoch": 5.262242972087017,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00037902453561140364,
+      "loss": 0.4874,
+      "step": 105950
+    },
+    {
+      "epoch": 5.262739644382636,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00037898480182775406,
+      "loss": 0.5046,
+      "step": 105960
+    },
+    {
+      "epoch": 5.263236316678256,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0003789450680441045,
+      "loss": 0.4964,
+      "step": 105970
+    },
+    {
+      "epoch": 5.263732988973875,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.000378905334260455,
+      "loss": 0.4827,
+      "step": 105980
+    },
+    {
+      "epoch": 5.2642296612694945,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003788656004768054,
+      "loss": 0.5055,
+      "step": 105990
+    },
+    {
+      "epoch": 5.264726333565114,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003788258666931559,
+      "loss": 0.5091,
+      "step": 106000
+    },
+    {
+      "epoch": 5.265223005860733,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003787861329095063,
+      "loss": 0.5217,
+      "step": 106010
+    },
+    {
+      "epoch": 5.265719678156352,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0003787463991258568,
+      "loss": 0.4808,
+      "step": 106020
+    },
+    {
+      "epoch": 5.2662163504519715,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00037870666534220725,
+      "loss": 0.5099,
+      "step": 106030
+    },
+    {
+      "epoch": 5.266713022747592,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00037866693155855767,
+      "loss": 0.5062,
+      "step": 106040
+    },
+    {
+      "epoch": 5.267209695043211,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00037862719777490814,
+      "loss": 0.5205,
+      "step": 106050
+    },
+    {
+      "epoch": 5.26770636733883,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0003785874639912586,
+      "loss": 0.5015,
+      "step": 106060
+    },
+    {
+      "epoch": 5.268203039634449,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.000378547730207609,
+      "loss": 0.5136,
+      "step": 106070
+    },
+    {
+      "epoch": 5.2686997119300685,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0003785079964239595,
+      "loss": 0.4776,
+      "step": 106080
+    },
+    {
+      "epoch": 5.269196384225688,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00037846826264030997,
+      "loss": 0.5012,
+      "step": 106090
+    },
+    {
+      "epoch": 5.269693056521307,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0003784285288566604,
+      "loss": 0.5194,
+      "step": 106100
+    },
+    {
+      "epoch": 5.270189728816927,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00037838879507301086,
+      "loss": 0.515,
+      "step": 106110
+    },
+    {
+      "epoch": 5.270686401112546,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0003783490612893613,
+      "loss": 0.4908,
+      "step": 106120
+    },
+    {
+      "epoch": 5.2711830734081655,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00037830932750571174,
+      "loss": 0.4953,
+      "step": 106130
+    },
+    {
+      "epoch": 5.271679745703785,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0003782695937220622,
+      "loss": 0.493,
+      "step": 106140
+    },
+    {
+      "epoch": 5.272176417999404,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00037822985993841263,
+      "loss": 0.4988,
+      "step": 106150
+    },
+    {
+      "epoch": 5.272673090295023,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003781901261547631,
+      "loss": 0.4874,
+      "step": 106160
+    },
+    {
+      "epoch": 5.273169762590642,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003781503923711136,
+      "loss": 0.539,
+      "step": 106170
+    },
+    {
+      "epoch": 5.273666434886262,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.000378110658587464,
+      "loss": 0.512,
+      "step": 106180
+    },
+    {
+      "epoch": 5.274163107181882,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00037807092480381446,
+      "loss": 0.4819,
+      "step": 106190
+    },
+    {
+      "epoch": 5.274659779477501,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.00037803119102016494,
+      "loss": 0.5073,
+      "step": 106200
+    },
+    {
+      "epoch": 5.27515645177312,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00037799145723651535,
+      "loss": 0.4994,
+      "step": 106210
+    },
+    {
+      "epoch": 5.275653124068739,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0003779517234528658,
+      "loss": 0.5123,
+      "step": 106220
+    },
+    {
+      "epoch": 5.276149796364359,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00037791198966921624,
+      "loss": 0.4797,
+      "step": 106230
+    },
+    {
+      "epoch": 5.276646468659978,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0003778722558855667,
+      "loss": 0.5056,
+      "step": 106240
+    },
+    {
+      "epoch": 5.277143140955597,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0003778325221019172,
+      "loss": 0.5202,
+      "step": 106250
+    },
+    {
+      "epoch": 5.277639813251217,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003777927883182676,
+      "loss": 0.5255,
+      "step": 106260
+    },
+    {
+      "epoch": 5.2781364855468365,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0003777530545346181,
+      "loss": 0.4985,
+      "step": 106270
+    },
+    {
+      "epoch": 5.278633157842456,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00037771332075096854,
+      "loss": 0.5092,
+      "step": 106280
+    },
+    {
+      "epoch": 5.279129830138075,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00037767358696731896,
+      "loss": 0.5219,
+      "step": 106290
+    },
+    {
+      "epoch": 5.279626502433694,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00037763385318366943,
+      "loss": 0.5119,
+      "step": 106300
+    },
+    {
+      "epoch": 5.280123174729313,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00037759411940001985,
+      "loss": 0.4844,
+      "step": 106310
+    },
+    {
+      "epoch": 5.280619847024933,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0003775543856163704,
+      "loss": 0.493,
+      "step": 106320
+    },
+    {
+      "epoch": 5.281116519320553,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0003775146518327208,
+      "loss": 0.5006,
+      "step": 106330
+    },
+    {
+      "epoch": 5.281613191616172,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003774749180490712,
+      "loss": 0.494,
+      "step": 106340
+    },
+    {
+      "epoch": 5.282109863911791,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003774351842654217,
+      "loss": 0.4872,
+      "step": 106350
+    },
+    {
+      "epoch": 5.28260653620741,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00037739545048177215,
+      "loss": 0.4894,
+      "step": 106360
+    },
+    {
+      "epoch": 5.28310320850303,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0003773557166981226,
+      "loss": 0.5106,
+      "step": 106370
+    },
+    {
+      "epoch": 5.283599880798649,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00037731598291447304,
+      "loss": 0.4896,
+      "step": 106380
+    },
+    {
+      "epoch": 5.284096553094268,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003772762491308235,
+      "loss": 0.513,
+      "step": 106390
+    },
+    {
+      "epoch": 5.284593225389887,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.000377236515347174,
+      "loss": 0.5174,
+      "step": 106400
+    },
+    {
+      "epoch": 5.285089897685507,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003771967815635244,
+      "loss": 0.5198,
+      "step": 106410
+    },
+    {
+      "epoch": 5.285586569981127,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0003771570477798748,
+      "loss": 0.513,
+      "step": 106420
+    },
+    {
+      "epoch": 5.286083242276746,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00037711731399622534,
+      "loss": 0.4695,
+      "step": 106430
+    },
+    {
+      "epoch": 5.286579914572365,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00037707758021257576,
+      "loss": 0.5157,
+      "step": 106440
+    },
+    {
+      "epoch": 5.287076586867984,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00037703784642892623,
+      "loss": 0.4926,
+      "step": 106450
+    },
+    {
+      "epoch": 5.2875732591636035,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00037699811264527665,
+      "loss": 0.5179,
+      "step": 106460
+    },
+    {
+      "epoch": 5.288069931459223,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0003769583788616271,
+      "loss": 0.5059,
+      "step": 106470
+    },
+    {
+      "epoch": 5.288566603754843,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0003769186450779776,
+      "loss": 0.5065,
+      "step": 106480
+    },
+    {
+      "epoch": 5.289063276050462,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.000376878911294328,
+      "loss": 0.4938,
+      "step": 106490
+    },
+    {
+      "epoch": 5.289559948346081,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0003768391775106785,
+      "loss": 0.5068,
+      "step": 106500
+    },
+    {
+      "epoch": 5.290056620641701,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00037679944372702895,
+      "loss": 0.5231,
+      "step": 106510
+    },
+    {
+      "epoch": 5.29055329293732,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00037675970994337937,
+      "loss": 0.5003,
+      "step": 106520
+    },
+    {
+      "epoch": 5.291049965232939,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00037671997615972984,
+      "loss": 0.4978,
+      "step": 106530
+    },
+    {
+      "epoch": 5.291546637528558,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00037668024237608026,
+      "loss": 0.4801,
+      "step": 106540
+    },
+    {
+      "epoch": 5.292043309824178,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00037664050859243073,
+      "loss": 0.5235,
+      "step": 106550
+    },
+    {
+      "epoch": 5.292539982119798,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0003766007748087812,
+      "loss": 0.4832,
+      "step": 106560
+    },
+    {
+      "epoch": 5.293036654415417,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0003765610410251316,
+      "loss": 0.5165,
+      "step": 106570
+    },
+    {
+      "epoch": 5.293533326711036,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0003765213072414821,
+      "loss": 0.5267,
+      "step": 106580
+    },
+    {
+      "epoch": 5.294029999006655,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00037648157345783256,
+      "loss": 0.5177,
+      "step": 106590
+    },
+    {
+      "epoch": 5.2945266713022745,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.000376441839674183,
+      "loss": 0.5257,
+      "step": 106600
+    },
+    {
+      "epoch": 5.295023343597894,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00037640210589053345,
+      "loss": 0.498,
+      "step": 106610
+    },
+    {
+      "epoch": 5.295520015893514,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003763623721068839,
+      "loss": 0.5108,
+      "step": 106620
+    },
+    {
+      "epoch": 5.296016688189133,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00037632263832323434,
+      "loss": 0.4712,
+      "step": 106630
+    },
+    {
+      "epoch": 5.296513360484752,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003762829045395848,
+      "loss": 0.4676,
+      "step": 106640
+    },
+    {
+      "epoch": 5.2970100327803715,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003762431707559352,
+      "loss": 0.4873,
+      "step": 106650
+    },
+    {
+      "epoch": 5.297506705075991,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003762034369722857,
+      "loss": 0.5055,
+      "step": 106660
+    },
+    {
+      "epoch": 5.29800337737161,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00037616370318863617,
+      "loss": 0.507,
+      "step": 106670
+    },
+    {
+      "epoch": 5.298500049667229,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0003761239694049866,
+      "loss": 0.5004,
+      "step": 106680
+    },
+    {
+      "epoch": 5.298996721962849,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00037608423562133706,
+      "loss": 0.4958,
+      "step": 106690
+    },
+    {
+      "epoch": 5.2994933942584685,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00037604450183768753,
+      "loss": 0.5015,
+      "step": 106700
+    },
+    {
+      "epoch": 5.299990066554088,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00037600476805403795,
+      "loss": 0.5173,
+      "step": 106710
+    },
+    {
+      "epoch": 5.300486738849707,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003759650342703884,
+      "loss": 0.5207,
+      "step": 106720
+    },
+    {
+      "epoch": 5.300983411145326,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003759253004867389,
+      "loss": 0.506,
+      "step": 106730
+    },
+    {
+      "epoch": 5.3014800834409455,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003758855667030893,
+      "loss": 0.4992,
+      "step": 106740
+    },
+    {
+      "epoch": 5.301976755736565,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003758458329194398,
+      "loss": 0.4862,
+      "step": 106750
+    },
+    {
+      "epoch": 5.302473428032185,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0003758060991357902,
+      "loss": 0.5146,
+      "step": 106760
+    },
+    {
+      "epoch": 5.302970100327804,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0003757663653521407,
+      "loss": 0.5075,
+      "step": 106770
+    },
+    {
+      "epoch": 5.303466772623423,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00037572663156849114,
+      "loss": 0.5059,
+      "step": 106780
+    },
+    {
+      "epoch": 5.3039634449190425,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00037568689778484155,
+      "loss": 0.5584,
+      "step": 106790
+    },
+    {
+      "epoch": 5.304460117214662,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.000375647164001192,
+      "loss": 0.527,
+      "step": 106800
+    },
+    {
+      "epoch": 5.304956789510281,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003756074302175425,
+      "loss": 0.5224,
+      "step": 106810
+    },
+    {
+      "epoch": 5.3054534618059,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00037556769643389297,
+      "loss": 0.4929,
+      "step": 106820
+    },
+    {
+      "epoch": 5.30595013410152,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0003755279626502434,
+      "loss": 0.5004,
+      "step": 106830
+    },
+    {
+      "epoch": 5.3064468063971395,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003754882288665938,
+      "loss": 0.5135,
+      "step": 106840
+    },
+    {
+      "epoch": 5.306943478692759,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00037544849508294433,
+      "loss": 0.523,
+      "step": 106850
+    },
+    {
+      "epoch": 5.307440150988378,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00037540876129929474,
+      "loss": 0.5097,
+      "step": 106860
+    },
+    {
+      "epoch": 5.307936823283997,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00037536902751564516,
+      "loss": 0.4743,
+      "step": 106870
+    },
+    {
+      "epoch": 5.308433495579616,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00037532929373199563,
+      "loss": 0.4911,
+      "step": 106880
+    },
+    {
+      "epoch": 5.308930167875236,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0003752895599483461,
+      "loss": 0.4959,
+      "step": 106890
+    },
+    {
+      "epoch": 5.309426840170855,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0003752498261646966,
+      "loss": 0.4944,
+      "step": 106900
+    },
+    {
+      "epoch": 5.309923512466475,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.000375210092381047,
+      "loss": 0.4845,
+      "step": 106910
+    },
+    {
+      "epoch": 5.310420184762094,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00037517035859739746,
+      "loss": 0.5044,
+      "step": 106920
+    },
+    {
+      "epoch": 5.310916857057713,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00037513062481374794,
+      "loss": 0.5066,
+      "step": 106930
+    },
+    {
+      "epoch": 5.311413529353333,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00037509089103009835,
+      "loss": 0.5175,
+      "step": 106940
+    },
+    {
+      "epoch": 5.311910201648952,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00037505115724644877,
+      "loss": 0.4814,
+      "step": 106950
+    },
+    {
+      "epoch": 5.312406873944571,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003750114234627993,
+      "loss": 0.5106,
+      "step": 106960
+    },
+    {
+      "epoch": 5.31290354624019,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0003749716896791497,
+      "loss": 0.4795,
+      "step": 106970
+    },
+    {
+      "epoch": 5.3134002185358105,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003749319558955002,
+      "loss": 0.5057,
+      "step": 106980
+    },
+    {
+      "epoch": 5.31389689083143,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003748922221118506,
+      "loss": 0.4873,
+      "step": 106990
+    },
+    {
+      "epoch": 5.314393563127049,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00037485248832820107,
+      "loss": 0.5039,
+      "step": 107000
+    },
+    {
+      "epoch": 5.314890235422668,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00037481275454455154,
+      "loss": 0.5208,
+      "step": 107010
+    },
+    {
+      "epoch": 5.315386907718287,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00037477302076090196,
+      "loss": 0.4899,
+      "step": 107020
+    },
+    {
+      "epoch": 5.315883580013907,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00037473328697725243,
+      "loss": 0.5226,
+      "step": 107030
+    },
+    {
+      "epoch": 5.316380252309526,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0003746935531936029,
+      "loss": 0.5062,
+      "step": 107040
+    },
+    {
+      "epoch": 5.316876924605146,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0003746538194099533,
+      "loss": 0.5017,
+      "step": 107050
+    },
+    {
+      "epoch": 5.317373596900765,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0003746140856263038,
+      "loss": 0.5258,
+      "step": 107060
+    },
+    {
+      "epoch": 5.317870269196384,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00037457435184265426,
+      "loss": 0.4831,
+      "step": 107070
+    },
+    {
+      "epoch": 5.318366941492004,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0003745346180590047,
+      "loss": 0.4974,
+      "step": 107080
+    },
+    {
+      "epoch": 5.318863613787623,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00037449488427535515,
+      "loss": 0.5158,
+      "step": 107090
+    },
+    {
+      "epoch": 5.319360286083242,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00037445515049170557,
+      "loss": 0.5136,
+      "step": 107100
+    },
+    {
+      "epoch": 5.319856958378861,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00037441541670805604,
+      "loss": 0.4901,
+      "step": 107110
+    },
+    {
+      "epoch": 5.3203536306744805,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003743756829244065,
+      "loss": 0.4909,
+      "step": 107120
+    },
+    {
+      "epoch": 5.320850302970101,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00037433594914075693,
+      "loss": 0.5183,
+      "step": 107130
+    },
+    {
+      "epoch": 5.32134697526572,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0003742962153571074,
+      "loss": 0.4713,
+      "step": 107140
+    },
+    {
+      "epoch": 5.321843647561339,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00037425648157345787,
+      "loss": 0.5129,
+      "step": 107150
+    },
+    {
+      "epoch": 5.322340319856958,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0003742167477898083,
+      "loss": 0.4957,
+      "step": 107160
+    },
+    {
+      "epoch": 5.3228369921525776,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00037417701400615876,
+      "loss": 0.4968,
+      "step": 107170
+    },
+    {
+      "epoch": 5.323333664448197,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003741372802225092,
+      "loss": 0.4901,
+      "step": 107180
+    },
+    {
+      "epoch": 5.323830336743816,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00037409754643885965,
+      "loss": 0.5394,
+      "step": 107190
+    },
+    {
+      "epoch": 5.324327009039436,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003740578126552101,
+      "loss": 0.5215,
+      "step": 107200
+    },
+    {
+      "epoch": 5.324823681335055,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00037401807887156054,
+      "loss": 0.4856,
+      "step": 107210
+    },
+    {
+      "epoch": 5.325320353630675,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.000373978345087911,
+      "loss": 0.5135,
+      "step": 107220
+    },
+    {
+      "epoch": 5.325817025926294,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003739386113042615,
+      "loss": 0.5232,
+      "step": 107230
+    },
+    {
+      "epoch": 5.326313698221913,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0003738988775206119,
+      "loss": 0.5302,
+      "step": 107240
+    },
+    {
+      "epoch": 5.326810370517532,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00037385914373696237,
+      "loss": 0.4882,
+      "step": 107250
+    },
+    {
+      "epoch": 5.3273070428131515,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00037381940995331284,
+      "loss": 0.5092,
+      "step": 107260
+    },
+    {
+      "epoch": 5.327803715108772,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0003737796761696633,
+      "loss": 0.533,
+      "step": 107270
+    },
+    {
+      "epoch": 5.328300387404391,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00037373994238601373,
+      "loss": 0.5166,
+      "step": 107280
+    },
+    {
+      "epoch": 5.32879705970001,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00037370020860236415,
+      "loss": 0.5068,
+      "step": 107290
+    },
+    {
+      "epoch": 5.329293731995629,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00037366047481871467,
+      "loss": 0.4994,
+      "step": 107300
+    },
+    {
+      "epoch": 5.3297904042912485,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0003736207410350651,
+      "loss": 0.5246,
+      "step": 107310
+    },
+    {
+      "epoch": 5.330287076586868,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0003735810072514155,
+      "loss": 0.4979,
+      "step": 107320
+    },
+    {
+      "epoch": 5.330783748882487,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.000373541273467766,
+      "loss": 0.4869,
+      "step": 107330
+    },
+    {
+      "epoch": 5.331280421178107,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00037350153968411645,
+      "loss": 0.4768,
+      "step": 107340
+    },
+    {
+      "epoch": 5.331777093473726,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003734618059004669,
+      "loss": 0.478,
+      "step": 107350
+    },
+    {
+      "epoch": 5.3322737657693455,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00037342207211681734,
+      "loss": 0.5075,
+      "step": 107360
+    },
+    {
+      "epoch": 5.332770438064965,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.00037338233833316775,
+      "loss": 0.4907,
+      "step": 107370
+    },
+    {
+      "epoch": 5.333267110360584,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0003733426045495183,
+      "loss": 0.5123,
+      "step": 107380
+    },
+    {
+      "epoch": 5.333763782656203,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003733028707658687,
+      "loss": 0.4853,
+      "step": 107390
+    },
+    {
+      "epoch": 5.334260454951822,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003732631369822191,
+      "loss": 0.5294,
+      "step": 107400
+    },
+    {
+      "epoch": 5.3347571272474426,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003732234031985696,
+      "loss": 0.5243,
+      "step": 107410
+    },
+    {
+      "epoch": 5.335253799543062,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00037318366941492006,
+      "loss": 0.4995,
+      "step": 107420
+    },
+    {
+      "epoch": 5.335750471838681,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00037314393563127053,
+      "loss": 0.4997,
+      "step": 107430
+    },
+    {
+      "epoch": 5.3362471441343,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00037310420184762094,
+      "loss": 0.4856,
+      "step": 107440
+    },
+    {
+      "epoch": 5.3367438164299195,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0003730644680639714,
+      "loss": 0.4911,
+      "step": 107450
+    },
+    {
+      "epoch": 5.337240488725539,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0003730247342803219,
+      "loss": 0.5015,
+      "step": 107460
+    },
+    {
+      "epoch": 5.337737161021158,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003729850004966723,
+      "loss": 0.4819,
+      "step": 107470
+    },
+    {
+      "epoch": 5.338233833316778,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0003729452667130227,
+      "loss": 0.4923,
+      "step": 107480
+    },
+    {
+      "epoch": 5.338730505612397,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00037290553292937325,
+      "loss": 0.4997,
+      "step": 107490
+    },
+    {
+      "epoch": 5.3392271779080165,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00037286579914572366,
+      "loss": 0.502,
+      "step": 107500
+    },
+    {
+      "epoch": 5.339723850203636,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00037282606536207414,
+      "loss": 0.5228,
+      "step": 107510
+    },
+    {
+      "epoch": 5.340220522499255,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00037278633157842455,
+      "loss": 0.4805,
+      "step": 107520
+    },
+    {
+      "epoch": 5.340717194794874,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.000372746597794775,
+      "loss": 0.5407,
+      "step": 107530
+    },
+    {
+      "epoch": 5.341213867090493,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0003727068640111255,
+      "loss": 0.523,
+      "step": 107540
+    },
+    {
+      "epoch": 5.3417105393861135,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0003726671302274759,
+      "loss": 0.4976,
+      "step": 107550
+    },
+    {
+      "epoch": 5.342207211681733,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003726273964438264,
+      "loss": 0.4908,
+      "step": 107560
+    },
+    {
+      "epoch": 5.342703883977352,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00037258766266017686,
+      "loss": 0.5216,
+      "step": 107570
+    },
+    {
+      "epoch": 5.343200556272971,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00037254792887652727,
+      "loss": 0.4958,
+      "step": 107580
+    },
+    {
+      "epoch": 5.34369722856859,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00037250819509287774,
+      "loss": 0.4956,
+      "step": 107590
+    },
+    {
+      "epoch": 5.34419390086421,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0003724684613092282,
+      "loss": 0.4909,
+      "step": 107600
+    },
+    {
+      "epoch": 5.344690573159829,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00037242872752557863,
+      "loss": 0.4814,
+      "step": 107610
+    },
+    {
+      "epoch": 5.345187245455448,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003723889937419291,
+      "loss": 0.4906,
+      "step": 107620
+    },
+    {
+      "epoch": 5.345683917751068,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003723492599582795,
+      "loss": 0.4998,
+      "step": 107630
+    },
+    {
+      "epoch": 5.346180590046687,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00037230952617463,
+      "loss": 0.4901,
+      "step": 107640
+    },
+    {
+      "epoch": 5.346677262342307,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00037226979239098046,
+      "loss": 0.4998,
+      "step": 107650
+    },
+    {
+      "epoch": 5.347173934637926,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0003722300586073309,
+      "loss": 0.4889,
+      "step": 107660
+    },
+    {
+      "epoch": 5.347670606933545,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00037219032482368135,
+      "loss": 0.5287,
+      "step": 107670
+    },
+    {
+      "epoch": 5.348167279229164,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003721505910400318,
+      "loss": 0.5008,
+      "step": 107680
+    },
+    {
+      "epoch": 5.348663951524784,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00037211085725638224,
+      "loss": 0.4908,
+      "step": 107690
+    },
+    {
+      "epoch": 5.349160623820404,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0003720711234727327,
+      "loss": 0.4943,
+      "step": 107700
+    },
+    {
+      "epoch": 5.349657296116023,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00037203138968908313,
+      "loss": 0.5044,
+      "step": 107710
+    },
+    {
+      "epoch": 5.350153968411642,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00037199165590543365,
+      "loss": 0.4952,
+      "step": 107720
+    },
+    {
+      "epoch": 5.350650640707261,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00037195192212178407,
+      "loss": 0.5112,
+      "step": 107730
+    },
+    {
+      "epoch": 5.351147313002881,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003719121883381345,
+      "loss": 0.4978,
+      "step": 107740
+    },
+    {
+      "epoch": 5.3516439852985,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00037187245455448496,
+      "loss": 0.4872,
+      "step": 107750
+    },
+    {
+      "epoch": 5.352140657594119,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00037183272077083543,
+      "loss": 0.5244,
+      "step": 107760
+    },
+    {
+      "epoch": 5.352637329889738,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00037179298698718585,
+      "loss": 0.5032,
+      "step": 107770
+    },
+    {
+      "epoch": 5.353134002185358,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0003717532532035363,
+      "loss": 0.5246,
+      "step": 107780
+    },
+    {
+      "epoch": 5.353630674480978,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0003717135194198868,
+      "loss": 0.4951,
+      "step": 107790
+    },
+    {
+      "epoch": 5.354127346776597,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00037167378563623726,
+      "loss": 0.5311,
+      "step": 107800
+    },
+    {
+      "epoch": 5.354624019072216,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0003716340518525877,
+      "loss": 0.5226,
+      "step": 107810
+    },
+    {
+      "epoch": 5.355120691367835,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0003715943180689381,
+      "loss": 0.5087,
+      "step": 107820
+    },
+    {
+      "epoch": 5.3556173636634545,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003715545842852886,
+      "loss": 0.4915,
+      "step": 107830
+    },
+    {
+      "epoch": 5.356114035959074,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.00037151485050163904,
+      "loss": 0.4851,
+      "step": 107840
+    },
+    {
+      "epoch": 5.356610708254694,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00037147511671798946,
+      "loss": 0.4998,
+      "step": 107850
+    },
+    {
+      "epoch": 5.357107380550313,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00037143538293433993,
+      "loss": 0.5159,
+      "step": 107860
+    },
+    {
+      "epoch": 5.357604052845932,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0003713956491506904,
+      "loss": 0.5063,
+      "step": 107870
+    },
+    {
+      "epoch": 5.358100725141552,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00037135591536704087,
+      "loss": 0.505,
+      "step": 107880
+    },
+    {
+      "epoch": 5.358597397437171,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0003713161815833913,
+      "loss": 0.5133,
+      "step": 107890
+    },
+    {
+      "epoch": 5.35909406973279,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00037127644779974176,
+      "loss": 0.489,
+      "step": 107900
+    },
+    {
+      "epoch": 5.359590742028409,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00037123671401609223,
+      "loss": 0.5074,
+      "step": 107910
+    },
+    {
+      "epoch": 5.360087414324029,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00037119698023244265,
+      "loss": 0.4719,
+      "step": 107920
+    },
+    {
+      "epoch": 5.360584086619649,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00037115724644879307,
+      "loss": 0.4836,
+      "step": 107930
+    },
+    {
+      "epoch": 5.361080758915268,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00037111751266514354,
+      "loss": 0.5127,
+      "step": 107940
+    },
+    {
+      "epoch": 5.361577431210887,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.000371077778881494,
+      "loss": 0.5171,
+      "step": 107950
+    },
+    {
+      "epoch": 5.362074103506506,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003710380450978445,
+      "loss": 0.4905,
+      "step": 107960
+    },
+    {
+      "epoch": 5.3625707758021255,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0003709983113141949,
+      "loss": 0.4623,
+      "step": 107970
+    },
+    {
+      "epoch": 5.363067448097745,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00037095857753054537,
+      "loss": 0.5042,
+      "step": 107980
+    },
+    {
+      "epoch": 5.363564120393365,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00037091884374689584,
+      "loss": 0.523,
+      "step": 107990
+    },
+    {
+      "epoch": 5.364060792688984,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00037087910996324626,
+      "loss": 0.4975,
+      "step": 108000
+    },
+    {
+      "epoch": 5.364557464984603,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0003708393761795967,
+      "loss": 0.472,
+      "step": 108010
+    },
+    {
+      "epoch": 5.3650541372802225,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0003707996423959472,
+      "loss": 0.5176,
+      "step": 108020
+    },
+    {
+      "epoch": 5.365550809575842,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0003707599086122976,
+      "loss": 0.505,
+      "step": 108030
+    },
+    {
+      "epoch": 5.366047481871461,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003707201748286481,
+      "loss": 0.522,
+      "step": 108040
+    },
+    {
+      "epoch": 5.36654415416708,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0003706804410449985,
+      "loss": 0.503,
+      "step": 108050
+    },
+    {
+      "epoch": 5.3670408264627,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.000370640707261349,
+      "loss": 0.5157,
+      "step": 108060
+    },
+    {
+      "epoch": 5.3675374987583195,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00037060097347769945,
+      "loss": 0.497,
+      "step": 108070
+    },
+    {
+      "epoch": 5.368034171053939,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00037056123969404986,
+      "loss": 0.4749,
+      "step": 108080
+    },
+    {
+      "epoch": 5.368530843349558,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00037052150591040034,
+      "loss": 0.4949,
+      "step": 108090
+    },
+    {
+      "epoch": 5.369027515645177,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003704817721267508,
+      "loss": 0.4938,
+      "step": 108100
+    },
+    {
+      "epoch": 5.3695241879407964,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003704420383431012,
+      "loss": 0.48,
+      "step": 108110
+    },
+    {
+      "epoch": 5.370020860236416,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003704023045594517,
+      "loss": 0.536,
+      "step": 108120
+    },
+    {
+      "epoch": 5.370517532532036,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.00037036257077580217,
+      "loss": 0.4826,
+      "step": 108130
+    },
+    {
+      "epoch": 5.371014204827655,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003703228369921526,
+      "loss": 0.4985,
+      "step": 108140
+    },
+    {
+      "epoch": 5.371510877123274,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00037028310320850306,
+      "loss": 0.5027,
+      "step": 108150
+    },
+    {
+      "epoch": 5.3720075494188935,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00037024336942485347,
+      "loss": 0.4979,
+      "step": 108160
+    },
+    {
+      "epoch": 5.372504221714513,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.000370203635641204,
+      "loss": 0.4905,
+      "step": 108170
+    },
+    {
+      "epoch": 5.373000894010132,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0003701639018575544,
+      "loss": 0.5008,
+      "step": 108180
+    },
+    {
+      "epoch": 5.373497566305751,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00037012416807390483,
+      "loss": 0.4952,
+      "step": 108190
+    },
+    {
+      "epoch": 5.373994238601371,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003700844342902553,
+      "loss": 0.5018,
+      "step": 108200
+    },
+    {
+      "epoch": 5.3744909108969905,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0003700447005066058,
+      "loss": 0.486,
+      "step": 108210
+    },
+    {
+      "epoch": 5.37498758319261,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003700049667229562,
+      "loss": 0.5243,
+      "step": 108220
+    },
+    {
+      "epoch": 5.375484255488229,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00036996523293930666,
+      "loss": 0.4858,
+      "step": 108230
+    },
+    {
+      "epoch": 5.375980927783848,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0003699254991556571,
+      "loss": 0.4943,
+      "step": 108240
+    },
+    {
+      "epoch": 5.376477600079467,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0003698857653720076,
+      "loss": 0.5185,
+      "step": 108250
+    },
+    {
+      "epoch": 5.376974272375087,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.000369846031588358,
+      "loss": 0.4881,
+      "step": 108260
+    },
+    {
+      "epoch": 5.377470944670706,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00036980629780470844,
+      "loss": 0.501,
+      "step": 108270
+    },
+    {
+      "epoch": 5.377967616966326,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003697665640210589,
+      "loss": 0.5114,
+      "step": 108280
+    },
+    {
+      "epoch": 5.378464289261945,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003697268302374094,
+      "loss": 0.5176,
+      "step": 108290
+    },
+    {
+      "epoch": 5.378960961557564,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0003696870964537598,
+      "loss": 0.516,
+      "step": 108300
+    },
+    {
+      "epoch": 5.379457633853184,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00036964736267011027,
+      "loss": 0.5143,
+      "step": 108310
+    },
+    {
+      "epoch": 5.379954306148803,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00036960762888646074,
+      "loss": 0.5209,
+      "step": 108320
+    },
+    {
+      "epoch": 5.380450978444422,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0003695678951028112,
+      "loss": 0.5073,
+      "step": 108330
+    },
+    {
+      "epoch": 5.380947650740041,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00036952816131916163,
+      "loss": 0.4927,
+      "step": 108340
+    },
+    {
+      "epoch": 5.3814443230356614,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00036948842753551205,
+      "loss": 0.4988,
+      "step": 108350
+    },
+    {
+      "epoch": 5.381940995331281,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0003694486937518626,
+      "loss": 0.5168,
+      "step": 108360
+    },
+    {
+      "epoch": 5.3824376676269,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.000369408959968213,
+      "loss": 0.4888,
+      "step": 108370
+    },
+    {
+      "epoch": 5.382934339922519,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0003693692261845634,
+      "loss": 0.5031,
+      "step": 108380
+    },
+    {
+      "epoch": 5.383431012218138,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0003693294924009139,
+      "loss": 0.4912,
+      "step": 108390
+    },
+    {
+      "epoch": 5.383927684513758,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00036928975861726435,
+      "loss": 0.5249,
+      "step": 108400
+    },
+    {
+      "epoch": 5.384424356809377,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003692500248336148,
+      "loss": 0.4943,
+      "step": 108410
+    },
+    {
+      "epoch": 5.384921029104997,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00036921029104996524,
+      "loss": 0.4961,
+      "step": 108420
+    },
+    {
+      "epoch": 5.385417701400616,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0003691705572663157,
+      "loss": 0.521,
+      "step": 108430
+    },
+    {
+      "epoch": 5.385914373696235,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003691308234826662,
+      "loss": 0.5057,
+      "step": 108440
+    },
+    {
+      "epoch": 5.386411045991855,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0003690910896990166,
+      "loss": 0.4832,
+      "step": 108450
+    },
+    {
+      "epoch": 5.386907718287474,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.000369051355915367,
+      "loss": 0.4958,
+      "step": 108460
+    },
+    {
+      "epoch": 5.387404390583093,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00036901162213171754,
+      "loss": 0.5162,
+      "step": 108470
+    },
+    {
+      "epoch": 5.387901062878712,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00036897188834806796,
+      "loss": 0.4704,
+      "step": 108480
+    },
+    {
+      "epoch": 5.3883977351743315,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00036893215456441843,
+      "loss": 0.4988,
+      "step": 108490
+    },
+    {
+      "epoch": 5.388894407469952,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00036889242078076885,
+      "loss": 0.4893,
+      "step": 108500
+    },
+    {
+      "epoch": 5.389391079765571,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003688526869971193,
+      "loss": 0.509,
+      "step": 108510
+    },
+    {
+      "epoch": 5.38988775206119,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003688129532134698,
+      "loss": 0.5246,
+      "step": 108520
+    },
+    {
+      "epoch": 5.390384424356809,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003687732194298202,
+      "loss": 0.5198,
+      "step": 108530
+    },
+    {
+      "epoch": 5.3908810966524285,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0003687334856461707,
+      "loss": 0.4906,
+      "step": 108540
+    },
+    {
+      "epoch": 5.391377768948048,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00036869375186252115,
+      "loss": 0.494,
+      "step": 108550
+    },
+    {
+      "epoch": 5.391874441243667,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00036865401807887157,
+      "loss": 0.5274,
+      "step": 108560
+    },
+    {
+      "epoch": 5.392371113539287,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00036861428429522204,
+      "loss": 0.5298,
+      "step": 108570
+    },
+    {
+      "epoch": 5.392867785834906,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00036857455051157246,
+      "loss": 0.4937,
+      "step": 108580
+    },
+    {
+      "epoch": 5.393364458130526,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00036853481672792293,
+      "loss": 0.5091,
+      "step": 108590
+    },
+    {
+      "epoch": 5.393861130426145,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003684950829442734,
+      "loss": 0.5224,
+      "step": 108600
+    },
+    {
+      "epoch": 5.394357802721764,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003684553491606238,
+      "loss": 0.5062,
+      "step": 108610
+    },
+    {
+      "epoch": 5.394854475017383,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0003684156153769743,
+      "loss": 0.4836,
+      "step": 108620
+    },
+    {
+      "epoch": 5.3953511473130025,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00036837588159332476,
+      "loss": 0.5242,
+      "step": 108630
+    },
+    {
+      "epoch": 5.395847819608623,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003683361478096752,
+      "loss": 0.5123,
+      "step": 108640
+    },
+    {
+      "epoch": 5.396344491904242,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00036829641402602565,
+      "loss": 0.4747,
+      "step": 108650
+    },
+    {
+      "epoch": 5.396841164199861,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0003682566802423761,
+      "loss": 0.4758,
+      "step": 108660
+    },
+    {
+      "epoch": 5.39733783649548,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00036821694645872654,
+      "loss": 0.5006,
+      "step": 108670
+    },
+    {
+      "epoch": 5.3978345087910995,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.000368177212675077,
+      "loss": 0.4897,
+      "step": 108680
+    },
+    {
+      "epoch": 5.398331181086719,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0003681374788914274,
+      "loss": 0.5123,
+      "step": 108690
+    },
+    {
+      "epoch": 5.398827853382338,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00036809774510777795,
+      "loss": 0.4788,
+      "step": 108700
+    },
+    {
+      "epoch": 5.399324525677958,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00036805801132412837,
+      "loss": 0.4871,
+      "step": 108710
+    },
+    {
+      "epoch": 5.399821197973577,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003680182775404788,
+      "loss": 0.4988,
+      "step": 108720
+    },
+    {
+      "epoch": 5.4003178702691965,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00036797854375682926,
+      "loss": 0.49,
+      "step": 108730
+    },
+    {
+      "epoch": 5.400814542564816,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00036793880997317973,
+      "loss": 0.5178,
+      "step": 108740
+    },
+    {
+      "epoch": 5.401311214860435,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00036789907618953014,
+      "loss": 0.5088,
+      "step": 108750
+    },
+    {
+      "epoch": 5.401807887156054,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003678593424058806,
+      "loss": 0.502,
+      "step": 108760
+    },
+    {
+      "epoch": 5.402304559451673,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00036781960862223103,
+      "loss": 0.4923,
+      "step": 108770
+    },
+    {
+      "epoch": 5.4028012317472935,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00036777987483858156,
+      "loss": 0.4984,
+      "step": 108780
+    },
+    {
+      "epoch": 5.403297904042913,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.000367740141054932,
+      "loss": 0.5066,
+      "step": 108790
+    },
+    {
+      "epoch": 5.403794576338532,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0003677004072712824,
+      "loss": 0.5195,
+      "step": 108800
+    },
+    {
+      "epoch": 5.404291248634151,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00036766067348763286,
+      "loss": 0.5054,
+      "step": 108810
+    },
+    {
+      "epoch": 5.4047879209297705,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00036762093970398334,
+      "loss": 0.5097,
+      "step": 108820
+    },
+    {
+      "epoch": 5.40528459322539,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00036758120592033375,
+      "loss": 0.504,
+      "step": 108830
+    },
+    {
+      "epoch": 5.405781265521009,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003675414721366842,
+      "loss": 0.5155,
+      "step": 108840
+    },
+    {
+      "epoch": 5.406277937816629,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003675017383530347,
+      "loss": 0.4882,
+      "step": 108850
+    },
+    {
+      "epoch": 5.406774610112248,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00036746200456938517,
+      "loss": 0.5012,
+      "step": 108860
+    },
+    {
+      "epoch": 5.4072712824078675,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0003674222707857356,
+      "loss": 0.5125,
+      "step": 108870
+    },
+    {
+      "epoch": 5.407767954703487,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.000367382537002086,
+      "loss": 0.5049,
+      "step": 108880
+    },
+    {
+      "epoch": 5.408264626999106,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003673428032184365,
+      "loss": 0.5241,
+      "step": 108890
+    },
+    {
+      "epoch": 5.408761299294725,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00036730306943478694,
+      "loss": 0.5068,
+      "step": 108900
+    },
+    {
+      "epoch": 5.409257971590344,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003672633356511374,
+      "loss": 0.496,
+      "step": 108910
+    },
+    {
+      "epoch": 5.4097546438859645,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00036722360186748783,
+      "loss": 0.5095,
+      "step": 108920
+    },
+    {
+      "epoch": 5.410251316181584,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0003671838680838383,
+      "loss": 0.503,
+      "step": 108930
+    },
+    {
+      "epoch": 5.410747988477203,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003671441343001888,
+      "loss": 0.4936,
+      "step": 108940
+    },
+    {
+      "epoch": 5.411244660772822,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0003671044005165392,
+      "loss": 0.4863,
+      "step": 108950
+    },
+    {
+      "epoch": 5.411741333068441,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00036706466673288966,
+      "loss": 0.5501,
+      "step": 108960
+    },
+    {
+      "epoch": 5.412238005364061,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00036702493294924013,
+      "loss": 0.4985,
+      "step": 108970
+    },
+    {
+      "epoch": 5.41273467765968,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00036698519916559055,
+      "loss": 0.5307,
+      "step": 108980
+    },
+    {
+      "epoch": 5.413231349955299,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.000366945465381941,
+      "loss": 0.5281,
+      "step": 108990
+    },
+    {
+      "epoch": 5.413728022250919,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003669057315982915,
+      "loss": 0.5336,
+      "step": 109000
+    },
+    {
+      "epoch": 5.414224694546538,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0003668659978146419,
+      "loss": 0.4793,
+      "step": 109010
+    },
+    {
+      "epoch": 5.414721366842158,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0003668262640309924,
+      "loss": 0.5175,
+      "step": 109020
+    },
+    {
+      "epoch": 5.415218039137777,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003667865302473428,
+      "loss": 0.4748,
+      "step": 109030
+    },
+    {
+      "epoch": 5.415714711433396,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00036674679646369327,
+      "loss": 0.5304,
+      "step": 109040
+    },
+    {
+      "epoch": 5.416211383729015,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00036670706268004374,
+      "loss": 0.5149,
+      "step": 109050
+    },
+    {
+      "epoch": 5.416708056024635,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00036666732889639416,
+      "loss": 0.4983,
+      "step": 109060
+    },
+    {
+      "epoch": 5.417204728320255,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00036662759511274463,
+      "loss": 0.5059,
+      "step": 109070
+    },
+    {
+      "epoch": 5.417701400615874,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0003665878613290951,
+      "loss": 0.4924,
+      "step": 109080
+    },
+    {
+      "epoch": 5.418198072911493,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0003665481275454455,
+      "loss": 0.4965,
+      "step": 109090
+    },
+    {
+      "epoch": 5.418694745207112,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.000366508393761796,
+      "loss": 0.4815,
+      "step": 109100
+    },
+    {
+      "epoch": 5.419191417502732,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0003664686599781464,
+      "loss": 0.5138,
+      "step": 109110
+    },
+    {
+      "epoch": 5.419688089798351,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003664289261944969,
+      "loss": 0.4985,
+      "step": 109120
+    },
+    {
+      "epoch": 5.42018476209397,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00036638919241084735,
+      "loss": 0.5111,
+      "step": 109130
+    },
+    {
+      "epoch": 5.42068143438959,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00036634945862719777,
+      "loss": 0.5004,
+      "step": 109140
+    },
+    {
+      "epoch": 5.421178106685209,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00036630972484354824,
+      "loss": 0.4852,
+      "step": 109150
+    },
+    {
+      "epoch": 5.421674778980829,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003662699910598987,
+      "loss": 0.5202,
+      "step": 109160
+    },
+    {
+      "epoch": 5.422171451276448,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00036623025727624913,
+      "loss": 0.4982,
+      "step": 109170
+    },
+    {
+      "epoch": 5.422668123572067,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0003661905234925996,
+      "loss": 0.4969,
+      "step": 109180
+    },
+    {
+      "epoch": 5.423164795867686,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00036615078970895007,
+      "loss": 0.5026,
+      "step": 109190
+    },
+    {
+      "epoch": 5.4236614681633055,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0003661110559253005,
+      "loss": 0.5074,
+      "step": 109200
+    },
+    {
+      "epoch": 5.424158140458925,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00036607132214165096,
+      "loss": 0.4977,
+      "step": 109210
+    },
+    {
+      "epoch": 5.424654812754545,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0003660315883580014,
+      "loss": 0.5052,
+      "step": 109220
+    },
+    {
+      "epoch": 5.425151485050164,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003659918545743519,
+      "loss": 0.5185,
+      "step": 109230
+    },
+    {
+      "epoch": 5.425648157345783,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0003659521207907023,
+      "loss": 0.5034,
+      "step": 109240
+    },
+    {
+      "epoch": 5.4261448296414025,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00036591238700705274,
+      "loss": 0.5235,
+      "step": 109250
+    },
+    {
+      "epoch": 5.426641501937022,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003658726532234032,
+      "loss": 0.5166,
+      "step": 109260
+    },
+    {
+      "epoch": 5.427138174232641,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0003658329194397537,
+      "loss": 0.4626,
+      "step": 109270
+    },
+    {
+      "epoch": 5.42763484652826,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003657931856561041,
+      "loss": 0.5009,
+      "step": 109280
+    },
+    {
+      "epoch": 5.42813151882388,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00036575345187245457,
+      "loss": 0.4857,
+      "step": 109290
+    },
+    {
+      "epoch": 5.4286281911195,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00036571371808880504,
+      "loss": 0.5296,
+      "step": 109300
+    },
+    {
+      "epoch": 5.429124863415119,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003656739843051555,
+      "loss": 0.5115,
+      "step": 109310
+    },
+    {
+      "epoch": 5.429621535710738,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00036563425052150593,
+      "loss": 0.506,
+      "step": 109320
+    },
+    {
+      "epoch": 5.430118208006357,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00036559451673785634,
+      "loss": 0.4808,
+      "step": 109330
+    },
+    {
+      "epoch": 5.4306148803019765,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0003655547829542068,
+      "loss": 0.513,
+      "step": 109340
+    },
+    {
+      "epoch": 5.431111552597596,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0003655150491705573,
+      "loss": 0.4674,
+      "step": 109350
+    },
+    {
+      "epoch": 5.431608224893216,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00036547531538690776,
+      "loss": 0.5004,
+      "step": 109360
+    },
+    {
+      "epoch": 5.432104897188835,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003654355816032582,
+      "loss": 0.5174,
+      "step": 109370
+    },
+    {
+      "epoch": 5.432601569484454,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00036539584781960865,
+      "loss": 0.5371,
+      "step": 109380
+    },
+    {
+      "epoch": 5.4330982417800735,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003653561140359591,
+      "loss": 0.5203,
+      "step": 109390
+    },
+    {
+      "epoch": 5.433594914075693,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00036531638025230954,
+      "loss": 0.4986,
+      "step": 109400
+    },
+    {
+      "epoch": 5.434091586371312,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00036527664646865995,
+      "loss": 0.4866,
+      "step": 109410
+    },
+    {
+      "epoch": 5.434588258666931,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003652369126850105,
+      "loss": 0.5068,
+      "step": 109420
+    },
+    {
+      "epoch": 5.435084930962551,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003651971789013609,
+      "loss": 0.5019,
+      "step": 109430
+    },
+    {
+      "epoch": 5.4355816032581705,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00036515744511771137,
+      "loss": 0.4791,
+      "step": 109440
+    },
+    {
+      "epoch": 5.43607827555379,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0003651177113340618,
+      "loss": 0.5114,
+      "step": 109450
+    },
+    {
+      "epoch": 5.436574947849409,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00036507797755041226,
+      "loss": 0.491,
+      "step": 109460
+    },
+    {
+      "epoch": 5.437071620145028,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003650382437667627,
+      "loss": 0.504,
+      "step": 109470
+    },
+    {
+      "epoch": 5.437568292440647,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00036499850998311314,
+      "loss": 0.5213,
+      "step": 109480
+    },
+    {
+      "epoch": 5.438064964736267,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003649587761994636,
+      "loss": 0.5141,
+      "step": 109490
+    },
+    {
+      "epoch": 5.438561637031887,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0003649190424158141,
+      "loss": 0.5127,
+      "step": 109500
+    },
+    {
+      "epoch": 5.439058309327506,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0003648793086321645,
+      "loss": 0.49,
+      "step": 109510
+    },
+    {
+      "epoch": 5.439554981623125,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.000364839574848515,
+      "loss": 0.5063,
+      "step": 109520
+    },
+    {
+      "epoch": 5.4400516539187445,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00036479984106486545,
+      "loss": 0.4929,
+      "step": 109530
+    },
+    {
+      "epoch": 5.440548326214364,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00036476010728121586,
+      "loss": 0.5066,
+      "step": 109540
+    },
+    {
+      "epoch": 5.441044998509983,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00036472037349756633,
+      "loss": 0.5232,
+      "step": 109550
+    },
+    {
+      "epoch": 5.441541670805602,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00036468063971391675,
+      "loss": 0.5274,
+      "step": 109560
+    },
+    {
+      "epoch": 5.442038343101222,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003646409059302672,
+      "loss": 0.491,
+      "step": 109570
+    },
+    {
+      "epoch": 5.4425350153968415,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0003646011721466177,
+      "loss": 0.4898,
+      "step": 109580
+    },
+    {
+      "epoch": 5.443031687692461,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0003645614383629681,
+      "loss": 0.5091,
+      "step": 109590
+    },
+    {
+      "epoch": 5.44352835998808,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003645217045793186,
+      "loss": 0.5018,
+      "step": 109600
+    },
+    {
+      "epoch": 5.444025032283699,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00036448197079566905,
+      "loss": 0.5004,
+      "step": 109610
+    },
+    {
+      "epoch": 5.444521704579318,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00036444223701201947,
+      "loss": 0.4917,
+      "step": 109620
+    },
+    {
+      "epoch": 5.445018376874938,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00036440250322836994,
+      "loss": 0.5041,
+      "step": 109630
+    },
+    {
+      "epoch": 5.445515049170558,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00036436276944472036,
+      "loss": 0.5227,
+      "step": 109640
+    },
+    {
+      "epoch": 5.446011721466177,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00036432303566107083,
+      "loss": 0.509,
+      "step": 109650
+    },
+    {
+      "epoch": 5.446508393761796,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003642833018774213,
+      "loss": 0.4936,
+      "step": 109660
+    },
+    {
+      "epoch": 5.447005066057415,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0003642435680937717,
+      "loss": 0.4939,
+      "step": 109670
+    },
+    {
+      "epoch": 5.447501738353035,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003642038343101222,
+      "loss": 0.4946,
+      "step": 109680
+    },
+    {
+      "epoch": 5.447998410648654,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00036416410052647266,
+      "loss": 0.5041,
+      "step": 109690
+    },
+    {
+      "epoch": 5.448495082944273,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0003641243667428231,
+      "loss": 0.4819,
+      "step": 109700
+    },
+    {
+      "epoch": 5.448991755239892,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00036408463295917355,
+      "loss": 0.5064,
+      "step": 109710
+    },
+    {
+      "epoch": 5.449488427535512,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.000364044899175524,
+      "loss": 0.5081,
+      "step": 109720
+    },
+    {
+      "epoch": 5.449985099831132,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003640051653918745,
+      "loss": 0.503,
+      "step": 109730
+    },
+    {
+      "epoch": 5.450481772126751,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0003639654316082249,
+      "loss": 0.52,
+      "step": 109740
+    },
+    {
+      "epoch": 5.45097844442237,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00036392569782457533,
+      "loss": 0.4944,
+      "step": 109750
+    },
+    {
+      "epoch": 5.451475116717989,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00036388596404092585,
+      "loss": 0.5124,
+      "step": 109760
+    },
+    {
+      "epoch": 5.451971789013609,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00036384623025727627,
+      "loss": 0.5257,
+      "step": 109770
+    },
+    {
+      "epoch": 5.452468461309228,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003638064964736267,
+      "loss": 0.5101,
+      "step": 109780
+    },
+    {
+      "epoch": 5.452965133604848,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00036376676268997716,
+      "loss": 0.4994,
+      "step": 109790
+    },
+    {
+      "epoch": 5.453461805900467,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00036372702890632763,
+      "loss": 0.5259,
+      "step": 109800
+    },
+    {
+      "epoch": 5.453958478196086,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0003636872951226781,
+      "loss": 0.4796,
+      "step": 109810
+    },
+    {
+      "epoch": 5.454455150491706,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0003636475613390285,
+      "loss": 0.477,
+      "step": 109820
+    },
+    {
+      "epoch": 5.454951822787325,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.000363607827555379,
+      "loss": 0.493,
+      "step": 109830
+    },
+    {
+      "epoch": 5.455448495082944,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00036356809377172946,
+      "loss": 0.516,
+      "step": 109840
+    },
+    {
+      "epoch": 5.455945167378563,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0003635283599880799,
+      "loss": 0.4964,
+      "step": 109850
+    },
+    {
+      "epoch": 5.4564418396741825,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003634886262044303,
+      "loss": 0.5178,
+      "step": 109860
+    },
+    {
+      "epoch": 5.456938511969803,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003634488924207808,
+      "loss": 0.5005,
+      "step": 109870
+    },
+    {
+      "epoch": 5.457435184265422,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00036340915863713124,
+      "loss": 0.4756,
+      "step": 109880
+    },
+    {
+      "epoch": 5.457931856561041,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0003633694248534817,
+      "loss": 0.5116,
+      "step": 109890
+    },
+    {
+      "epoch": 5.45842852885666,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00036332969106983213,
+      "loss": 0.511,
+      "step": 109900
+    },
+    {
+      "epoch": 5.4589252011522795,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003632899572861826,
+      "loss": 0.4732,
+      "step": 109910
+    },
+    {
+      "epoch": 5.459421873447899,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00036325022350253307,
+      "loss": 0.487,
+      "step": 109920
+    },
+    {
+      "epoch": 5.459918545743518,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003632104897188835,
+      "loss": 0.5186,
+      "step": 109930
+    },
+    {
+      "epoch": 5.460415218039138,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0003631707559352339,
+      "loss": 0.5075,
+      "step": 109940
+    },
+    {
+      "epoch": 5.460911890334757,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00036313102215158443,
+      "loss": 0.5035,
+      "step": 109950
+    },
+    {
+      "epoch": 5.4614085626303766,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.00036309128836793485,
+      "loss": 0.4867,
+      "step": 109960
+    },
+    {
+      "epoch": 5.461905234925996,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0003630515545842853,
+      "loss": 0.4873,
+      "step": 109970
+    },
+    {
+      "epoch": 5.462401907221615,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00036301182080063574,
+      "loss": 0.5273,
+      "step": 109980
+    },
+    {
+      "epoch": 5.462898579517234,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0003629720870169862,
+      "loss": 0.4841,
+      "step": 109990
+    },
+    {
+      "epoch": 5.4633952518128535,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003629323532333367,
+      "loss": 0.4807,
+      "step": 110000
+    },
+    {
+      "epoch": 5.463891924108474,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003628926194496871,
+      "loss": 0.5149,
+      "step": 110010
+    },
+    {
+      "epoch": 5.464388596404093,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00036285288566603757,
+      "loss": 0.5286,
+      "step": 110020
+    },
+    {
+      "epoch": 5.464885268699712,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00036281315188238804,
+      "loss": 0.5148,
+      "step": 110030
+    },
+    {
+      "epoch": 5.465381940995331,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00036277341809873846,
+      "loss": 0.5212,
+      "step": 110040
+    },
+    {
+      "epoch": 5.4658786132909505,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003627336843150889,
+      "loss": 0.5017,
+      "step": 110050
+    },
+    {
+      "epoch": 5.46637528558657,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0003626939505314394,
+      "loss": 0.5164,
+      "step": 110060
+    },
+    {
+      "epoch": 5.466871957882189,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0003626542167477898,
+      "loss": 0.5149,
+      "step": 110070
+    },
+    {
+      "epoch": 5.467368630177809,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0003626144829641403,
+      "loss": 0.4969,
+      "step": 110080
+    },
+    {
+      "epoch": 5.467865302473428,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0003625747491804907,
+      "loss": 0.5312,
+      "step": 110090
+    },
+    {
+      "epoch": 5.4683619747690475,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003625350153968412,
+      "loss": 0.4916,
+      "step": 110100
+    },
+    {
+      "epoch": 5.468858647064667,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00036249528161319165,
+      "loss": 0.4932,
+      "step": 110110
+    },
+    {
+      "epoch": 5.469355319360286,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00036245554782954206,
+      "loss": 0.5117,
+      "step": 110120
+    },
+    {
+      "epoch": 5.469851991655905,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00036241581404589254,
+      "loss": 0.5093,
+      "step": 110130
+    },
+    {
+      "epoch": 5.470348663951524,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.000362376080262243,
+      "loss": 0.5008,
+      "step": 110140
+    },
+    {
+      "epoch": 5.4708453362471445,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0003623363464785934,
+      "loss": 0.4835,
+      "step": 110150
+    },
+    {
+      "epoch": 5.471342008542764,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0003622966126949439,
+      "loss": 0.5138,
+      "step": 110160
+    },
+    {
+      "epoch": 5.471838680838383,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00036225687891129437,
+      "loss": 0.4963,
+      "step": 110170
+    },
+    {
+      "epoch": 5.472335353134002,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00036221714512764484,
+      "loss": 0.5067,
+      "step": 110180
+    },
+    {
+      "epoch": 5.472832025429621,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00036217741134399525,
+      "loss": 0.4944,
+      "step": 110190
+    },
+    {
+      "epoch": 5.473328697725241,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00036213767756034567,
+      "loss": 0.5065,
+      "step": 110200
+    },
+    {
+      "epoch": 5.47382537002086,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00036209794377669614,
+      "loss": 0.5064,
+      "step": 110210
+    },
+    {
+      "epoch": 5.47432204231648,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0003620582099930466,
+      "loss": 0.4773,
+      "step": 110220
+    },
+    {
+      "epoch": 5.474818714612099,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00036201847620939703,
+      "loss": 0.4551,
+      "step": 110230
+    },
+    {
+      "epoch": 5.4753153869077185,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0003619787424257475,
+      "loss": 0.4933,
+      "step": 110240
+    },
+    {
+      "epoch": 5.475812059203338,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.000361939008642098,
+      "loss": 0.4873,
+      "step": 110250
+    },
+    {
+      "epoch": 5.476308731498957,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00036189927485844845,
+      "loss": 0.4914,
+      "step": 110260
+    },
+    {
+      "epoch": 5.476805403794576,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00036185954107479886,
+      "loss": 0.518,
+      "step": 110270
+    },
+    {
+      "epoch": 5.477302076090195,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0003618198072911493,
+      "loss": 0.4907,
+      "step": 110280
+    },
+    {
+      "epoch": 5.4777987483858155,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003617800735074998,
+      "loss": 0.4896,
+      "step": 110290
+    },
+    {
+      "epoch": 5.478295420681435,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0003617403397238502,
+      "loss": 0.4898,
+      "step": 110300
+    },
+    {
+      "epoch": 5.478792092977054,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00036170060594020064,
+      "loss": 0.5006,
+      "step": 110310
+    },
+    {
+      "epoch": 5.479288765272673,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0003616608721565511,
+      "loss": 0.5047,
+      "step": 110320
+    },
+    {
+      "epoch": 5.479785437568292,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003616211383729016,
+      "loss": 0.5147,
+      "step": 110330
+    },
+    {
+      "epoch": 5.480282109863912,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00036158140458925205,
+      "loss": 0.5337,
+      "step": 110340
+    },
+    {
+      "epoch": 5.480778782159531,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00036154167080560247,
+      "loss": 0.5247,
+      "step": 110350
+    },
+    {
+      "epoch": 5.48127545445515,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00036150193702195294,
+      "loss": 0.5077,
+      "step": 110360
+    },
+    {
+      "epoch": 5.48177212675077,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0003614622032383034,
+      "loss": 0.4701,
+      "step": 110370
+    },
+    {
+      "epoch": 5.482268799046389,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00036142246945465383,
+      "loss": 0.5172,
+      "step": 110380
+    },
+    {
+      "epoch": 5.482765471342009,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00036138273567100425,
+      "loss": 0.5148,
+      "step": 110390
+    },
+    {
+      "epoch": 5.483262143637628,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003613430018873548,
+      "loss": 0.5129,
+      "step": 110400
+    },
+    {
+      "epoch": 5.483758815933247,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003613032681037052,
+      "loss": 0.4989,
+      "step": 110410
+    },
+    {
+      "epoch": 5.484255488228866,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00036126353432005566,
+      "loss": 0.4895,
+      "step": 110420
+    },
+    {
+      "epoch": 5.4847521605244856,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0003612238005364061,
+      "loss": 0.4899,
+      "step": 110430
+    },
+    {
+      "epoch": 5.485248832820106,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00036118406675275655,
+      "loss": 0.525,
+      "step": 110440
+    },
+    {
+      "epoch": 5.485745505115725,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.000361144332969107,
+      "loss": 0.4797,
+      "step": 110450
+    },
+    {
+      "epoch": 5.486242177411344,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00036110459918545744,
+      "loss": 0.526,
+      "step": 110460
+    },
+    {
+      "epoch": 5.486738849706963,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00036106486540180786,
+      "loss": 0.5061,
+      "step": 110470
+    },
+    {
+      "epoch": 5.487235522002583,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0003610251316181584,
+      "loss": 0.4925,
+      "step": 110480
+    },
+    {
+      "epoch": 5.487732194298202,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003609853978345088,
+      "loss": 0.4995,
+      "step": 110490
+    },
+    {
+      "epoch": 5.488228866593821,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00036094566405085927,
+      "loss": 0.495,
+      "step": 110500
+    },
+    {
+      "epoch": 5.488725538889441,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003609059302672097,
+      "loss": 0.4999,
+      "step": 110510
+    },
+    {
+      "epoch": 5.48922221118506,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00036086619648356016,
+      "loss": 0.4985,
+      "step": 110520
+    },
+    {
+      "epoch": 5.48971888348068,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00036082646269991063,
+      "loss": 0.4989,
+      "step": 110530
+    },
+    {
+      "epoch": 5.490215555776299,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00036078672891626105,
+      "loss": 0.5001,
+      "step": 110540
+    },
+    {
+      "epoch": 5.490712228071918,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003607469951326115,
+      "loss": 0.4665,
+      "step": 110550
+    },
+    {
+      "epoch": 5.491208900367537,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.000360707261348962,
+      "loss": 0.5204,
+      "step": 110560
+    },
+    {
+      "epoch": 5.4917055726631565,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0003606675275653124,
+      "loss": 0.4917,
+      "step": 110570
+    },
+    {
+      "epoch": 5.492202244958776,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003606277937816629,
+      "loss": 0.4965,
+      "step": 110580
+    },
+    {
+      "epoch": 5.492698917254396,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00036058805999801335,
+      "loss": 0.5131,
+      "step": 110590
+    },
+    {
+      "epoch": 5.493195589550015,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00036054832621436377,
+      "loss": 0.4856,
+      "step": 110600
+    },
+    {
+      "epoch": 5.493692261845634,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00036050859243071424,
+      "loss": 0.4854,
+      "step": 110610
+    },
+    {
+      "epoch": 5.4941889341412535,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00036046885864706466,
+      "loss": 0.4969,
+      "step": 110620
+    },
+    {
+      "epoch": 5.494685606436873,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003604291248634152,
+      "loss": 0.5093,
+      "step": 110630
+    },
+    {
+      "epoch": 5.495182278732492,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0003603893910797656,
+      "loss": 0.5054,
+      "step": 110640
+    },
+    {
+      "epoch": 5.495678951028111,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.000360349657296116,
+      "loss": 0.4877,
+      "step": 110650
+    },
+    {
+      "epoch": 5.496175623323731,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003603099235124665,
+      "loss": 0.5036,
+      "step": 110660
+    },
+    {
+      "epoch": 5.496672295619351,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00036027018972881696,
+      "loss": 0.5093,
+      "step": 110670
+    },
+    {
+      "epoch": 5.49716896791497,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0003602304559451674,
+      "loss": 0.4902,
+      "step": 110680
+    },
+    {
+      "epoch": 5.497665640210589,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00036019072216151785,
+      "loss": 0.4923,
+      "step": 110690
+    },
+    {
+      "epoch": 5.498162312506208,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003601509883778683,
+      "loss": 0.5205,
+      "step": 110700
+    },
+    {
+      "epoch": 5.4986589848018275,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0003601112545942188,
+      "loss": 0.5088,
+      "step": 110710
+    },
+    {
+      "epoch": 5.499155657097447,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003600715208105692,
+      "loss": 0.5202,
+      "step": 110720
+    },
+    {
+      "epoch": 5.499652329393067,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003600317870269196,
+      "loss": 0.5072,
+      "step": 110730
+    },
+    {
+      "epoch": 5.500149001688686,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0003599920532432701,
+      "loss": 0.5187,
+      "step": 110740
+    },
+    {
+      "epoch": 5.500645673984305,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00035995231945962057,
+      "loss": 0.4875,
+      "step": 110750
+    },
+    {
+      "epoch": 5.5011423462799245,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.000359912585675971,
+      "loss": 0.4937,
+      "step": 110760
+    },
+    {
+      "epoch": 5.501639018575544,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00035987285189232146,
+      "loss": 0.5154,
+      "step": 110770
+    },
+    {
+      "epoch": 5.502135690871163,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003598331181086719,
+      "loss": 0.521,
+      "step": 110780
+    },
+    {
+      "epoch": 5.502632363166782,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0003597933843250224,
+      "loss": 0.5052,
+      "step": 110790
+    },
+    {
+      "epoch": 5.503129035462402,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003597536505413728,
+      "loss": 0.4976,
+      "step": 110800
+    },
+    {
+      "epoch": 5.5036257077580215,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00035971391675772323,
+      "loss": 0.4796,
+      "step": 110810
+    },
+    {
+      "epoch": 5.504122380053641,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00035967418297407376,
+      "loss": 0.491,
+      "step": 110820
+    },
+    {
+      "epoch": 5.50461905234926,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003596344491904242,
+      "loss": 0.5089,
+      "step": 110830
+    },
+    {
+      "epoch": 5.505115724644879,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0003595947154067746,
+      "loss": 0.5173,
+      "step": 110840
+    },
+    {
+      "epoch": 5.505612396940498,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.00035955498162312506,
+      "loss": 0.4792,
+      "step": 110850
+    },
+    {
+      "epoch": 5.506109069236118,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00035951524783947553,
+      "loss": 0.5183,
+      "step": 110860
+    },
+    {
+      "epoch": 5.506605741531738,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.000359475514055826,
+      "loss": 0.4963,
+      "step": 110870
+    },
+    {
+      "epoch": 5.507102413827357,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0003594357802721764,
+      "loss": 0.5182,
+      "step": 110880
+    },
+    {
+      "epoch": 5.507599086122976,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0003593960464885269,
+      "loss": 0.5177,
+      "step": 110890
+    },
+    {
+      "epoch": 5.5080957584185954,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00035935631270487737,
+      "loss": 0.5087,
+      "step": 110900
+    },
+    {
+      "epoch": 5.508592430714215,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0003593165789212278,
+      "loss": 0.5081,
+      "step": 110910
+    },
+    {
+      "epoch": 5.509089103009834,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003592768451375782,
+      "loss": 0.5092,
+      "step": 110920
+    },
+    {
+      "epoch": 5.509585775305453,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0003592371113539287,
+      "loss": 0.501,
+      "step": 110930
+    },
+    {
+      "epoch": 5.510082447601073,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00035919737757027914,
+      "loss": 0.5157,
+      "step": 110940
+    },
+    {
+      "epoch": 5.5105791198966925,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0003591576437866296,
+      "loss": 0.4858,
+      "step": 110950
+    },
+    {
+      "epoch": 5.511075792192312,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00035911791000298003,
+      "loss": 0.5092,
+      "step": 110960
+    },
+    {
+      "epoch": 5.511572464487931,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0003590781762193305,
+      "loss": 0.5241,
+      "step": 110970
+    },
+    {
+      "epoch": 5.51206913678355,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.000359038442435681,
+      "loss": 0.4872,
+      "step": 110980
+    },
+    {
+      "epoch": 5.512565809079169,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003589987086520314,
+      "loss": 0.5347,
+      "step": 110990
+    },
+    {
+      "epoch": 5.513062481374789,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00035895897486838186,
+      "loss": 0.535,
+      "step": 111000
+    },
+    {
+      "epoch": 5.513559153670409,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00035891924108473233,
+      "loss": 0.4963,
+      "step": 111010
+    },
+    {
+      "epoch": 5.514055825966028,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00035887950730108275,
+      "loss": 0.4886,
+      "step": 111020
+    },
+    {
+      "epoch": 5.514552498261647,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0003588397735174332,
+      "loss": 0.5294,
+      "step": 111030
+    },
+    {
+      "epoch": 5.515049170557266,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00035880003973378364,
+      "loss": 0.5258,
+      "step": 111040
+    },
+    {
+      "epoch": 5.515545842852886,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003587603059501341,
+      "loss": 0.513,
+      "step": 111050
+    },
+    {
+      "epoch": 5.516042515148505,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003587205721664846,
+      "loss": 0.4839,
+      "step": 111060
+    },
+    {
+      "epoch": 5.516539187444124,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.000358680838382835,
+      "loss": 0.49,
+      "step": 111070
+    },
+    {
+      "epoch": 5.517035859739744,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00035864110459918547,
+      "loss": 0.545,
+      "step": 111080
+    },
+    {
+      "epoch": 5.517532532035363,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00035860137081553594,
+      "loss": 0.478,
+      "step": 111090
+    },
+    {
+      "epoch": 5.518029204330983,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00035856163703188636,
+      "loss": 0.4944,
+      "step": 111100
+    },
+    {
+      "epoch": 5.518525876626602,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00035852190324823683,
+      "loss": 0.5055,
+      "step": 111110
+    },
+    {
+      "epoch": 5.519022548922221,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0003584821694645873,
+      "loss": 0.4963,
+      "step": 111120
+    },
+    {
+      "epoch": 5.51951922121784,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0003584424356809377,
+      "loss": 0.522,
+      "step": 111130
+    },
+    {
+      "epoch": 5.52001589351346,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0003584027018972882,
+      "loss": 0.5008,
+      "step": 111140
+    },
+    {
+      "epoch": 5.520512565809079,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0003583629681136386,
+      "loss": 0.5025,
+      "step": 111150
+    },
+    {
+      "epoch": 5.521009238104699,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00035832323432998913,
+      "loss": 0.5103,
+      "step": 111160
+    },
+    {
+      "epoch": 5.521505910400318,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00035828350054633955,
+      "loss": 0.5237,
+      "step": 111170
+    },
+    {
+      "epoch": 5.522002582695937,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00035824376676268997,
+      "loss": 0.5219,
+      "step": 111180
+    },
+    {
+      "epoch": 5.522499254991557,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00035820403297904044,
+      "loss": 0.5106,
+      "step": 111190
+    },
+    {
+      "epoch": 5.522995927287176,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0003581642991953909,
+      "loss": 0.5001,
+      "step": 111200
+    },
+    {
+      "epoch": 5.523492599582795,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00035812456541174133,
+      "loss": 0.5162,
+      "step": 111210
+    },
+    {
+      "epoch": 5.523989271878414,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0003580848316280918,
+      "loss": 0.5089,
+      "step": 111220
+    },
+    {
+      "epoch": 5.5244859441740335,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00035804509784444227,
+      "loss": 0.5231,
+      "step": 111230
+    },
+    {
+      "epoch": 5.524982616469654,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00035800536406079274,
+      "loss": 0.4949,
+      "step": 111240
+    },
+    {
+      "epoch": 5.525479288765273,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00035796563027714316,
+      "loss": 0.5107,
+      "step": 111250
+    },
+    {
+      "epoch": 5.525975961060892,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0003579258964934936,
+      "loss": 0.5279,
+      "step": 111260
+    },
+    {
+      "epoch": 5.526472633356511,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003578861627098441,
+      "loss": 0.5147,
+      "step": 111270
+    },
+    {
+      "epoch": 5.5269693056521305,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0003578464289261945,
+      "loss": 0.4759,
+      "step": 111280
+    },
+    {
+      "epoch": 5.52746597794775,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00035780669514254494,
+      "loss": 0.4928,
+      "step": 111290
+    },
+    {
+      "epoch": 5.527962650243369,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0003577669613588954,
+      "loss": 0.4903,
+      "step": 111300
+    },
+    {
+      "epoch": 5.528459322538989,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0003577272275752459,
+      "loss": 0.5192,
+      "step": 111310
+    },
+    {
+      "epoch": 5.528955994834608,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00035768749379159635,
+      "loss": 0.5061,
+      "step": 111320
+    },
+    {
+      "epoch": 5.5294526671302275,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00035764776000794677,
+      "loss": 0.4941,
+      "step": 111330
+    },
+    {
+      "epoch": 5.529949339425847,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0003576080262242972,
+      "loss": 0.5264,
+      "step": 111340
+    },
+    {
+      "epoch": 5.530446011721466,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003575682924406477,
+      "loss": 0.4951,
+      "step": 111350
+    },
+    {
+      "epoch": 5.530942684017085,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0003575285586569981,
+      "loss": 0.5178,
+      "step": 111360
+    },
+    {
+      "epoch": 5.5314393563127044,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00035748882487334854,
+      "loss": 0.5282,
+      "step": 111370
+    },
+    {
+      "epoch": 5.531936028608325,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.000357449091089699,
+      "loss": 0.5222,
+      "step": 111380
+    },
+    {
+      "epoch": 5.532432700903944,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0003574093573060495,
+      "loss": 0.5135,
+      "step": 111390
+    },
+    {
+      "epoch": 5.532929373199563,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00035736962352239996,
+      "loss": 0.4777,
+      "step": 111400
+    },
+    {
+      "epoch": 5.533426045495182,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0003573298897387504,
+      "loss": 0.4899,
+      "step": 111410
+    },
+    {
+      "epoch": 5.5339227177908015,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00035729015595510085,
+      "loss": 0.524,
+      "step": 111420
+    },
+    {
+      "epoch": 5.534419390086421,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0003572504221714513,
+      "loss": 0.4869,
+      "step": 111430
+    },
+    {
+      "epoch": 5.53491606238204,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00035721068838780173,
+      "loss": 0.4676,
+      "step": 111440
+    },
+    {
+      "epoch": 5.53541273467766,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003571709546041522,
+      "loss": 0.4894,
+      "step": 111450
+    },
+    {
+      "epoch": 5.535909406973279,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003571312208205027,
+      "loss": 0.4867,
+      "step": 111460
+    },
+    {
+      "epoch": 5.5364060792688985,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003570914870368531,
+      "loss": 0.5043,
+      "step": 111470
+    },
+    {
+      "epoch": 5.536902751564518,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00035705175325320357,
+      "loss": 0.5002,
+      "step": 111480
+    },
+    {
+      "epoch": 5.537399423860137,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.000357012019469554,
+      "loss": 0.5109,
+      "step": 111490
+    },
+    {
+      "epoch": 5.537896096155756,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00035697228568590445,
+      "loss": 0.4679,
+      "step": 111500
+    },
+    {
+      "epoch": 5.538392768451375,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003569325519022549,
+      "loss": 0.4981,
+      "step": 111510
+    },
+    {
+      "epoch": 5.5388894407469955,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00035689281811860534,
+      "loss": 0.4789,
+      "step": 111520
+    },
+    {
+      "epoch": 5.539386113042615,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0003568530843349558,
+      "loss": 0.508,
+      "step": 111530
+    },
+    {
+      "epoch": 5.539882785338234,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003568133505513063,
+      "loss": 0.5006,
+      "step": 111540
+    },
+    {
+      "epoch": 5.540379457633853,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003567736167676567,
+      "loss": 0.4994,
+      "step": 111550
+    },
+    {
+      "epoch": 5.540876129929472,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003567338829840072,
+      "loss": 0.5043,
+      "step": 111560
+    },
+    {
+      "epoch": 5.541372802225092,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00035669414920035765,
+      "loss": 0.5191,
+      "step": 111570
+    },
+    {
+      "epoch": 5.541869474520711,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00035665441541670806,
+      "loss": 0.5158,
+      "step": 111580
+    },
+    {
+      "epoch": 5.542366146816331,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00035661468163305853,
+      "loss": 0.5008,
+      "step": 111590
+    },
+    {
+      "epoch": 5.54286281911195,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00035657494784940895,
+      "loss": 0.4865,
+      "step": 111600
+    },
+    {
+      "epoch": 5.5433594914075695,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003565352140657594,
+      "loss": 0.4886,
+      "step": 111610
+    },
+    {
+      "epoch": 5.543856163703189,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0003564954802821099,
+      "loss": 0.512,
+      "step": 111620
+    },
+    {
+      "epoch": 5.544352835998808,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003564557464984603,
+      "loss": 0.4671,
+      "step": 111630
+    },
+    {
+      "epoch": 5.544849508294427,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0003564160127148108,
+      "loss": 0.527,
+      "step": 111640
+    },
+    {
+      "epoch": 5.545346180590046,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00035637627893116125,
+      "loss": 0.506,
+      "step": 111650
+    },
+    {
+      "epoch": 5.5458428528856665,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00035633654514751167,
+      "loss": 0.5005,
+      "step": 111660
+    },
+    {
+      "epoch": 5.546339525181286,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00035629681136386214,
+      "loss": 0.5076,
+      "step": 111670
+    },
+    {
+      "epoch": 5.546836197476905,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00035625707758021256,
+      "loss": 0.496,
+      "step": 111680
+    },
+    {
+      "epoch": 5.547332869772524,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003562173437965631,
+      "loss": 0.4994,
+      "step": 111690
+    },
+    {
+      "epoch": 5.547829542068143,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0003561776100129135,
+      "loss": 0.4964,
+      "step": 111700
+    },
+    {
+      "epoch": 5.548326214363763,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003561378762292639,
+      "loss": 0.5088,
+      "step": 111710
+    },
+    {
+      "epoch": 5.548822886659382,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003560981424456144,
+      "loss": 0.4856,
+      "step": 111720
+    },
+    {
+      "epoch": 5.549319558955002,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00035605840866196486,
+      "loss": 0.5071,
+      "step": 111730
+    },
+    {
+      "epoch": 5.549816231250621,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003560186748783153,
+      "loss": 0.5245,
+      "step": 111740
+    },
+    {
+      "epoch": 5.55031290354624,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00035597894109466575,
+      "loss": 0.4938,
+      "step": 111750
+    },
+    {
+      "epoch": 5.55080957584186,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0003559392073110162,
+      "loss": 0.4827,
+      "step": 111760
+    },
+    {
+      "epoch": 5.551306248137479,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0003558994735273667,
+      "loss": 0.5306,
+      "step": 111770
+    },
+    {
+      "epoch": 5.551802920433098,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003558597397437171,
+      "loss": 0.5068,
+      "step": 111780
+    },
+    {
+      "epoch": 5.552299592728717,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00035582000596006753,
+      "loss": 0.5106,
+      "step": 111790
+    },
+    {
+      "epoch": 5.552796265024337,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00035578027217641805,
+      "loss": 0.5352,
+      "step": 111800
+    },
+    {
+      "epoch": 5.553292937319957,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00035574053839276847,
+      "loss": 0.4932,
+      "step": 111810
+    },
+    {
+      "epoch": 5.553789609615576,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0003557008046091189,
+      "loss": 0.5136,
+      "step": 111820
+    },
+    {
+      "epoch": 5.554286281911195,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00035566107082546936,
+      "loss": 0.5193,
+      "step": 111830
+    },
+    {
+      "epoch": 5.554782954206814,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00035562133704181983,
+      "loss": 0.4814,
+      "step": 111840
+    },
+    {
+      "epoch": 5.555279626502434,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003555816032581703,
+      "loss": 0.5252,
+      "step": 111850
+    },
+    {
+      "epoch": 5.555776298798053,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003555418694745207,
+      "loss": 0.5123,
+      "step": 111860
+    },
+    {
+      "epoch": 5.556272971093672,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003555021356908712,
+      "loss": 0.4962,
+      "step": 111870
+    },
+    {
+      "epoch": 5.556769643389291,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00035546240190722166,
+      "loss": 0.4998,
+      "step": 111880
+    },
+    {
+      "epoch": 5.557266315684911,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0003554226681235721,
+      "loss": 0.5142,
+      "step": 111890
+    },
+    {
+      "epoch": 5.557762987980531,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00035538293433992255,
+      "loss": 0.4966,
+      "step": 111900
+    },
+    {
+      "epoch": 5.55825966027615,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00035534320055627297,
+      "loss": 0.5049,
+      "step": 111910
+    },
+    {
+      "epoch": 5.558756332571769,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00035530346677262344,
+      "loss": 0.5032,
+      "step": 111920
+    },
+    {
+      "epoch": 5.559253004867388,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003552637329889739,
+      "loss": 0.5364,
+      "step": 111930
+    },
+    {
+      "epoch": 5.5597496771630075,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003552239992053243,
+      "loss": 0.5056,
+      "step": 111940
+    },
+    {
+      "epoch": 5.560246349458627,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0003551842654216748,
+      "loss": 0.4908,
+      "step": 111950
+    },
+    {
+      "epoch": 5.560743021754247,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00035514453163802527,
+      "loss": 0.4613,
+      "step": 111960
+    },
+    {
+      "epoch": 5.561239694049866,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0003551047978543757,
+      "loss": 0.5262,
+      "step": 111970
+    },
+    {
+      "epoch": 5.561736366345485,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00035506506407072616,
+      "loss": 0.4986,
+      "step": 111980
+    },
+    {
+      "epoch": 5.5622330386411045,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00035502533028707663,
+      "loss": 0.5007,
+      "step": 111990
+    },
+    {
+      "epoch": 5.562729710936724,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00035498559650342705,
+      "loss": 0.5097,
+      "step": 112000
+    },
+    {
+      "epoch": 5.563226383232343,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0003549458627197775,
+      "loss": 0.5015,
+      "step": 112010
+    },
+    {
+      "epoch": 5.563723055527962,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00035490612893612794,
+      "loss": 0.5014,
+      "step": 112020
+    },
+    {
+      "epoch": 5.564219727823582,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0003548663951524784,
+      "loss": 0.4997,
+      "step": 112030
+    },
+    {
+      "epoch": 5.5647164001192015,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0003548266613688289,
+      "loss": 0.5313,
+      "step": 112040
+    },
+    {
+      "epoch": 5.565213072414821,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0003547869275851793,
+      "loss": 0.5256,
+      "step": 112050
+    },
+    {
+      "epoch": 5.56570974471044,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00035474719380152977,
+      "loss": 0.4737,
+      "step": 112060
+    },
+    {
+      "epoch": 5.566206417006059,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00035470746001788024,
+      "loss": 0.4921,
+      "step": 112070
+    },
+    {
+      "epoch": 5.5667030893016785,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.00035466772623423065,
+      "loss": 0.5041,
+      "step": 112080
+    },
+    {
+      "epoch": 5.567199761597298,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0003546279924505811,
+      "loss": 0.5132,
+      "step": 112090
+    },
+    {
+      "epoch": 5.567696433892918,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0003545882586669316,
+      "loss": 0.511,
+      "step": 112100
+    },
+    {
+      "epoch": 5.568193106188537,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.000354548524883282,
+      "loss": 0.5041,
+      "step": 112110
+    },
+    {
+      "epoch": 5.568689778484156,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003545087910996325,
+      "loss": 0.4775,
+      "step": 112120
+    },
+    {
+      "epoch": 5.5691864507797755,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003544690573159829,
+      "loss": 0.5107,
+      "step": 112130
+    },
+    {
+      "epoch": 5.569683123075395,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003544293235323334,
+      "loss": 0.4981,
+      "step": 112140
+    },
+    {
+      "epoch": 5.570179795371014,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00035438958974868385,
+      "loss": 0.5178,
+      "step": 112150
+    },
+    {
+      "epoch": 5.570676467666633,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00035434985596503426,
+      "loss": 0.515,
+      "step": 112160
+    },
+    {
+      "epoch": 5.571173139962253,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00035431012218138473,
+      "loss": 0.5216,
+      "step": 112170
+    },
+    {
+      "epoch": 5.5716698122578725,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0003542703883977352,
+      "loss": 0.4892,
+      "step": 112180
+    },
+    {
+      "epoch": 5.572166484553492,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0003542306546140856,
+      "loss": 0.4874,
+      "step": 112190
+    },
+    {
+      "epoch": 5.572663156849111,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003541909208304361,
+      "loss": 0.4912,
+      "step": 112200
+    },
+    {
+      "epoch": 5.57315982914473,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003541511870467865,
+      "loss": 0.5085,
+      "step": 112210
+    },
+    {
+      "epoch": 5.573656501440349,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00035411145326313704,
+      "loss": 0.5043,
+      "step": 112220
+    },
+    {
+      "epoch": 5.574153173735969,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00035407171947948745,
+      "loss": 0.5144,
+      "step": 112230
+    },
+    {
+      "epoch": 5.574649846031589,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00035403198569583787,
+      "loss": 0.5169,
+      "step": 112240
+    },
+    {
+      "epoch": 5.575146518327208,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00035399225191218834,
+      "loss": 0.4937,
+      "step": 112250
+    },
+    {
+      "epoch": 5.575643190622827,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0003539525181285388,
+      "loss": 0.5045,
+      "step": 112260
+    },
+    {
+      "epoch": 5.576139862918446,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003539127843448893,
+      "loss": 0.5102,
+      "step": 112270
+    },
+    {
+      "epoch": 5.576636535214066,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003538730505612397,
+      "loss": 0.5018,
+      "step": 112280
+    },
+    {
+      "epoch": 5.577133207509685,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0003538333167775902,
+      "loss": 0.4976,
+      "step": 112290
+    },
+    {
+      "epoch": 5.577629879805304,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00035379358299394065,
+      "loss": 0.5036,
+      "step": 112300
+    },
+    {
+      "epoch": 5.578126552100924,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00035375384921029106,
+      "loss": 0.4883,
+      "step": 112310
+    },
+    {
+      "epoch": 5.5786232243965435,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0003537141154266415,
+      "loss": 0.4795,
+      "step": 112320
+    },
+    {
+      "epoch": 5.579119896692163,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.000353674381642992,
+      "loss": 0.5187,
+      "step": 112330
+    },
+    {
+      "epoch": 5.579616568987782,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003536346478593424,
+      "loss": 0.4851,
+      "step": 112340
+    },
+    {
+      "epoch": 5.580113241283401,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0003535949140756929,
+      "loss": 0.5324,
+      "step": 112350
+    },
+    {
+      "epoch": 5.58060991357902,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0003535551802920433,
+      "loss": 0.5057,
+      "step": 112360
+    },
+    {
+      "epoch": 5.58110658587464,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0003535154465083938,
+      "loss": 0.5042,
+      "step": 112370
+    },
+    {
+      "epoch": 5.58160325817026,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00035347571272474425,
+      "loss": 0.4992,
+      "step": 112380
+    },
+    {
+      "epoch": 5.582099930465879,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00035343597894109467,
+      "loss": 0.4987,
+      "step": 112390
+    },
+    {
+      "epoch": 5.582596602761498,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00035339624515744514,
+      "loss": 0.4869,
+      "step": 112400
+    },
+    {
+      "epoch": 5.583093275057117,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003533565113737956,
+      "loss": 0.5111,
+      "step": 112410
+    },
+    {
+      "epoch": 5.583589947352737,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00035331677759014603,
+      "loss": 0.4857,
+      "step": 112420
+    },
+    {
+      "epoch": 5.584086619648356,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0003532770438064965,
+      "loss": 0.5344,
+      "step": 112430
+    },
+    {
+      "epoch": 5.584583291943975,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0003532373100228469,
+      "loss": 0.5189,
+      "step": 112440
+    },
+    {
+      "epoch": 5.585079964239595,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003531975762391974,
+      "loss": 0.4859,
+      "step": 112450
+    },
+    {
+      "epoch": 5.585576636535214,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00035315784245554786,
+      "loss": 0.511,
+      "step": 112460
+    },
+    {
+      "epoch": 5.586073308830834,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003531181086718983,
+      "loss": 0.4974,
+      "step": 112470
+    },
+    {
+      "epoch": 5.586569981126453,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00035307837488824875,
+      "loss": 0.4927,
+      "step": 112480
+    },
+    {
+      "epoch": 5.587066653422072,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0003530386411045992,
+      "loss": 0.4816,
+      "step": 112490
+    },
+    {
+      "epoch": 5.587563325717691,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00035299890732094964,
+      "loss": 0.466,
+      "step": 112500
+    },
+    {
+      "epoch": 5.5880599980133105,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003529591735373001,
+      "loss": 0.5036,
+      "step": 112510
+    },
+    {
+      "epoch": 5.588556670308931,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0003529194397536506,
+      "loss": 0.5224,
+      "step": 112520
+    },
+    {
+      "epoch": 5.58905334260455,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.000352879705970001,
+      "loss": 0.4723,
+      "step": 112530
+    },
+    {
+      "epoch": 5.589550014900169,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.00035283997218635147,
+      "loss": 0.5103,
+      "step": 112540
+    },
+    {
+      "epoch": 5.590046687195788,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0003528002384027019,
+      "loss": 0.502,
+      "step": 112550
+    },
+    {
+      "epoch": 5.590543359491408,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00035276050461905236,
+      "loss": 0.5116,
+      "step": 112560
+    },
+    {
+      "epoch": 5.591040031787027,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00035272077083540283,
+      "loss": 0.5403,
+      "step": 112570
+    },
+    {
+      "epoch": 5.591536704082646,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00035268103705175325,
+      "loss": 0.5146,
+      "step": 112580
+    },
+    {
+      "epoch": 5.592033376378265,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003526413032681037,
+      "loss": 0.5076,
+      "step": 112590
+    },
+    {
+      "epoch": 5.5925300486738845,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003526015694844542,
+      "loss": 0.5097,
+      "step": 112600
+    },
+    {
+      "epoch": 5.593026720969505,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003525618357008046,
+      "loss": 0.4999,
+      "step": 112610
+    },
+    {
+      "epoch": 5.593523393265124,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003525221019171551,
+      "loss": 0.4954,
+      "step": 112620
+    },
+    {
+      "epoch": 5.594020065560743,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00035248236813350555,
+      "loss": 0.5037,
+      "step": 112630
+    },
+    {
+      "epoch": 5.594516737856362,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00035244263434985597,
+      "loss": 0.5443,
+      "step": 112640
+    },
+    {
+      "epoch": 5.5950134101519815,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00035240290056620644,
+      "loss": 0.4872,
+      "step": 112650
+    },
+    {
+      "epoch": 5.595510082447601,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00035236316678255686,
+      "loss": 0.5063,
+      "step": 112660
+    },
+    {
+      "epoch": 5.59600675474322,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0003523234329989074,
+      "loss": 0.518,
+      "step": 112670
+    },
+    {
+      "epoch": 5.59650342703884,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003522836992152578,
+      "loss": 0.5125,
+      "step": 112680
+    },
+    {
+      "epoch": 5.597000099334459,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0003522439654316082,
+      "loss": 0.5128,
+      "step": 112690
+    },
+    {
+      "epoch": 5.5974967716300785,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0003522042316479587,
+      "loss": 0.5089,
+      "step": 112700
+    },
+    {
+      "epoch": 5.597993443925698,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00035216449786430916,
+      "loss": 0.4967,
+      "step": 112710
+    },
+    {
+      "epoch": 5.598490116221317,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00035212476408065963,
+      "loss": 0.5239,
+      "step": 112720
+    },
+    {
+      "epoch": 5.598986788516936,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00035208503029701005,
+      "loss": 0.512,
+      "step": 112730
+    },
+    {
+      "epoch": 5.599483460812555,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00035204529651336046,
+      "loss": 0.4772,
+      "step": 112740
+    },
+    {
+      "epoch": 5.5999801331081756,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.000352005562729711,
+      "loss": 0.5299,
+      "step": 112750
+    },
+    {
+      "epoch": 5.600476805403795,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003519658289460614,
+      "loss": 0.4781,
+      "step": 112760
+    },
+    {
+      "epoch": 5.600973477699414,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0003519260951624118,
+      "loss": 0.5015,
+      "step": 112770
+    },
+    {
+      "epoch": 5.601470149995033,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0003518863613787623,
+      "loss": 0.4884,
+      "step": 112780
+    },
+    {
+      "epoch": 5.6019668222906525,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00035184662759511277,
+      "loss": 0.4633,
+      "step": 112790
+    },
+    {
+      "epoch": 5.602463494586272,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00035180689381146324,
+      "loss": 0.4953,
+      "step": 112800
+    },
+    {
+      "epoch": 5.602960166881891,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00035176716002781365,
+      "loss": 0.5085,
+      "step": 112810
+    },
+    {
+      "epoch": 5.603456839177511,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0003517274262441641,
+      "loss": 0.4908,
+      "step": 112820
+    },
+    {
+      "epoch": 5.60395351147313,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0003516876924605146,
+      "loss": 0.4784,
+      "step": 112830
+    },
+    {
+      "epoch": 5.6044501837687495,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.000351647958676865,
+      "loss": 0.5038,
+      "step": 112840
+    },
+    {
+      "epoch": 5.604946856064369,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00035160822489321543,
+      "loss": 0.5014,
+      "step": 112850
+    },
+    {
+      "epoch": 5.605443528359988,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00035156849110956596,
+      "loss": 0.5004,
+      "step": 112860
+    },
+    {
+      "epoch": 5.605940200655607,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0003515287573259164,
+      "loss": 0.4771,
+      "step": 112870
+    },
+    {
+      "epoch": 5.606436872951226,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00035148902354226685,
+      "loss": 0.5097,
+      "step": 112880
+    },
+    {
+      "epoch": 5.6069335452468465,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00035144928975861726,
+      "loss": 0.4958,
+      "step": 112890
+    },
+    {
+      "epoch": 5.607430217542466,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00035140955597496773,
+      "loss": 0.4939,
+      "step": 112900
+    },
+    {
+      "epoch": 5.607926889838085,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003513698221913182,
+      "loss": 0.5129,
+      "step": 112910
+    },
+    {
+      "epoch": 5.608423562133704,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003513300884076686,
+      "loss": 0.4886,
+      "step": 112920
+    },
+    {
+      "epoch": 5.608920234429323,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0003512903546240191,
+      "loss": 0.5186,
+      "step": 112930
+    },
+    {
+      "epoch": 5.609416906724943,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00035125062084036957,
+      "loss": 0.4782,
+      "step": 112940
+    },
+    {
+      "epoch": 5.609913579020562,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00035121088705672,
+      "loss": 0.518,
+      "step": 112950
+    },
+    {
+      "epoch": 5.610410251316182,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00035117115327307045,
+      "loss": 0.505,
+      "step": 112960
+    },
+    {
+      "epoch": 5.610906923611801,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003511314194894209,
+      "loss": 0.51,
+      "step": 112970
+    },
+    {
+      "epoch": 5.61140359590742,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00035109168570577134,
+      "loss": 0.5075,
+      "step": 112980
+    },
+    {
+      "epoch": 5.61190026820304,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003510519519221218,
+      "loss": 0.5147,
+      "step": 112990
+    },
+    {
+      "epoch": 5.612396940498659,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00035101221813847223,
+      "loss": 0.4915,
+      "step": 113000
+    },
+    {
+      "epoch": 5.612893612794278,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003509724843548227,
+      "loss": 0.5033,
+      "step": 113010
+    },
+    {
+      "epoch": 5.613390285089897,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003509327505711732,
+      "loss": 0.5055,
+      "step": 113020
+    },
+    {
+      "epoch": 5.6138869573855175,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0003508930167875236,
+      "loss": 0.496,
+      "step": 113030
+    },
+    {
+      "epoch": 5.614383629681137,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00035085328300387406,
+      "loss": 0.5035,
+      "step": 113040
+    },
+    {
+      "epoch": 5.614880301976756,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00035081354922022453,
+      "loss": 0.4637,
+      "step": 113050
+    },
+    {
+      "epoch": 5.615376974272375,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00035077381543657495,
+      "loss": 0.5175,
+      "step": 113060
+    },
+    {
+      "epoch": 5.615873646567994,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003507340816529254,
+      "loss": 0.5448,
+      "step": 113070
+    },
+    {
+      "epoch": 5.616370318863614,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00035069434786927584,
+      "loss": 0.5084,
+      "step": 113080
+    },
+    {
+      "epoch": 5.616866991159233,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0003506546140856263,
+      "loss": 0.4975,
+      "step": 113090
+    },
+    {
+      "epoch": 5.617363663454853,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0003506148803019768,
+      "loss": 0.5024,
+      "step": 113100
+    },
+    {
+      "epoch": 5.617860335750472,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0003505751465183272,
+      "loss": 0.5177,
+      "step": 113110
+    },
+    {
+      "epoch": 5.618357008046091,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00035053541273467767,
+      "loss": 0.4979,
+      "step": 113120
+    },
+    {
+      "epoch": 5.618853680341711,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00035049567895102814,
+      "loss": 0.5113,
+      "step": 113130
+    },
+    {
+      "epoch": 5.61935035263733,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00035045594516737856,
+      "loss": 0.5245,
+      "step": 113140
+    },
+    {
+      "epoch": 5.619847024932949,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00035041621138372903,
+      "loss": 0.535,
+      "step": 113150
+    },
+    {
+      "epoch": 5.620343697228568,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0003503764776000795,
+      "loss": 0.5191,
+      "step": 113160
+    },
+    {
+      "epoch": 5.620840369524188,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00035033674381642997,
+      "loss": 0.501,
+      "step": 113170
+    },
+    {
+      "epoch": 5.621337041819808,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0003502970100327804,
+      "loss": 0.509,
+      "step": 113180
+    },
+    {
+      "epoch": 5.621833714115427,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0003502572762491308,
+      "loss": 0.507,
+      "step": 113190
+    },
+    {
+      "epoch": 5.622330386411046,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00035021754246548133,
+      "loss": 0.5035,
+      "step": 113200
+    },
+    {
+      "epoch": 5.622827058706665,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00035017780868183175,
+      "loss": 0.5092,
+      "step": 113210
+    },
+    {
+      "epoch": 5.6233237310022846,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00035013807489818217,
+      "loss": 0.4871,
+      "step": 113220
+    },
+    {
+      "epoch": 5.623820403297904,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00035009834111453264,
+      "loss": 0.4793,
+      "step": 113230
+    },
+    {
+      "epoch": 5.624317075593523,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003500586073308831,
+      "loss": 0.4974,
+      "step": 113240
+    },
+    {
+      "epoch": 5.624813747889143,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003500188735472336,
+      "loss": 0.5088,
+      "step": 113250
+    },
+    {
+      "epoch": 5.625310420184762,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.000349979139763584,
+      "loss": 0.4957,
+      "step": 113260
+    },
+    {
+      "epoch": 5.625807092480382,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00034993940597993447,
+      "loss": 0.4833,
+      "step": 113270
+    },
+    {
+      "epoch": 5.626303764776001,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00034989967219628494,
+      "loss": 0.4983,
+      "step": 113280
+    },
+    {
+      "epoch": 5.62680043707162,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00034985993841263536,
+      "loss": 0.4798,
+      "step": 113290
+    },
+    {
+      "epoch": 5.627297109367239,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0003498202046289858,
+      "loss": 0.5027,
+      "step": 113300
+    },
+    {
+      "epoch": 5.6277937816628585,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00034978047084533625,
+      "loss": 0.4887,
+      "step": 113310
+    },
+    {
+      "epoch": 5.628290453958478,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003497407370616867,
+      "loss": 0.4926,
+      "step": 113320
+    },
+    {
+      "epoch": 5.628787126254098,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003497010032780372,
+      "loss": 0.5032,
+      "step": 113330
+    },
+    {
+      "epoch": 5.629283798549717,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003496612694943876,
+      "loss": 0.4918,
+      "step": 113340
+    },
+    {
+      "epoch": 5.629780470845336,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0003496215357107381,
+      "loss": 0.5331,
+      "step": 113350
+    },
+    {
+      "epoch": 5.6302771431409555,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00034958180192708855,
+      "loss": 0.5183,
+      "step": 113360
+    },
+    {
+      "epoch": 5.630773815436575,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00034954206814343897,
+      "loss": 0.5034,
+      "step": 113370
+    },
+    {
+      "epoch": 5.631270487732194,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003495023343597894,
+      "loss": 0.503,
+      "step": 113380
+    },
+    {
+      "epoch": 5.631767160027813,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0003494626005761399,
+      "loss": 0.5195,
+      "step": 113390
+    },
+    {
+      "epoch": 5.632263832323433,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0003494228667924903,
+      "loss": 0.4737,
+      "step": 113400
+    },
+    {
+      "epoch": 5.6327605046190525,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003493831330088408,
+      "loss": 0.5063,
+      "step": 113410
+    },
+    {
+      "epoch": 5.633257176914672,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003493433992251912,
+      "loss": 0.5096,
+      "step": 113420
+    },
+    {
+      "epoch": 5.633753849210291,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003493036654415417,
+      "loss": 0.4873,
+      "step": 113430
+    },
+    {
+      "epoch": 5.63425052150591,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00034926393165789216,
+      "loss": 0.4895,
+      "step": 113440
+    },
+    {
+      "epoch": 5.634747193801529,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003492241978742426,
+      "loss": 0.5058,
+      "step": 113450
+    },
+    {
+      "epoch": 5.635243866097149,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.00034918446409059305,
+      "loss": 0.4999,
+      "step": 113460
+    },
+    {
+      "epoch": 5.635740538392769,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0003491447303069435,
+      "loss": 0.5198,
+      "step": 113470
+    },
+    {
+      "epoch": 5.636237210688388,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00034910499652329393,
+      "loss": 0.5367,
+      "step": 113480
+    },
+    {
+      "epoch": 5.636733882984007,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003490652627396444,
+      "loss": 0.5206,
+      "step": 113490
+    },
+    {
+      "epoch": 5.6372305552796265,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0003490255289559949,
+      "loss": 0.4997,
+      "step": 113500
+    },
+    {
+      "epoch": 5.637727227575246,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0003489857951723453,
+      "loss": 0.4914,
+      "step": 113510
+    },
+    {
+      "epoch": 5.638223899870865,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00034894606138869577,
+      "loss": 0.5156,
+      "step": 113520
+    },
+    {
+      "epoch": 5.638720572166484,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0003489063276050462,
+      "loss": 0.5263,
+      "step": 113530
+    },
+    {
+      "epoch": 5.639217244462104,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003488665938213967,
+      "loss": 0.5106,
+      "step": 113540
+    },
+    {
+      "epoch": 5.6397139167577235,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0003488268600377471,
+      "loss": 0.4905,
+      "step": 113550
+    },
+    {
+      "epoch": 5.640210589053343,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00034878712625409754,
+      "loss": 0.4771,
+      "step": 113560
+    },
+    {
+      "epoch": 5.640707261348962,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.000348747392470448,
+      "loss": 0.4879,
+      "step": 113570
+    },
+    {
+      "epoch": 5.641203933644581,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003487076586867985,
+      "loss": 0.5118,
+      "step": 113580
+    },
+    {
+      "epoch": 5.6417006059402,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003486679249031489,
+      "loss": 0.5268,
+      "step": 113590
+    },
+    {
+      "epoch": 5.64219727823582,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0003486281911194994,
+      "loss": 0.5118,
+      "step": 113600
+    },
+    {
+      "epoch": 5.64269395053144,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0003485884573358498,
+      "loss": 0.501,
+      "step": 113610
+    },
+    {
+      "epoch": 5.643190622827059,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003485487235522003,
+      "loss": 0.528,
+      "step": 113620
+    },
+    {
+      "epoch": 5.643687295122678,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00034850898976855073,
+      "loss": 0.5168,
+      "step": 113630
+    },
+    {
+      "epoch": 5.644183967418297,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00034846925598490115,
+      "loss": 0.5091,
+      "step": 113640
+    },
+    {
+      "epoch": 5.644680639713917,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0003484295222012516,
+      "loss": 0.5126,
+      "step": 113650
+    },
+    {
+      "epoch": 5.645177312009536,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0003483897884176021,
+      "loss": 0.5085,
+      "step": 113660
+    },
+    {
+      "epoch": 5.645673984305155,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0003483500546339525,
+      "loss": 0.5356,
+      "step": 113670
+    },
+    {
+      "epoch": 5.646170656600775,
+      "grad_norm": 0.125,
+      "learning_rate": 0.000348310320850303,
+      "loss": 0.5088,
+      "step": 113680
+    },
+    {
+      "epoch": 5.6466673288963944,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00034827058706665345,
+      "loss": 0.5075,
+      "step": 113690
+    },
+    {
+      "epoch": 5.647164001192014,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0003482308532830039,
+      "loss": 0.5144,
+      "step": 113700
+    },
+    {
+      "epoch": 5.647660673487633,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00034819111949935434,
+      "loss": 0.4816,
+      "step": 113710
+    },
+    {
+      "epoch": 5.648157345783252,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00034815138571570476,
+      "loss": 0.523,
+      "step": 113720
+    },
+    {
+      "epoch": 5.648654018078871,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0003481116519320553,
+      "loss": 0.5075,
+      "step": 113730
+    },
+    {
+      "epoch": 5.649150690374491,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0003480719181484057,
+      "loss": 0.5195,
+      "step": 113740
+    },
+    {
+      "epoch": 5.649647362670111,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003480321843647561,
+      "loss": 0.5362,
+      "step": 113750
+    },
+    {
+      "epoch": 5.65014403496573,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0003479924505811066,
+      "loss": 0.515,
+      "step": 113760
+    },
+    {
+      "epoch": 5.650640707261349,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00034795271679745706,
+      "loss": 0.491,
+      "step": 113770
+    },
+    {
+      "epoch": 5.651137379556968,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00034791298301380753,
+      "loss": 0.4865,
+      "step": 113780
+    },
+    {
+      "epoch": 5.651634051852588,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00034787324923015795,
+      "loss": 0.5131,
+      "step": 113790
+    },
+    {
+      "epoch": 5.652130724148207,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003478335154465084,
+      "loss": 0.5108,
+      "step": 113800
+    },
+    {
+      "epoch": 5.652627396443826,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0003477937816628589,
+      "loss": 0.5055,
+      "step": 113810
+    },
+    {
+      "epoch": 5.653124068739446,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0003477540478792093,
+      "loss": 0.4903,
+      "step": 113820
+    },
+    {
+      "epoch": 5.653620741035065,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0003477143140955597,
+      "loss": 0.5069,
+      "step": 113830
+    },
+    {
+      "epoch": 5.654117413330685,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003476745803119102,
+      "loss": 0.4916,
+      "step": 113840
+    },
+    {
+      "epoch": 5.654614085626304,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00034763484652826067,
+      "loss": 0.5005,
+      "step": 113850
+    },
+    {
+      "epoch": 5.655110757921923,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00034759511274461114,
+      "loss": 0.4995,
+      "step": 113860
+    },
+    {
+      "epoch": 5.655607430217542,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00034755537896096156,
+      "loss": 0.5523,
+      "step": 113870
+    },
+    {
+      "epoch": 5.6561041025131615,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00034751564517731203,
+      "loss": 0.522,
+      "step": 113880
+    },
+    {
+      "epoch": 5.656600774808782,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003474759113936625,
+      "loss": 0.4748,
+      "step": 113890
+    },
+    {
+      "epoch": 5.657097447104401,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003474361776100129,
+      "loss": 0.5299,
+      "step": 113900
+    },
+    {
+      "epoch": 5.65759411940002,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00034739644382636334,
+      "loss": 0.5318,
+      "step": 113910
+    },
+    {
+      "epoch": 5.658090791695639,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00034735671004271386,
+      "loss": 0.524,
+      "step": 113920
+    },
+    {
+      "epoch": 5.658587463991259,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0003473169762590643,
+      "loss": 0.494,
+      "step": 113930
+    },
+    {
+      "epoch": 5.659084136286878,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00034727724247541475,
+      "loss": 0.4914,
+      "step": 113940
+    },
+    {
+      "epoch": 5.659580808582497,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00034723750869176517,
+      "loss": 0.4848,
+      "step": 113950
+    },
+    {
+      "epoch": 5.660077480878116,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00034719777490811564,
+      "loss": 0.4869,
+      "step": 113960
+    },
+    {
+      "epoch": 5.660574153173736,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003471580411244661,
+      "loss": 0.4768,
+      "step": 113970
+    },
+    {
+      "epoch": 5.661070825469356,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0003471183073408165,
+      "loss": 0.491,
+      "step": 113980
+    },
+    {
+      "epoch": 5.661567497764975,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.000347078573557167,
+      "loss": 0.4945,
+      "step": 113990
+    },
+    {
+      "epoch": 5.662064170060594,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00034703883977351747,
+      "loss": 0.5128,
+      "step": 114000
+    },
+    {
+      "epoch": 5.662560842356213,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003469991059898679,
+      "loss": 0.4963,
+      "step": 114010
+    },
+    {
+      "epoch": 5.6630575146518325,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00034695937220621836,
+      "loss": 0.5241,
+      "step": 114020
+    },
+    {
+      "epoch": 5.663554186947452,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00034691963842256883,
+      "loss": 0.5257,
+      "step": 114030
+    },
+    {
+      "epoch": 5.664050859243071,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00034687990463891925,
+      "loss": 0.5243,
+      "step": 114040
+    },
+    {
+      "epoch": 5.664547531538691,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0003468401708552697,
+      "loss": 0.4992,
+      "step": 114050
+    },
+    {
+      "epoch": 5.66504420383431,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00034680043707162013,
+      "loss": 0.5164,
+      "step": 114060
+    },
+    {
+      "epoch": 5.6655408761299295,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00034676070328797066,
+      "loss": 0.5001,
+      "step": 114070
+    },
+    {
+      "epoch": 5.666037548425549,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0003467209695043211,
+      "loss": 0.5042,
+      "step": 114080
+    },
+    {
+      "epoch": 5.666534220721168,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0003466812357206715,
+      "loss": 0.5113,
+      "step": 114090
+    },
+    {
+      "epoch": 5.667030893016787,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00034664150193702197,
+      "loss": 0.5155,
+      "step": 114100
+    },
+    {
+      "epoch": 5.667527565312406,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00034660176815337244,
+      "loss": 0.5023,
+      "step": 114110
+    },
+    {
+      "epoch": 5.6680242376080265,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00034656203436972285,
+      "loss": 0.4907,
+      "step": 114120
+    },
+    {
+      "epoch": 5.668520909903646,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003465223005860733,
+      "loss": 0.481,
+      "step": 114130
+    },
+    {
+      "epoch": 5.669017582199265,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00034648256680242374,
+      "loss": 0.5019,
+      "step": 114140
+    },
+    {
+      "epoch": 5.669514254494884,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00034644283301877427,
+      "loss": 0.4908,
+      "step": 114150
+    },
+    {
+      "epoch": 5.6700109267905034,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0003464030992351247,
+      "loss": 0.5105,
+      "step": 114160
+    },
+    {
+      "epoch": 5.670507599086123,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0003463633654514751,
+      "loss": 0.5104,
+      "step": 114170
+    },
+    {
+      "epoch": 5.671004271381742,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003463236316678256,
+      "loss": 0.5042,
+      "step": 114180
+    },
+    {
+      "epoch": 5.671500943677362,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00034628389788417605,
+      "loss": 0.5159,
+      "step": 114190
+    },
+    {
+      "epoch": 5.671997615972981,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00034624416410052646,
+      "loss": 0.5134,
+      "step": 114200
+    },
+    {
+      "epoch": 5.6724942882686005,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00034620443031687693,
+      "loss": 0.5087,
+      "step": 114210
+    },
+    {
+      "epoch": 5.67299096056422,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003461646965332274,
+      "loss": 0.5168,
+      "step": 114220
+    },
+    {
+      "epoch": 5.673487632859839,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003461249627495779,
+      "loss": 0.5074,
+      "step": 114230
+    },
+    {
+      "epoch": 5.673984305155458,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0003460852289659283,
+      "loss": 0.5065,
+      "step": 114240
+    },
+    {
+      "epoch": 5.674480977451077,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003460454951822787,
+      "loss": 0.5326,
+      "step": 114250
+    },
+    {
+      "epoch": 5.6749776497466975,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00034600576139862924,
+      "loss": 0.4766,
+      "step": 114260
+    },
+    {
+      "epoch": 5.675474322042317,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00034596602761497965,
+      "loss": 0.5074,
+      "step": 114270
+    },
+    {
+      "epoch": 5.675970994337936,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00034592629383133007,
+      "loss": 0.497,
+      "step": 114280
+    },
+    {
+      "epoch": 5.676467666633555,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00034588656004768054,
+      "loss": 0.5005,
+      "step": 114290
+    },
+    {
+      "epoch": 5.676964338929174,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.000345846826264031,
+      "loss": 0.4988,
+      "step": 114300
+    },
+    {
+      "epoch": 5.677461011224794,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003458070924803815,
+      "loss": 0.5226,
+      "step": 114310
+    },
+    {
+      "epoch": 5.677957683520413,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0003457673586967319,
+      "loss": 0.5349,
+      "step": 114320
+    },
+    {
+      "epoch": 5.678454355816033,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0003457276249130824,
+      "loss": 0.4942,
+      "step": 114330
+    },
+    {
+      "epoch": 5.678951028111652,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00034568789112943284,
+      "loss": 0.4801,
+      "step": 114340
+    },
+    {
+      "epoch": 5.679447700407271,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00034564815734578326,
+      "loss": 0.5183,
+      "step": 114350
+    },
+    {
+      "epoch": 5.679944372702891,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0003456084235621337,
+      "loss": 0.4792,
+      "step": 114360
+    },
+    {
+      "epoch": 5.68044104499851,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003455686897784842,
+      "loss": 0.5151,
+      "step": 114370
+    },
+    {
+      "epoch": 5.680937717294129,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0003455289559948346,
+      "loss": 0.5321,
+      "step": 114380
+    },
+    {
+      "epoch": 5.681434389589748,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0003454892222111851,
+      "loss": 0.4892,
+      "step": 114390
+    },
+    {
+      "epoch": 5.6819310618853685,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0003454494884275355,
+      "loss": 0.4954,
+      "step": 114400
+    },
+    {
+      "epoch": 5.682427734180988,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.000345409754643886,
+      "loss": 0.4942,
+      "step": 114410
+    },
+    {
+      "epoch": 5.682924406476607,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00034537002086023645,
+      "loss": 0.5226,
+      "step": 114420
+    },
+    {
+      "epoch": 5.683421078772226,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00034533028707658687,
+      "loss": 0.5203,
+      "step": 114430
+    },
+    {
+      "epoch": 5.683917751067845,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00034529055329293734,
+      "loss": 0.5137,
+      "step": 114440
+    },
+    {
+      "epoch": 5.684414423363465,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003452508195092878,
+      "loss": 0.5044,
+      "step": 114450
+    },
+    {
+      "epoch": 5.684911095659084,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00034521108572563823,
+      "loss": 0.5375,
+      "step": 114460
+    },
+    {
+      "epoch": 5.685407767954704,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0003451713519419887,
+      "loss": 0.4996,
+      "step": 114470
+    },
+    {
+      "epoch": 5.685904440250323,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0003451316181583391,
+      "loss": 0.4906,
+      "step": 114480
+    },
+    {
+      "epoch": 5.686401112545942,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003450918843746896,
+      "loss": 0.5102,
+      "step": 114490
+    },
+    {
+      "epoch": 5.686897784841562,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00034505215059104006,
+      "loss": 0.5323,
+      "step": 114500
+    },
+    {
+      "epoch": 5.687394457137181,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0003450124168073905,
+      "loss": 0.5173,
+      "step": 114510
+    },
+    {
+      "epoch": 5.6878911294328,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00034497268302374095,
+      "loss": 0.5176,
+      "step": 114520
+    },
+    {
+      "epoch": 5.688387801728419,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003449329492400914,
+      "loss": 0.5059,
+      "step": 114530
+    },
+    {
+      "epoch": 5.688884474024039,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00034489321545644184,
+      "loss": 0.5014,
+      "step": 114540
+    },
+    {
+      "epoch": 5.689381146319659,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0003448534816727923,
+      "loss": 0.51,
+      "step": 114550
+    },
+    {
+      "epoch": 5.689877818615278,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0003448137478891428,
+      "loss": 0.5091,
+      "step": 114560
+    },
+    {
+      "epoch": 5.690374490910897,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003447740141054932,
+      "loss": 0.5274,
+      "step": 114570
+    },
+    {
+      "epoch": 5.690871163206516,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00034473428032184367,
+      "loss": 0.5088,
+      "step": 114580
+    },
+    {
+      "epoch": 5.6913678355021355,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003446945465381941,
+      "loss": 0.4935,
+      "step": 114590
+    },
+    {
+      "epoch": 5.691864507797755,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003446548127545446,
+      "loss": 0.527,
+      "step": 114600
+    },
+    {
+      "epoch": 5.692361180093375,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00034461507897089503,
+      "loss": 0.5384,
+      "step": 114610
+    },
+    {
+      "epoch": 5.692857852388994,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00034457534518724545,
+      "loss": 0.5439,
+      "step": 114620
+    },
+    {
+      "epoch": 5.693354524684613,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0003445356114035959,
+      "loss": 0.5353,
+      "step": 114630
+    },
+    {
+      "epoch": 5.693851196980233,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0003444958776199464,
+      "loss": 0.4961,
+      "step": 114640
+    },
+    {
+      "epoch": 5.694347869275852,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003444561438362968,
+      "loss": 0.5086,
+      "step": 114650
+    },
+    {
+      "epoch": 5.694844541571471,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0003444164100526473,
+      "loss": 0.5043,
+      "step": 114660
+    },
+    {
+      "epoch": 5.69534121386709,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.00034437667626899775,
+      "loss": 0.4958,
+      "step": 114670
+    },
+    {
+      "epoch": 5.6958378861627095,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0003443369424853482,
+      "loss": 0.4841,
+      "step": 114680
+    },
+    {
+      "epoch": 5.696334558458329,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00034429720870169864,
+      "loss": 0.4919,
+      "step": 114690
+    },
+    {
+      "epoch": 5.696831230753949,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00034425747491804905,
+      "loss": 0.5071,
+      "step": 114700
+    },
+    {
+      "epoch": 5.697327903049568,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0003442177411343995,
+      "loss": 0.4982,
+      "step": 114710
+    },
+    {
+      "epoch": 5.697824575345187,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00034417800735075,
+      "loss": 0.533,
+      "step": 114720
+    },
+    {
+      "epoch": 5.6983212476408065,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0003441382735671004,
+      "loss": 0.4986,
+      "step": 114730
+    },
+    {
+      "epoch": 5.698817919936426,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003440985397834509,
+      "loss": 0.5243,
+      "step": 114740
+    },
+    {
+      "epoch": 5.699314592232045,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00034405880599980136,
+      "loss": 0.503,
+      "step": 114750
+    },
+    {
+      "epoch": 5.699811264527664,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00034401907221615183,
+      "loss": 0.494,
+      "step": 114760
+    },
+    {
+      "epoch": 5.700307936823284,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00034397933843250225,
+      "loss": 0.5262,
+      "step": 114770
+    },
+    {
+      "epoch": 5.7008046091189035,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00034393960464885266,
+      "loss": 0.5391,
+      "step": 114780
+    },
+    {
+      "epoch": 5.701301281414523,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003438998708652032,
+      "loss": 0.5111,
+      "step": 114790
+    },
+    {
+      "epoch": 5.701797953710142,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003438601370815536,
+      "loss": 0.4906,
+      "step": 114800
+    },
+    {
+      "epoch": 5.702294626005761,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003438204032979041,
+      "loss": 0.4968,
+      "step": 114810
+    },
+    {
+      "epoch": 5.70279129830138,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003437806695142545,
+      "loss": 0.4965,
+      "step": 114820
+    },
+    {
+      "epoch": 5.703287970597,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00034374093573060497,
+      "loss": 0.5153,
+      "step": 114830
+    },
+    {
+      "epoch": 5.70378464289262,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00034370120194695544,
+      "loss": 0.5069,
+      "step": 114840
+    },
+    {
+      "epoch": 5.704281315188239,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00034366146816330585,
+      "loss": 0.4992,
+      "step": 114850
+    },
+    {
+      "epoch": 5.704777987483858,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0003436217343796563,
+      "loss": 0.4968,
+      "step": 114860
+    },
+    {
+      "epoch": 5.7052746597794775,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003435820005960068,
+      "loss": 0.482,
+      "step": 114870
+    },
+    {
+      "epoch": 5.705771332075097,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0003435422668123572,
+      "loss": 0.4977,
+      "step": 114880
+    },
+    {
+      "epoch": 5.706268004370716,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003435025330287077,
+      "loss": 0.5002,
+      "step": 114890
+    },
+    {
+      "epoch": 5.706764676666335,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00034346279924505816,
+      "loss": 0.5028,
+      "step": 114900
+    },
+    {
+      "epoch": 5.707261348961955,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003434230654614086,
+      "loss": 0.522,
+      "step": 114910
+    },
+    {
+      "epoch": 5.7077580212575745,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00034338333167775904,
+      "loss": 0.4802,
+      "step": 114920
+    },
+    {
+      "epoch": 5.708254693553194,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00034334359789410946,
+      "loss": 0.4978,
+      "step": 114930
+    },
+    {
+      "epoch": 5.708751365848813,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00034330386411045993,
+      "loss": 0.5344,
+      "step": 114940
+    },
+    {
+      "epoch": 5.709248038144432,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0003432641303268104,
+      "loss": 0.5075,
+      "step": 114950
+    },
+    {
+      "epoch": 5.709744710440051,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003432243965431608,
+      "loss": 0.4632,
+      "step": 114960
+    },
+    {
+      "epoch": 5.710241382735671,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003431846627595113,
+      "loss": 0.495,
+      "step": 114970
+    },
+    {
+      "epoch": 5.710738055031291,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00034314492897586176,
+      "loss": 0.5062,
+      "step": 114980
+    },
+    {
+      "epoch": 5.71123472732691,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003431051951922122,
+      "loss": 0.5016,
+      "step": 114990
+    },
+    {
+      "epoch": 5.711731399622529,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00034306546140856265,
+      "loss": 0.4985,
+      "step": 115000
+    },
+    {
+      "epoch": 5.712228071918148,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00034302572762491307,
+      "loss": 0.5135,
+      "step": 115010
+    },
+    {
+      "epoch": 5.712724744213768,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00034298599384126354,
+      "loss": 0.5065,
+      "step": 115020
+    },
+    {
+      "epoch": 5.713221416509387,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.000342946260057614,
+      "loss": 0.4908,
+      "step": 115030
+    },
+    {
+      "epoch": 5.713718088805006,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00034290652627396443,
+      "loss": 0.5072,
+      "step": 115040
+    },
+    {
+      "epoch": 5.714214761100626,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003428667924903149,
+      "loss": 0.4969,
+      "step": 115050
+    },
+    {
+      "epoch": 5.714711433396245,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00034282705870666537,
+      "loss": 0.4945,
+      "step": 115060
+    },
+    {
+      "epoch": 5.715208105691865,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003427873249230158,
+      "loss": 0.502,
+      "step": 115070
+    },
+    {
+      "epoch": 5.715704777987484,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00034274759113936626,
+      "loss": 0.5,
+      "step": 115080
+    },
+    {
+      "epoch": 5.716201450283103,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00034270785735571673,
+      "loss": 0.49,
+      "step": 115090
+    },
+    {
+      "epoch": 5.716698122578722,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00034266812357206715,
+      "loss": 0.4713,
+      "step": 115100
+    },
+    {
+      "epoch": 5.717194794874342,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0003426283897884176,
+      "loss": 0.4878,
+      "step": 115110
+    },
+    {
+      "epoch": 5.717691467169962,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00034258865600476804,
+      "loss": 0.4958,
+      "step": 115120
+    },
+    {
+      "epoch": 5.718188139465581,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00034254892222111856,
+      "loss": 0.4822,
+      "step": 115130
+    },
+    {
+      "epoch": 5.7186848117612,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.000342509188437469,
+      "loss": 0.5026,
+      "step": 115140
+    },
+    {
+      "epoch": 5.719181484056819,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003424694546538194,
+      "loss": 0.5005,
+      "step": 115150
+    },
+    {
+      "epoch": 5.719678156352439,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00034242972087016987,
+      "loss": 0.5175,
+      "step": 115160
+    },
+    {
+      "epoch": 5.720174828648058,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00034238998708652034,
+      "loss": 0.4952,
+      "step": 115170
+    },
+    {
+      "epoch": 5.720671500943677,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00034235025330287076,
+      "loss": 0.4982,
+      "step": 115180
+    },
+    {
+      "epoch": 5.721168173239297,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00034231051951922123,
+      "loss": 0.4914,
+      "step": 115190
+    },
+    {
+      "epoch": 5.721664845534916,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0003422707857355717,
+      "loss": 0.5042,
+      "step": 115200
+    },
+    {
+      "epoch": 5.722161517830536,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00034223105195192217,
+      "loss": 0.503,
+      "step": 115210
+    },
+    {
+      "epoch": 5.722658190126155,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0003421913181682726,
+      "loss": 0.4848,
+      "step": 115220
+    },
+    {
+      "epoch": 5.723154862421774,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.000342151584384623,
+      "loss": 0.4676,
+      "step": 115230
+    },
+    {
+      "epoch": 5.723651534717393,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.00034211185060097353,
+      "loss": 0.4628,
+      "step": 115240
+    },
+    {
+      "epoch": 5.7241482070130125,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00034207211681732395,
+      "loss": 0.5114,
+      "step": 115250
+    },
+    {
+      "epoch": 5.724644879308633,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003420323830336744,
+      "loss": 0.5178,
+      "step": 115260
+    },
+    {
+      "epoch": 5.725141551604252,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00034199264925002484,
+      "loss": 0.49,
+      "step": 115270
+    },
+    {
+      "epoch": 5.725638223899871,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003419529154663753,
+      "loss": 0.4979,
+      "step": 115280
+    },
+    {
+      "epoch": 5.72613489619549,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0003419131816827258,
+      "loss": 0.52,
+      "step": 115290
+    },
+    {
+      "epoch": 5.7266315684911095,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0003418734478990762,
+      "loss": 0.4943,
+      "step": 115300
+    },
+    {
+      "epoch": 5.727128240786729,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003418337141154266,
+      "loss": 0.5051,
+      "step": 115310
+    },
+    {
+      "epoch": 5.727624913082348,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00034179398033177714,
+      "loss": 0.4983,
+      "step": 115320
+    },
+    {
+      "epoch": 5.728121585377967,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00034175424654812756,
+      "loss": 0.5153,
+      "step": 115330
+    },
+    {
+      "epoch": 5.728618257673587,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00034171451276447803,
+      "loss": 0.4977,
+      "step": 115340
+    },
+    {
+      "epoch": 5.729114929969207,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00034167477898082845,
+      "loss": 0.5151,
+      "step": 115350
+    },
+    {
+      "epoch": 5.729611602264826,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003416350451971789,
+      "loss": 0.4898,
+      "step": 115360
+    },
+    {
+      "epoch": 5.730108274560445,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003415953114135294,
+      "loss": 0.4805,
+      "step": 115370
+    },
+    {
+      "epoch": 5.730604946856064,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0003415555776298798,
+      "loss": 0.4821,
+      "step": 115380
+    },
+    {
+      "epoch": 5.7311016191516835,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0003415158438462303,
+      "loss": 0.5175,
+      "step": 115390
+    },
+    {
+      "epoch": 5.731598291447303,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00034147611006258075,
+      "loss": 0.5031,
+      "step": 115400
+    },
+    {
+      "epoch": 5.732094963742922,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00034143637627893117,
+      "loss": 0.5136,
+      "step": 115410
+    },
+    {
+      "epoch": 5.732591636038542,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00034139664249528164,
+      "loss": 0.5181,
+      "step": 115420
+    },
+    {
+      "epoch": 5.733088308334161,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0003413569087116321,
+      "loss": 0.4761,
+      "step": 115430
+    },
+    {
+      "epoch": 5.7335849806297805,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0003413171749279825,
+      "loss": 0.513,
+      "step": 115440
+    },
+    {
+      "epoch": 5.7340816529254,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.000341277441144333,
+      "loss": 0.4977,
+      "step": 115450
+    },
+    {
+      "epoch": 5.734578325221019,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003412377073606834,
+      "loss": 0.4923,
+      "step": 115460
+    },
+    {
+      "epoch": 5.735074997516638,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003411979735770339,
+      "loss": 0.4929,
+      "step": 115470
+    },
+    {
+      "epoch": 5.735571669812257,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00034115823979338436,
+      "loss": 0.5056,
+      "step": 115480
+    },
+    {
+      "epoch": 5.7360683421078775,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0003411185060097348,
+      "loss": 0.487,
+      "step": 115490
+    },
+    {
+      "epoch": 5.736565014403497,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00034107877222608524,
+      "loss": 0.483,
+      "step": 115500
+    },
+    {
+      "epoch": 5.737061686699116,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003410390384424357,
+      "loss": 0.5022,
+      "step": 115510
+    },
+    {
+      "epoch": 5.737558358994735,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00034099930465878613,
+      "loss": 0.4968,
+      "step": 115520
+    },
+    {
+      "epoch": 5.738055031290354,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003409595708751366,
+      "loss": 0.5115,
+      "step": 115530
+    },
+    {
+      "epoch": 5.738551703585974,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.000340919837091487,
+      "loss": 0.4868,
+      "step": 115540
+    },
+    {
+      "epoch": 5.739048375881593,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003408801033078375,
+      "loss": 0.5299,
+      "step": 115550
+    },
+    {
+      "epoch": 5.739545048177213,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00034084036952418796,
+      "loss": 0.4985,
+      "step": 115560
+    },
+    {
+      "epoch": 5.740041720472832,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003408006357405384,
+      "loss": 0.4886,
+      "step": 115570
+    },
+    {
+      "epoch": 5.7405383927684515,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00034076090195688885,
+      "loss": 0.4821,
+      "step": 115580
+    },
+    {
+      "epoch": 5.741035065064071,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003407211681732393,
+      "loss": 0.5072,
+      "step": 115590
+    },
+    {
+      "epoch": 5.74153173735969,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00034068143438958974,
+      "loss": 0.5099,
+      "step": 115600
+    },
+    {
+      "epoch": 5.742028409655309,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0003406417006059402,
+      "loss": 0.5313,
+      "step": 115610
+    },
+    {
+      "epoch": 5.742525081950928,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003406019668222907,
+      "loss": 0.5168,
+      "step": 115620
+    },
+    {
+      "epoch": 5.7430217542465485,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00034056223303864116,
+      "loss": 0.4783,
+      "step": 115630
+    },
+    {
+      "epoch": 5.743518426542168,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00034052249925499157,
+      "loss": 0.5161,
+      "step": 115640
+    },
+    {
+      "epoch": 5.744015098837787,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.000340482765471342,
+      "loss": 0.4956,
+      "step": 115650
+    },
+    {
+      "epoch": 5.744511771133406,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003404430316876925,
+      "loss": 0.4915,
+      "step": 115660
+    },
+    {
+      "epoch": 5.745008443429025,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00034040329790404293,
+      "loss": 0.5149,
+      "step": 115670
+    },
+    {
+      "epoch": 5.745505115724645,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00034036356412039335,
+      "loss": 0.4939,
+      "step": 115680
+    },
+    {
+      "epoch": 5.746001788020264,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003403238303367438,
+      "loss": 0.4763,
+      "step": 115690
+    },
+    {
+      "epoch": 5.746498460315884,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003402840965530943,
+      "loss": 0.5225,
+      "step": 115700
+    },
+    {
+      "epoch": 5.746995132611503,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00034024436276944476,
+      "loss": 0.5139,
+      "step": 115710
+    },
+    {
+      "epoch": 5.747491804907122,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003402046289857952,
+      "loss": 0.5062,
+      "step": 115720
+    },
+    {
+      "epoch": 5.747988477202742,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00034016489520214565,
+      "loss": 0.5167,
+      "step": 115730
+    },
+    {
+      "epoch": 5.748485149498361,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0003401251614184961,
+      "loss": 0.5147,
+      "step": 115740
+    },
+    {
+      "epoch": 5.74898182179398,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00034008542763484654,
+      "loss": 0.5476,
+      "step": 115750
+    },
+    {
+      "epoch": 5.749478494089599,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00034004569385119696,
+      "loss": 0.5189,
+      "step": 115760
+    },
+    {
+      "epoch": 5.749975166385219,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003400059600675475,
+      "loss": 0.4933,
+      "step": 115770
+    },
+    {
+      "epoch": 5.750471838680839,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0003399662262838979,
+      "loss": 0.5053,
+      "step": 115780
+    },
+    {
+      "epoch": 5.750968510976458,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00033992649250024837,
+      "loss": 0.5216,
+      "step": 115790
+    },
+    {
+      "epoch": 5.751465183272077,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003398867587165988,
+      "loss": 0.4888,
+      "step": 115800
+    },
+    {
+      "epoch": 5.751961855567696,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00033984702493294926,
+      "loss": 0.5226,
+      "step": 115810
+    },
+    {
+      "epoch": 5.752458527863316,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00033980729114929973,
+      "loss": 0.5049,
+      "step": 115820
+    },
+    {
+      "epoch": 5.752955200158935,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00033976755736565015,
+      "loss": 0.5168,
+      "step": 115830
+    },
+    {
+      "epoch": 5.753451872454555,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00033972782358200057,
+      "loss": 0.4807,
+      "step": 115840
+    },
+    {
+      "epoch": 5.753948544750174,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0003396880897983511,
+      "loss": 0.492,
+      "step": 115850
+    },
+    {
+      "epoch": 5.754445217045793,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003396483560147015,
+      "loss": 0.4788,
+      "step": 115860
+    },
+    {
+      "epoch": 5.754941889341413,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.000339608622231052,
+      "loss": 0.507,
+      "step": 115870
+    },
+    {
+      "epoch": 5.755438561637032,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0003395688884474024,
+      "loss": 0.5107,
+      "step": 115880
+    },
+    {
+      "epoch": 5.755935233932651,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00033952915466375287,
+      "loss": 0.4984,
+      "step": 115890
+    },
+    {
+      "epoch": 5.75643190622827,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00033948942088010334,
+      "loss": 0.5008,
+      "step": 115900
+    },
+    {
+      "epoch": 5.75692857852389,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00033944968709645376,
+      "loss": 0.508,
+      "step": 115910
+    },
+    {
+      "epoch": 5.75742525081951,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00033940995331280423,
+      "loss": 0.5232,
+      "step": 115920
+    },
+    {
+      "epoch": 5.757921923115129,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003393702195291547,
+      "loss": 0.4951,
+      "step": 115930
+    },
+    {
+      "epoch": 5.758418595410748,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003393304857455051,
+      "loss": 0.5019,
+      "step": 115940
+    },
+    {
+      "epoch": 5.758915267706367,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0003392907519618556,
+      "loss": 0.5148,
+      "step": 115950
+    },
+    {
+      "epoch": 5.7594119400019865,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00033925101817820606,
+      "loss": 0.5035,
+      "step": 115960
+    },
+    {
+      "epoch": 5.759908612297606,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003392112843945565,
+      "loss": 0.4882,
+      "step": 115970
+    },
+    {
+      "epoch": 5.760405284593226,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00033917155061090695,
+      "loss": 0.4819,
+      "step": 115980
+    },
+    {
+      "epoch": 5.760901956888845,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00033913181682725737,
+      "loss": 0.5107,
+      "step": 115990
+    },
+    {
+      "epoch": 5.761398629184464,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00033909208304360784,
+      "loss": 0.4972,
+      "step": 116000
+    },
+    {
+      "epoch": 5.7618953014800836,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003390523492599583,
+      "loss": 0.5162,
+      "step": 116010
+    },
+    {
+      "epoch": 5.762391973775703,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003390126154763087,
+      "loss": 0.5025,
+      "step": 116020
+    },
+    {
+      "epoch": 5.762888646071322,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003389728816926592,
+      "loss": 0.5068,
+      "step": 116030
+    },
+    {
+      "epoch": 5.763385318366941,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00033893314790900967,
+      "loss": 0.4853,
+      "step": 116040
+    },
+    {
+      "epoch": 5.7638819906625605,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003388934141253601,
+      "loss": 0.5271,
+      "step": 116050
+    },
+    {
+      "epoch": 5.764378662958181,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00033885368034171056,
+      "loss": 0.4749,
+      "step": 116060
+    },
+    {
+      "epoch": 5.7648753352538,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00033881394655806103,
+      "loss": 0.486,
+      "step": 116070
+    },
+    {
+      "epoch": 5.765372007549419,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0003387742127744115,
+      "loss": 0.4745,
+      "step": 116080
+    },
+    {
+      "epoch": 5.765868679845038,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003387344789907619,
+      "loss": 0.4979,
+      "step": 116090
+    },
+    {
+      "epoch": 5.7663653521406575,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00033869474520711233,
+      "loss": 0.4611,
+      "step": 116100
+    },
+    {
+      "epoch": 5.766862024436277,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003386550114234628,
+      "loss": 0.5248,
+      "step": 116110
+    },
+    {
+      "epoch": 5.767358696731896,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003386152776398133,
+      "loss": 0.5053,
+      "step": 116120
+    },
+    {
+      "epoch": 5.767855369027515,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0003385755438561637,
+      "loss": 0.5074,
+      "step": 116130
+    },
+    {
+      "epoch": 5.768352041323135,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00033853581007251416,
+      "loss": 0.5085,
+      "step": 116140
+    },
+    {
+      "epoch": 5.7688487136187545,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00033849607628886464,
+      "loss": 0.5461,
+      "step": 116150
+    },
+    {
+      "epoch": 5.769345385914374,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003384563425052151,
+      "loss": 0.5136,
+      "step": 116160
+    },
+    {
+      "epoch": 5.769842058209993,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0003384166087215655,
+      "loss": 0.5117,
+      "step": 116170
+    },
+    {
+      "epoch": 5.770338730505612,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00033837687493791594,
+      "loss": 0.5246,
+      "step": 116180
+    },
+    {
+      "epoch": 5.770835402801231,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00033833714115426647,
+      "loss": 0.4963,
+      "step": 116190
+    },
+    {
+      "epoch": 5.771332075096851,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0003382974073706169,
+      "loss": 0.5181,
+      "step": 116200
+    },
+    {
+      "epoch": 5.771828747392471,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0003382576735869673,
+      "loss": 0.4944,
+      "step": 116210
+    },
+    {
+      "epoch": 5.77232541968809,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003382179398033178,
+      "loss": 0.525,
+      "step": 116220
+    },
+    {
+      "epoch": 5.772822091983709,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00033817820601966824,
+      "loss": 0.4985,
+      "step": 116230
+    },
+    {
+      "epoch": 5.773318764279328,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003381384722360187,
+      "loss": 0.4936,
+      "step": 116240
+    },
+    {
+      "epoch": 5.773815436574948,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00033809873845236913,
+      "loss": 0.498,
+      "step": 116250
+    },
+    {
+      "epoch": 5.774312108870567,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0003380590046687196,
+      "loss": 0.5084,
+      "step": 116260
+    },
+    {
+      "epoch": 5.774808781166186,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0003380192708850701,
+      "loss": 0.4746,
+      "step": 116270
+    },
+    {
+      "epoch": 5.775305453461806,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003379795371014205,
+      "loss": 0.5184,
+      "step": 116280
+    },
+    {
+      "epoch": 5.7758021257574255,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0003379398033177709,
+      "loss": 0.5068,
+      "step": 116290
+    },
+    {
+      "epoch": 5.776298798053045,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00033790006953412144,
+      "loss": 0.5128,
+      "step": 116300
+    },
+    {
+      "epoch": 5.776795470348664,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00033786033575047185,
+      "loss": 0.4971,
+      "step": 116310
+    },
+    {
+      "epoch": 5.777292142644283,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0003378206019668223,
+      "loss": 0.5134,
+      "step": 116320
+    },
+    {
+      "epoch": 5.777788814939902,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00033778086818317274,
+      "loss": 0.4887,
+      "step": 116330
+    },
+    {
+      "epoch": 5.778285487235522,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003377411343995232,
+      "loss": 0.4808,
+      "step": 116340
+    },
+    {
+      "epoch": 5.778782159531142,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003377014006158737,
+      "loss": 0.5144,
+      "step": 116350
+    },
+    {
+      "epoch": 5.779278831826761,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0003376616668322241,
+      "loss": 0.4807,
+      "step": 116360
+    },
+    {
+      "epoch": 5.77977550412238,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00033762193304857457,
+      "loss": 0.4983,
+      "step": 116370
+    },
+    {
+      "epoch": 5.780272176417999,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00033758219926492504,
+      "loss": 0.4903,
+      "step": 116380
+    },
+    {
+      "epoch": 5.780768848713619,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00033754246548127546,
+      "loss": 0.5133,
+      "step": 116390
+    },
+    {
+      "epoch": 5.781265521009238,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00033750273169762593,
+      "loss": 0.5031,
+      "step": 116400
+    },
+    {
+      "epoch": 5.781762193304857,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00033746299791397635,
+      "loss": 0.4948,
+      "step": 116410
+    },
+    {
+      "epoch": 5.782258865600477,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003374232641303268,
+      "loss": 0.4974,
+      "step": 116420
+    },
+    {
+      "epoch": 5.782755537896096,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0003373835303466773,
+      "loss": 0.5041,
+      "step": 116430
+    },
+    {
+      "epoch": 5.783252210191716,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0003373437965630277,
+      "loss": 0.5212,
+      "step": 116440
+    },
+    {
+      "epoch": 5.783748882487335,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003373040627793782,
+      "loss": 0.4982,
+      "step": 116450
+    },
+    {
+      "epoch": 5.784245554782954,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00033726432899572865,
+      "loss": 0.5074,
+      "step": 116460
+    },
+    {
+      "epoch": 5.784742227078573,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00033722459521207907,
+      "loss": 0.5313,
+      "step": 116470
+    },
+    {
+      "epoch": 5.785238899374193,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00033718486142842954,
+      "loss": 0.521,
+      "step": 116480
+    },
+    {
+      "epoch": 5.785735571669813,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00033714512764478,
+      "loss": 0.5085,
+      "step": 116490
+    },
+    {
+      "epoch": 5.786232243965432,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00033710539386113043,
+      "loss": 0.5141,
+      "step": 116500
+    },
+    {
+      "epoch": 5.786728916261051,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003370656600774809,
+      "loss": 0.5,
+      "step": 116510
+    },
+    {
+      "epoch": 5.78722558855667,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003370259262938313,
+      "loss": 0.5004,
+      "step": 116520
+    },
+    {
+      "epoch": 5.78772226085229,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00033698619251018184,
+      "loss": 0.4756,
+      "step": 116530
+    },
+    {
+      "epoch": 5.788218933147909,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00033694645872653226,
+      "loss": 0.4956,
+      "step": 116540
+    },
+    {
+      "epoch": 5.788715605443528,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0003369067249428827,
+      "loss": 0.4909,
+      "step": 116550
+    },
+    {
+      "epoch": 5.789212277739148,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00033686699115923315,
+      "loss": 0.5151,
+      "step": 116560
+    },
+    {
+      "epoch": 5.789708950034767,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003368272573755836,
+      "loss": 0.5047,
+      "step": 116570
+    },
+    {
+      "epoch": 5.790205622330387,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00033678752359193404,
+      "loss": 0.5064,
+      "step": 116580
+    },
+    {
+      "epoch": 5.790702294626006,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003367477898082845,
+      "loss": 0.5127,
+      "step": 116590
+    },
+    {
+      "epoch": 5.791198966921625,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.000336708056024635,
+      "loss": 0.5037,
+      "step": 116600
+    },
+    {
+      "epoch": 5.791695639217244,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00033666832224098545,
+      "loss": 0.5015,
+      "step": 116610
+    },
+    {
+      "epoch": 5.7921923115128635,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00033662858845733587,
+      "loss": 0.524,
+      "step": 116620
+    },
+    {
+      "epoch": 5.792688983808484,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003365888546736863,
+      "loss": 0.4881,
+      "step": 116630
+    },
+    {
+      "epoch": 5.793185656104103,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003365491208900368,
+      "loss": 0.5178,
+      "step": 116640
+    },
+    {
+      "epoch": 5.793682328399722,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00033650938710638723,
+      "loss": 0.4994,
+      "step": 116650
+    },
+    {
+      "epoch": 5.794179000695341,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00033646965332273765,
+      "loss": 0.4946,
+      "step": 116660
+    },
+    {
+      "epoch": 5.7946756729909605,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003364299195390881,
+      "loss": 0.4832,
+      "step": 116670
+    },
+    {
+      "epoch": 5.79517234528658,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003363901857554386,
+      "loss": 0.5086,
+      "step": 116680
+    },
+    {
+      "epoch": 5.795669017582199,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00033635045197178906,
+      "loss": 0.5078,
+      "step": 116690
+    },
+    {
+      "epoch": 5.796165689877819,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0003363107181881395,
+      "loss": 0.5327,
+      "step": 116700
+    },
+    {
+      "epoch": 5.796662362173438,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0003362709844044899,
+      "loss": 0.4991,
+      "step": 116710
+    },
+    {
+      "epoch": 5.797159034469058,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0003362312506208404,
+      "loss": 0.4685,
+      "step": 116720
+    },
+    {
+      "epoch": 5.797655706764677,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00033619151683719084,
+      "loss": 0.5092,
+      "step": 116730
+    },
+    {
+      "epoch": 5.798152379060296,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00033615178305354125,
+      "loss": 0.4894,
+      "step": 116740
+    },
+    {
+      "epoch": 5.798649051355915,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0003361120492698917,
+      "loss": 0.4927,
+      "step": 116750
+    },
+    {
+      "epoch": 5.7991457236515345,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003360723154862422,
+      "loss": 0.5244,
+      "step": 116760
+    },
+    {
+      "epoch": 5.799642395947154,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00033603258170259267,
+      "loss": 0.5282,
+      "step": 116770
+    },
+    {
+      "epoch": 5.800139068242773,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0003359928479189431,
+      "loss": 0.5163,
+      "step": 116780
+    },
+    {
+      "epoch": 5.800635740538393,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00033595311413529356,
+      "loss": 0.5125,
+      "step": 116790
+    },
+    {
+      "epoch": 5.801132412834012,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00033591338035164403,
+      "loss": 0.5156,
+      "step": 116800
+    },
+    {
+      "epoch": 5.8016290851296315,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00033587364656799444,
+      "loss": 0.4645,
+      "step": 116810
+    },
+    {
+      "epoch": 5.802125757425251,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00033583391278434486,
+      "loss": 0.4997,
+      "step": 116820
+    },
+    {
+      "epoch": 5.80262242972087,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003357941790006954,
+      "loss": 0.5129,
+      "step": 116830
+    },
+    {
+      "epoch": 5.803119102016489,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003357544452170458,
+      "loss": 0.4992,
+      "step": 116840
+    },
+    {
+      "epoch": 5.803615774312108,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0003357147114333963,
+      "loss": 0.5252,
+      "step": 116850
+    },
+    {
+      "epoch": 5.8041124466077285,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003356749776497467,
+      "loss": 0.5282,
+      "step": 116860
+    },
+    {
+      "epoch": 5.804609118903348,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00033563524386609716,
+      "loss": 0.4747,
+      "step": 116870
+    },
+    {
+      "epoch": 5.805105791198967,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00033559551008244764,
+      "loss": 0.4811,
+      "step": 116880
+    },
+    {
+      "epoch": 5.805602463494586,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00033555577629879805,
+      "loss": 0.4758,
+      "step": 116890
+    },
+    {
+      "epoch": 5.806099135790205,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0003355160425151485,
+      "loss": 0.5076,
+      "step": 116900
+    },
+    {
+      "epoch": 5.806595808085825,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000335476308731499,
+      "loss": 0.4745,
+      "step": 116910
+    },
+    {
+      "epoch": 5.807092480381444,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0003354365749478494,
+      "loss": 0.5009,
+      "step": 116920
+    },
+    {
+      "epoch": 5.807589152677064,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003353968411641999,
+      "loss": 0.5389,
+      "step": 116930
+    },
+    {
+      "epoch": 5.808085824972683,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00033535710738055036,
+      "loss": 0.5034,
+      "step": 116940
+    },
+    {
+      "epoch": 5.8085824972683024,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00033531737359690077,
+      "loss": 0.5461,
+      "step": 116950
+    },
+    {
+      "epoch": 5.809079169563922,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00033527763981325124,
+      "loss": 0.4941,
+      "step": 116960
+    },
+    {
+      "epoch": 5.809575841859541,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00033523790602960166,
+      "loss": 0.5037,
+      "step": 116970
+    },
+    {
+      "epoch": 5.81007251415516,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.00033519817224595213,
+      "loss": 0.5012,
+      "step": 116980
+    },
+    {
+      "epoch": 5.810569186450779,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003351584384623026,
+      "loss": 0.4791,
+      "step": 116990
+    },
+    {
+      "epoch": 5.8110658587463995,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.000335118704678653,
+      "loss": 0.49,
+      "step": 117000
+    },
+    {
+      "epoch": 5.811562531042019,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0003350789708950035,
+      "loss": 0.4665,
+      "step": 117010
+    },
+    {
+      "epoch": 5.812059203337638,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00033503923711135396,
+      "loss": 0.5215,
+      "step": 117020
+    },
+    {
+      "epoch": 5.812555875633257,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0003349995033277044,
+      "loss": 0.5096,
+      "step": 117030
+    },
+    {
+      "epoch": 5.813052547928876,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00033495976954405485,
+      "loss": 0.4757,
+      "step": 117040
+    },
+    {
+      "epoch": 5.813549220224496,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00033492003576040527,
+      "loss": 0.4872,
+      "step": 117050
+    },
+    {
+      "epoch": 5.814045892520115,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0003348803019767558,
+      "loss": 0.4954,
+      "step": 117060
+    },
+    {
+      "epoch": 5.814542564815735,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0003348405681931062,
+      "loss": 0.5031,
+      "step": 117070
+    },
+    {
+      "epoch": 5.815039237111354,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00033480083440945663,
+      "loss": 0.5036,
+      "step": 117080
+    },
+    {
+      "epoch": 5.815535909406973,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0003347611006258071,
+      "loss": 0.4804,
+      "step": 117090
+    },
+    {
+      "epoch": 5.816032581702593,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.00033472136684215757,
+      "loss": 0.5143,
+      "step": 117100
+    },
+    {
+      "epoch": 5.816529253998212,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.000334681633058508,
+      "loss": 0.5067,
+      "step": 117110
+    },
+    {
+      "epoch": 5.817025926293831,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00033464189927485846,
+      "loss": 0.5047,
+      "step": 117120
+    },
+    {
+      "epoch": 5.81752259858945,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00033460216549120893,
+      "loss": 0.5313,
+      "step": 117130
+    },
+    {
+      "epoch": 5.81801927088507,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003345624317075594,
+      "loss": 0.5599,
+      "step": 117140
+    },
+    {
+      "epoch": 5.81851594318069,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0003345226979239098,
+      "loss": 0.4944,
+      "step": 117150
+    },
+    {
+      "epoch": 5.819012615476309,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00033448296414026024,
+      "loss": 0.5168,
+      "step": 117160
+    },
+    {
+      "epoch": 5.819509287771928,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00033444323035661076,
+      "loss": 0.4926,
+      "step": 117170
+    },
+    {
+      "epoch": 5.820005960067547,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0003344034965729612,
+      "loss": 0.4978,
+      "step": 117180
+    },
+    {
+      "epoch": 5.820502632363167,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0003343637627893116,
+      "loss": 0.5229,
+      "step": 117190
+    },
+    {
+      "epoch": 5.820999304658786,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00033432402900566207,
+      "loss": 0.5083,
+      "step": 117200
+    },
+    {
+      "epoch": 5.821495976954406,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00033428429522201254,
+      "loss": 0.5103,
+      "step": 117210
+    },
+    {
+      "epoch": 5.821992649250025,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.000334244561438363,
+      "loss": 0.508,
+      "step": 117220
+    },
+    {
+      "epoch": 5.822489321545644,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00033420482765471343,
+      "loss": 0.4983,
+      "step": 117230
+    },
+    {
+      "epoch": 5.822985993841264,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00033416509387106385,
+      "loss": 0.5179,
+      "step": 117240
+    },
+    {
+      "epoch": 5.823482666136883,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00033412536008741437,
+      "loss": 0.5201,
+      "step": 117250
+    },
+    {
+      "epoch": 5.823979338432502,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003340856263037648,
+      "loss": 0.5031,
+      "step": 117260
+    },
+    {
+      "epoch": 5.824476010728121,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0003340458925201152,
+      "loss": 0.5299,
+      "step": 117270
+    },
+    {
+      "epoch": 5.824972683023741,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003340061587364657,
+      "loss": 0.5158,
+      "step": 117280
+    },
+    {
+      "epoch": 5.825469355319361,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00033396642495281615,
+      "loss": 0.4898,
+      "step": 117290
+    },
+    {
+      "epoch": 5.82596602761498,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003339266911691666,
+      "loss": 0.5246,
+      "step": 117300
+    },
+    {
+      "epoch": 5.826462699910599,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00033388695738551704,
+      "loss": 0.5086,
+      "step": 117310
+    },
+    {
+      "epoch": 5.826959372206218,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0003338472236018675,
+      "loss": 0.4911,
+      "step": 117320
+    },
+    {
+      "epoch": 5.8274560445018375,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.000333807489818218,
+      "loss": 0.5211,
+      "step": 117330
+    },
+    {
+      "epoch": 5.827952716797457,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003337677560345684,
+      "loss": 0.514,
+      "step": 117340
+    },
+    {
+      "epoch": 5.828449389093077,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00033372802225091887,
+      "loss": 0.49,
+      "step": 117350
+    },
+    {
+      "epoch": 5.828946061388696,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00033368828846726934,
+      "loss": 0.511,
+      "step": 117360
+    },
+    {
+      "epoch": 5.829442733684315,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.00033364855468361976,
+      "loss": 0.5215,
+      "step": 117370
+    },
+    {
+      "epoch": 5.8299394059799345,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00033360882089997023,
+      "loss": 0.4963,
+      "step": 117380
+    },
+    {
+      "epoch": 5.830436078275554,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00033356908711632064,
+      "loss": 0.5116,
+      "step": 117390
+    },
+    {
+      "epoch": 5.830932750571173,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0003335293533326711,
+      "loss": 0.4927,
+      "step": 117400
+    },
+    {
+      "epoch": 5.831429422866792,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003334896195490216,
+      "loss": 0.4983,
+      "step": 117410
+    },
+    {
+      "epoch": 5.831926095162412,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.000333449885765372,
+      "loss": 0.4852,
+      "step": 117420
+    },
+    {
+      "epoch": 5.832422767458032,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0003334101519817225,
+      "loss": 0.5369,
+      "step": 117430
+    },
+    {
+      "epoch": 5.832919439753651,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00033337041819807295,
+      "loss": 0.5075,
+      "step": 117440
+    },
+    {
+      "epoch": 5.83341611204927,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00033333068441442336,
+      "loss": 0.5141,
+      "step": 117450
+    },
+    {
+      "epoch": 5.833912784344889,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00033329095063077384,
+      "loss": 0.5138,
+      "step": 117460
+    },
+    {
+      "epoch": 5.8344094566405085,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003332512168471243,
+      "loss": 0.5337,
+      "step": 117470
+    },
+    {
+      "epoch": 5.834906128936128,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0003332114830634747,
+      "loss": 0.5158,
+      "step": 117480
+    },
+    {
+      "epoch": 5.835402801231747,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003331717492798252,
+      "loss": 0.497,
+      "step": 117490
+    },
+    {
+      "epoch": 5.835899473527366,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003331320154961756,
+      "loss": 0.4804,
+      "step": 117500
+    },
+    {
+      "epoch": 5.836396145822986,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0003330922817125261,
+      "loss": 0.5316,
+      "step": 117510
+    },
+    {
+      "epoch": 5.8368928181186055,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00033305254792887656,
+      "loss": 0.5328,
+      "step": 117520
+    },
+    {
+      "epoch": 5.837389490414225,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00033301281414522697,
+      "loss": 0.4812,
+      "step": 117530
+    },
+    {
+      "epoch": 5.837886162709844,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00033297308036157744,
+      "loss": 0.4885,
+      "step": 117540
+    },
+    {
+      "epoch": 5.838382835005463,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0003329333465779279,
+      "loss": 0.4769,
+      "step": 117550
+    },
+    {
+      "epoch": 5.838879507301082,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00033289361279427833,
+      "loss": 0.4961,
+      "step": 117560
+    },
+    {
+      "epoch": 5.839376179596702,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003328538790106288,
+      "loss": 0.4942,
+      "step": 117570
+    },
+    {
+      "epoch": 5.839872851892322,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003328141452269792,
+      "loss": 0.5328,
+      "step": 117580
+    },
+    {
+      "epoch": 5.840369524187941,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00033277441144332975,
+      "loss": 0.4926,
+      "step": 117590
+    },
+    {
+      "epoch": 5.84086619648356,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00033273467765968016,
+      "loss": 0.5053,
+      "step": 117600
+    },
+    {
+      "epoch": 5.841362868779179,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003326949438760306,
+      "loss": 0.498,
+      "step": 117610
+    },
+    {
+      "epoch": 5.841859541074799,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00033265521009238105,
+      "loss": 0.4987,
+      "step": 117620
+    },
+    {
+      "epoch": 5.842356213370418,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003326154763087315,
+      "loss": 0.4759,
+      "step": 117630
+    },
+    {
+      "epoch": 5.842852885666037,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00033257574252508194,
+      "loss": 0.5073,
+      "step": 117640
+    },
+    {
+      "epoch": 5.843349557961657,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0003325360087414324,
+      "loss": 0.4777,
+      "step": 117650
+    },
+    {
+      "epoch": 5.8438462302572765,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003324962749577829,
+      "loss": 0.5161,
+      "step": 117660
+    },
+    {
+      "epoch": 5.844342902552896,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00033245654117413335,
+      "loss": 0.4854,
+      "step": 117670
+    },
+    {
+      "epoch": 5.844839574848515,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00033241680739048377,
+      "loss": 0.517,
+      "step": 117680
+    },
+    {
+      "epoch": 5.845336247144134,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003323770736068342,
+      "loss": 0.5202,
+      "step": 117690
+    },
+    {
+      "epoch": 5.845832919439753,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003323373398231847,
+      "loss": 0.4899,
+      "step": 117700
+    },
+    {
+      "epoch": 5.846329591735373,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00033229760603953513,
+      "loss": 0.4762,
+      "step": 117710
+    },
+    {
+      "epoch": 5.846826264030993,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00033225787225588555,
+      "loss": 0.4953,
+      "step": 117720
+    },
+    {
+      "epoch": 5.847322936326612,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.000332218138472236,
+      "loss": 0.4721,
+      "step": 117730
+    },
+    {
+      "epoch": 5.847819608622231,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0003321784046885865,
+      "loss": 0.503,
+      "step": 117740
+    },
+    {
+      "epoch": 5.84831628091785,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00033213867090493696,
+      "loss": 0.4832,
+      "step": 117750
+    },
+    {
+      "epoch": 5.84881295321347,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0003320989371212874,
+      "loss": 0.4994,
+      "step": 117760
+    },
+    {
+      "epoch": 5.849309625509089,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00033205920333763785,
+      "loss": 0.4854,
+      "step": 117770
+    },
+    {
+      "epoch": 5.849806297804708,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003320194695539883,
+      "loss": 0.4945,
+      "step": 117780
+    },
+    {
+      "epoch": 5.850302970100328,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00033197973577033874,
+      "loss": 0.49,
+      "step": 117790
+    },
+    {
+      "epoch": 5.850799642395947,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0003319400019866892,
+      "loss": 0.4972,
+      "step": 117800
+    },
+    {
+      "epoch": 5.851296314691567,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00033190026820303963,
+      "loss": 0.4971,
+      "step": 117810
+    },
+    {
+      "epoch": 5.851792986987186,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003318605344193901,
+      "loss": 0.5217,
+      "step": 117820
+    },
+    {
+      "epoch": 5.852289659282805,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00033182080063574057,
+      "loss": 0.5168,
+      "step": 117830
+    },
+    {
+      "epoch": 5.852786331578424,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.000331781066852091,
+      "loss": 0.4887,
+      "step": 117840
+    },
+    {
+      "epoch": 5.8532830038740435,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00033174133306844146,
+      "loss": 0.4975,
+      "step": 117850
+    },
+    {
+      "epoch": 5.853779676169664,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00033170159928479193,
+      "loss": 0.5065,
+      "step": 117860
+    },
+    {
+      "epoch": 5.854276348465283,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.00033166186550114235,
+      "loss": 0.4773,
+      "step": 117870
+    },
+    {
+      "epoch": 5.854773020760902,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0003316221317174928,
+      "loss": 0.507,
+      "step": 117880
+    },
+    {
+      "epoch": 5.855269693056521,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003315823979338433,
+      "loss": 0.4938,
+      "step": 117890
+    },
+    {
+      "epoch": 5.855766365352141,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0003315426641501937,
+      "loss": 0.5063,
+      "step": 117900
+    },
+    {
+      "epoch": 5.85626303764776,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0003315029303665442,
+      "loss": 0.5022,
+      "step": 117910
+    },
+    {
+      "epoch": 5.856759709943379,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0003314631965828946,
+      "loss": 0.5145,
+      "step": 117920
+    },
+    {
+      "epoch": 5.857256382238999,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00033142346279924507,
+      "loss": 0.5042,
+      "step": 117930
+    },
+    {
+      "epoch": 5.857753054534618,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00033138372901559554,
+      "loss": 0.4761,
+      "step": 117940
+    },
+    {
+      "epoch": 5.858249726830238,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00033134399523194596,
+      "loss": 0.4655,
+      "step": 117950
+    },
+    {
+      "epoch": 5.858746399125857,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00033130426144829643,
+      "loss": 0.5248,
+      "step": 117960
+    },
+    {
+      "epoch": 5.859243071421476,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003312645276646469,
+      "loss": 0.5076,
+      "step": 117970
+    },
+    {
+      "epoch": 5.859739743717095,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0003312247938809973,
+      "loss": 0.5143,
+      "step": 117980
+    },
+    {
+      "epoch": 5.8602364160127145,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0003311850600973478,
+      "loss": 0.5171,
+      "step": 117990
+    },
+    {
+      "epoch": 5.860733088308335,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00033114532631369826,
+      "loss": 0.4982,
+      "step": 118000
+    },
+    {
+      "epoch": 5.861229760603954,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003311055925300487,
+      "loss": 0.4797,
+      "step": 118010
+    },
+    {
+      "epoch": 5.861726432899573,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00033106585874639915,
+      "loss": 0.4773,
+      "step": 118020
+    },
+    {
+      "epoch": 5.862223105195192,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00033102612496274956,
+      "loss": 0.5194,
+      "step": 118030
+    },
+    {
+      "epoch": 5.8627197774908115,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0003309863911791001,
+      "loss": 0.5051,
+      "step": 118040
+    },
+    {
+      "epoch": 5.863216449786431,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0003309466573954505,
+      "loss": 0.4984,
+      "step": 118050
+    },
+    {
+      "epoch": 5.86371312208205,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003309069236118009,
+      "loss": 0.4948,
+      "step": 118060
+    },
+    {
+      "epoch": 5.86420979437767,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003308671898281514,
+      "loss": 0.486,
+      "step": 118070
+    },
+    {
+      "epoch": 5.864706466673289,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.00033082745604450187,
+      "loss": 0.5001,
+      "step": 118080
+    },
+    {
+      "epoch": 5.8652031389689085,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003307877222608523,
+      "loss": 0.4748,
+      "step": 118090
+    },
+    {
+      "epoch": 5.865699811264528,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00033074798847720276,
+      "loss": 0.5161,
+      "step": 118100
+    },
+    {
+      "epoch": 5.866196483560147,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0003307082546935532,
+      "loss": 0.521,
+      "step": 118110
+    },
+    {
+      "epoch": 5.866693155855766,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0003306685209099037,
+      "loss": 0.5077,
+      "step": 118120
+    },
+    {
+      "epoch": 5.8671898281513855,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0003306287871262541,
+      "loss": 0.5289,
+      "step": 118130
+    },
+    {
+      "epoch": 5.867686500447005,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00033058905334260453,
+      "loss": 0.5263,
+      "step": 118140
+    },
+    {
+      "epoch": 5.868183172742625,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.000330549319558955,
+      "loss": 0.5321,
+      "step": 118150
+    },
+    {
+      "epoch": 5.868679845038244,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0003305095857753055,
+      "loss": 0.501,
+      "step": 118160
+    },
+    {
+      "epoch": 5.869176517333863,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00033046985199165595,
+      "loss": 0.5024,
+      "step": 118170
+    },
+    {
+      "epoch": 5.8696731896294825,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00033043011820800636,
+      "loss": 0.5344,
+      "step": 118180
+    },
+    {
+      "epoch": 5.870169861925102,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00033039038442435684,
+      "loss": 0.5226,
+      "step": 118190
+    },
+    {
+      "epoch": 5.870666534220721,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003303506506407073,
+      "loss": 0.5285,
+      "step": 118200
+    },
+    {
+      "epoch": 5.87116320651634,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0003303109168570577,
+      "loss": 0.4912,
+      "step": 118210
+    },
+    {
+      "epoch": 5.871659878811959,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00033027118307340814,
+      "loss": 0.4776,
+      "step": 118220
+    },
+    {
+      "epoch": 5.8721565511075795,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.00033023144928975867,
+      "loss": 0.4904,
+      "step": 118230
+    },
+    {
+      "epoch": 5.872653223403199,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003301917155061091,
+      "loss": 0.5071,
+      "step": 118240
+    },
+    {
+      "epoch": 5.873149895698818,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00033015198172245956,
+      "loss": 0.4906,
+      "step": 118250
+    },
+    {
+      "epoch": 5.873646567994437,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00033011224793880997,
+      "loss": 0.495,
+      "step": 118260
+    },
+    {
+      "epoch": 5.874143240290056,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00033007251415516044,
+      "loss": 0.4985,
+      "step": 118270
+    },
+    {
+      "epoch": 5.874639912585676,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0003300327803715109,
+      "loss": 0.4986,
+      "step": 118280
+    },
+    {
+      "epoch": 5.875136584881295,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00032999304658786133,
+      "loss": 0.4812,
+      "step": 118290
+    },
+    {
+      "epoch": 5.875633257176915,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003299533128042118,
+      "loss": 0.4896,
+      "step": 118300
+    },
+    {
+      "epoch": 5.876129929472534,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0003299135790205623,
+      "loss": 0.4845,
+      "step": 118310
+    },
+    {
+      "epoch": 5.876626601768153,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003298738452369127,
+      "loss": 0.5161,
+      "step": 118320
+    },
+    {
+      "epoch": 5.877123274063773,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00032983411145326316,
+      "loss": 0.5102,
+      "step": 118330
+    },
+    {
+      "epoch": 5.877619946359392,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00032979437766961363,
+      "loss": 0.4987,
+      "step": 118340
+    },
+    {
+      "epoch": 5.878116618655011,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00032975464388596405,
+      "loss": 0.5095,
+      "step": 118350
+    },
+    {
+      "epoch": 5.87861329095063,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0003297149101023145,
+      "loss": 0.5482,
+      "step": 118360
+    },
+    {
+      "epoch": 5.8791099632462505,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00032967517631866494,
+      "loss": 0.5039,
+      "step": 118370
+    },
+    {
+      "epoch": 5.87960663554187,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0003296354425350154,
+      "loss": 0.5345,
+      "step": 118380
+    },
+    {
+      "epoch": 5.880103307837489,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0003295957087513659,
+      "loss": 0.4986,
+      "step": 118390
+    },
+    {
+      "epoch": 5.880599980133108,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0003295559749677163,
+      "loss": 0.5201,
+      "step": 118400
+    },
+    {
+      "epoch": 5.881096652428727,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00032951624118406677,
+      "loss": 0.4885,
+      "step": 118410
+    },
+    {
+      "epoch": 5.881593324724347,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00032947650740041724,
+      "loss": 0.505,
+      "step": 118420
+    },
+    {
+      "epoch": 5.882089997019966,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00032943677361676766,
+      "loss": 0.4985,
+      "step": 118430
+    },
+    {
+      "epoch": 5.882586669315586,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00032939703983311813,
+      "loss": 0.5121,
+      "step": 118440
+    },
+    {
+      "epoch": 5.883083341611205,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00032935730604946855,
+      "loss": 0.5012,
+      "step": 118450
+    },
+    {
+      "epoch": 5.883580013906824,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.000329317572265819,
+      "loss": 0.519,
+      "step": 118460
+    },
+    {
+      "epoch": 5.884076686202444,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003292778384821695,
+      "loss": 0.5172,
+      "step": 118470
+    },
+    {
+      "epoch": 5.884573358498063,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003292381046985199,
+      "loss": 0.4944,
+      "step": 118480
+    },
+    {
+      "epoch": 5.885070030793682,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0003291983709148704,
+      "loss": 0.5043,
+      "step": 118490
+    },
+    {
+      "epoch": 5.885566703089301,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00032915863713122085,
+      "loss": 0.4903,
+      "step": 118500
+    },
+    {
+      "epoch": 5.886063375384921,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00032911890334757127,
+      "loss": 0.5015,
+      "step": 118510
+    },
+    {
+      "epoch": 5.886560047680541,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00032907916956392174,
+      "loss": 0.4816,
+      "step": 118520
+    },
+    {
+      "epoch": 5.88705671997616,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0003290394357802722,
+      "loss": 0.5231,
+      "step": 118530
+    },
+    {
+      "epoch": 5.887553392271779,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00032899970199662263,
+      "loss": 0.4856,
+      "step": 118540
+    },
+    {
+      "epoch": 5.888050064567398,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0003289599682129731,
+      "loss": 0.5284,
+      "step": 118550
+    },
+    {
+      "epoch": 5.8885467368630175,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0003289202344293235,
+      "loss": 0.4831,
+      "step": 118560
+    },
+    {
+      "epoch": 5.889043409158637,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00032888050064567404,
+      "loss": 0.5038,
+      "step": 118570
+    },
+    {
+      "epoch": 5.889540081454257,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00032884076686202446,
+      "loss": 0.5066,
+      "step": 118580
+    },
+    {
+      "epoch": 5.890036753749876,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003288010330783749,
+      "loss": 0.5145,
+      "step": 118590
+    },
+    {
+      "epoch": 5.890533426045495,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00032876129929472535,
+      "loss": 0.5495,
+      "step": 118600
+    },
+    {
+      "epoch": 5.891030098341115,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0003287215655110758,
+      "loss": 0.4683,
+      "step": 118610
+    },
+    {
+      "epoch": 5.891526770636734,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003286818317274263,
+      "loss": 0.4913,
+      "step": 118620
+    },
+    {
+      "epoch": 5.892023442932353,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0003286420979437767,
+      "loss": 0.5002,
+      "step": 118630
+    },
+    {
+      "epoch": 5.892520115227972,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0003286023641601272,
+      "loss": 0.4984,
+      "step": 118640
+    },
+    {
+      "epoch": 5.893016787523592,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00032856263037647765,
+      "loss": 0.5122,
+      "step": 118650
+    },
+    {
+      "epoch": 5.893513459819212,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00032852289659282807,
+      "loss": 0.532,
+      "step": 118660
+    },
+    {
+      "epoch": 5.894010132114831,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0003284831628091785,
+      "loss": 0.5041,
+      "step": 118670
+    },
+    {
+      "epoch": 5.89450680441045,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00032844342902552896,
+      "loss": 0.5026,
+      "step": 118680
+    },
+    {
+      "epoch": 5.895003476706069,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00032840369524187943,
+      "loss": 0.4592,
+      "step": 118690
+    },
+    {
+      "epoch": 5.8955001490016885,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003283639614582299,
+      "loss": 0.5028,
+      "step": 118700
+    },
+    {
+      "epoch": 5.895996821297308,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003283242276745803,
+      "loss": 0.4951,
+      "step": 118710
+    },
+    {
+      "epoch": 5.896493493592928,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003282844938909308,
+      "loss": 0.5101,
+      "step": 118720
+    },
+    {
+      "epoch": 5.896990165888547,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00032824476010728126,
+      "loss": 0.4962,
+      "step": 118730
+    },
+    {
+      "epoch": 5.897486838184166,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003282050263236317,
+      "loss": 0.4997,
+      "step": 118740
+    },
+    {
+      "epoch": 5.8979835104797855,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003281652925399821,
+      "loss": 0.5008,
+      "step": 118750
+    },
+    {
+      "epoch": 5.898480182775405,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0003281255587563326,
+      "loss": 0.5007,
+      "step": 118760
+    },
+    {
+      "epoch": 5.898976855071024,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00032808582497268304,
+      "loss": 0.4933,
+      "step": 118770
+    },
+    {
+      "epoch": 5.899473527366643,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003280460911890335,
+      "loss": 0.508,
+      "step": 118780
+    },
+    {
+      "epoch": 5.899970199662263,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0003280063574053839,
+      "loss": 0.4906,
+      "step": 118790
+    },
+    {
+      "epoch": 5.9004668719578826,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0003279666236217344,
+      "loss": 0.5132,
+      "step": 118800
+    },
+    {
+      "epoch": 5.900963544253502,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00032792688983808487,
+      "loss": 0.5058,
+      "step": 118810
+    },
+    {
+      "epoch": 5.901460216549121,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003278871560544353,
+      "loss": 0.5065,
+      "step": 118820
+    },
+    {
+      "epoch": 5.90195688884474,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00032784742227078576,
+      "loss": 0.5009,
+      "step": 118830
+    },
+    {
+      "epoch": 5.9024535611403595,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003278076884871362,
+      "loss": 0.5332,
+      "step": 118840
+    },
+    {
+      "epoch": 5.902950233435979,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00032776795470348664,
+      "loss": 0.493,
+      "step": 118850
+    },
+    {
+      "epoch": 5.903446905731598,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0003277282209198371,
+      "loss": 0.4811,
+      "step": 118860
+    },
+    {
+      "epoch": 5.903943578027217,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003276884871361876,
+      "loss": 0.4869,
+      "step": 118870
+    },
+    {
+      "epoch": 5.904440250322837,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000327648753352538,
+      "loss": 0.5196,
+      "step": 118880
+    },
+    {
+      "epoch": 5.9049369226184565,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003276090195688885,
+      "loss": 0.4773,
+      "step": 118890
+    },
+    {
+      "epoch": 5.905433594914076,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003275692857852389,
+      "loss": 0.5117,
+      "step": 118900
+    },
+    {
+      "epoch": 5.905930267209695,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00032752955200158936,
+      "loss": 0.5024,
+      "step": 118910
+    },
+    {
+      "epoch": 5.906426939505314,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00032748981821793983,
+      "loss": 0.4918,
+      "step": 118920
+    },
+    {
+      "epoch": 5.906923611800933,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00032745008443429025,
+      "loss": 0.503,
+      "step": 118930
+    },
+    {
+      "epoch": 5.907420284096553,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003274103506506407,
+      "loss": 0.5204,
+      "step": 118940
+    },
+    {
+      "epoch": 5.907916956392173,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003273706168669912,
+      "loss": 0.4752,
+      "step": 118950
+    },
+    {
+      "epoch": 5.908413628687792,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003273308830833416,
+      "loss": 0.5238,
+      "step": 118960
+    },
+    {
+      "epoch": 5.908910300983411,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003272911492996921,
+      "loss": 0.4999,
+      "step": 118970
+    },
+    {
+      "epoch": 5.90940697327903,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0003272514155160425,
+      "loss": 0.5208,
+      "step": 118980
+    },
+    {
+      "epoch": 5.90990364557465,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00032721168173239297,
+      "loss": 0.5137,
+      "step": 118990
+    },
+    {
+      "epoch": 5.910400317870269,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00032717194794874344,
+      "loss": 0.4515,
+      "step": 119000
+    },
+    {
+      "epoch": 5.910896990165888,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00032713221416509386,
+      "loss": 0.5002,
+      "step": 119010
+    },
+    {
+      "epoch": 5.911393662461508,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00032709248038144433,
+      "loss": 0.5258,
+      "step": 119020
+    },
+    {
+      "epoch": 5.911890334757127,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0003270527465977948,
+      "loss": 0.4869,
+      "step": 119030
+    },
+    {
+      "epoch": 5.912387007052747,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0003270130128141452,
+      "loss": 0.513,
+      "step": 119040
+    },
+    {
+      "epoch": 5.912883679348366,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003269732790304957,
+      "loss": 0.4948,
+      "step": 119050
+    },
+    {
+      "epoch": 5.913380351643985,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00032693354524684616,
+      "loss": 0.5159,
+      "step": 119060
+    },
+    {
+      "epoch": 5.913877023939604,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00032689381146319663,
+      "loss": 0.4906,
+      "step": 119070
+    },
+    {
+      "epoch": 5.914373696235224,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00032685407767954705,
+      "loss": 0.4982,
+      "step": 119080
+    },
+    {
+      "epoch": 5.914870368530844,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00032681434389589747,
+      "loss": 0.5223,
+      "step": 119090
+    },
+    {
+      "epoch": 5.915367040826463,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.000326774610112248,
+      "loss": 0.5294,
+      "step": 119100
+    },
+    {
+      "epoch": 5.915863713122082,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003267348763285984,
+      "loss": 0.5004,
+      "step": 119110
+    },
+    {
+      "epoch": 5.916360385417701,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00032669514254494883,
+      "loss": 0.5063,
+      "step": 119120
+    },
+    {
+      "epoch": 5.916857057713321,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0003266554087612993,
+      "loss": 0.5225,
+      "step": 119130
+    },
+    {
+      "epoch": 5.91735373000894,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00032661567497764977,
+      "loss": 0.5236,
+      "step": 119140
+    },
+    {
+      "epoch": 5.917850402304559,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00032657594119400024,
+      "loss": 0.5233,
+      "step": 119150
+    },
+    {
+      "epoch": 5.918347074600179,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00032653620741035066,
+      "loss": 0.5082,
+      "step": 119160
+    },
+    {
+      "epoch": 5.918843746895798,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00032649647362670113,
+      "loss": 0.5145,
+      "step": 119170
+    },
+    {
+      "epoch": 5.919340419191418,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0003264567398430516,
+      "loss": 0.4983,
+      "step": 119180
+    },
+    {
+      "epoch": 5.919837091487037,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.000326417006059402,
+      "loss": 0.5243,
+      "step": 119190
+    },
+    {
+      "epoch": 5.920333763782656,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00032637727227575244,
+      "loss": 0.5181,
+      "step": 119200
+    },
+    {
+      "epoch": 5.920830436078275,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0003263375384921029,
+      "loss": 0.5084,
+      "step": 119210
+    },
+    {
+      "epoch": 5.9213271083738945,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0003262978047084534,
+      "loss": 0.4905,
+      "step": 119220
+    },
+    {
+      "epoch": 5.921823780669515,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00032625807092480385,
+      "loss": 0.5001,
+      "step": 119230
+    },
+    {
+      "epoch": 5.922320452965134,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00032621833714115427,
+      "loss": 0.5056,
+      "step": 119240
+    },
+    {
+      "epoch": 5.922817125260753,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00032617860335750474,
+      "loss": 0.4853,
+      "step": 119250
+    },
+    {
+      "epoch": 5.923313797556372,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0003261388695738552,
+      "loss": 0.4933,
+      "step": 119260
+    },
+    {
+      "epoch": 5.9238104698519916,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00032609913579020563,
+      "loss": 0.5344,
+      "step": 119270
+    },
+    {
+      "epoch": 5.924307142147611,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00032605940200655604,
+      "loss": 0.5063,
+      "step": 119280
+    },
+    {
+      "epoch": 5.92480381444323,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00032601966822290657,
+      "loss": 0.5223,
+      "step": 119290
+    },
+    {
+      "epoch": 5.92530048673885,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.000325979934439257,
+      "loss": 0.5342,
+      "step": 119300
+    },
+    {
+      "epoch": 5.925797159034469,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00032594020065560746,
+      "loss": 0.5046,
+      "step": 119310
+    },
+    {
+      "epoch": 5.926293831330089,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003259004668719579,
+      "loss": 0.5225,
+      "step": 119320
+    },
+    {
+      "epoch": 5.926790503625708,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00032586073308830835,
+      "loss": 0.4972,
+      "step": 119330
+    },
+    {
+      "epoch": 5.927287175921327,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0003258209993046588,
+      "loss": 0.5219,
+      "step": 119340
+    },
+    {
+      "epoch": 5.927783848216946,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00032578126552100924,
+      "loss": 0.4868,
+      "step": 119350
+    },
+    {
+      "epoch": 5.9282805205125655,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003257415317373597,
+      "loss": 0.5057,
+      "step": 119360
+    },
+    {
+      "epoch": 5.928777192808186,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003257017979537102,
+      "loss": 0.4783,
+      "step": 119370
+    },
+    {
+      "epoch": 5.929273865103805,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0003256620641700606,
+      "loss": 0.5347,
+      "step": 119380
+    },
+    {
+      "epoch": 5.929770537399424,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00032562233038641107,
+      "loss": 0.4995,
+      "step": 119390
+    },
+    {
+      "epoch": 5.930267209695043,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00032558259660276154,
+      "loss": 0.5117,
+      "step": 119400
+    },
+    {
+      "epoch": 5.9307638819906625,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.00032554286281911196,
+      "loss": 0.4947,
+      "step": 119410
+    },
+    {
+      "epoch": 5.931260554286282,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0003255031290354624,
+      "loss": 0.5476,
+      "step": 119420
+    },
+    {
+      "epoch": 5.931757226581901,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00032546339525181284,
+      "loss": 0.5358,
+      "step": 119430
+    },
+    {
+      "epoch": 5.932253898877521,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00032542366146816337,
+      "loss": 0.4916,
+      "step": 119440
+    },
+    {
+      "epoch": 5.93275057117314,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0003253839276845138,
+      "loss": 0.5134,
+      "step": 119450
+    },
+    {
+      "epoch": 5.9332472434687595,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003253441939008642,
+      "loss": 0.5055,
+      "step": 119460
+    },
+    {
+      "epoch": 5.933743915764379,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0003253044601172147,
+      "loss": 0.5068,
+      "step": 119470
+    },
+    {
+      "epoch": 5.934240588059998,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00032526472633356515,
+      "loss": 0.5089,
+      "step": 119480
+    },
+    {
+      "epoch": 5.934737260355617,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00032522499254991556,
+      "loss": 0.539,
+      "step": 119490
+    },
+    {
+      "epoch": 5.935233932651236,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00032518525876626604,
+      "loss": 0.5004,
+      "step": 119500
+    },
+    {
+      "epoch": 5.935730604946857,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00032514552498261645,
+      "loss": 0.5103,
+      "step": 119510
+    },
+    {
+      "epoch": 5.936227277242476,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.000325105791198967,
+      "loss": 0.4969,
+      "step": 119520
+    },
+    {
+      "epoch": 5.936723949538095,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0003250660574153174,
+      "loss": 0.5127,
+      "step": 119530
+    },
+    {
+      "epoch": 5.937220621833714,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0003250263236316678,
+      "loss": 0.5122,
+      "step": 119540
+    },
+    {
+      "epoch": 5.9377172941293335,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0003249865898480183,
+      "loss": 0.5198,
+      "step": 119550
+    },
+    {
+      "epoch": 5.938213966424953,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00032494685606436875,
+      "loss": 0.5059,
+      "step": 119560
+    },
+    {
+      "epoch": 5.938710638720572,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00032490712228071917,
+      "loss": 0.5045,
+      "step": 119570
+    },
+    {
+      "epoch": 5.939207311016191,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00032486738849706964,
+      "loss": 0.498,
+      "step": 119580
+    },
+    {
+      "epoch": 5.93970398331181,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003248276547134201,
+      "loss": 0.49,
+      "step": 119590
+    },
+    {
+      "epoch": 5.9402006556074305,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003247879209297706,
+      "loss": 0.5009,
+      "step": 119600
+    },
+    {
+      "epoch": 5.94069732790305,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.000324748187146121,
+      "loss": 0.5049,
+      "step": 119610
+    },
+    {
+      "epoch": 5.941194000198669,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0003247084533624714,
+      "loss": 0.4934,
+      "step": 119620
+    },
+    {
+      "epoch": 5.941690672494288,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00032466871957882195,
+      "loss": 0.5154,
+      "step": 119630
+    },
+    {
+      "epoch": 5.942187344789907,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00032462898579517236,
+      "loss": 0.5056,
+      "step": 119640
+    },
+    {
+      "epoch": 5.942684017085527,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0003245892520115228,
+      "loss": 0.4994,
+      "step": 119650
+    },
+    {
+      "epoch": 5.943180689381146,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00032454951822787325,
+      "loss": 0.5202,
+      "step": 119660
+    },
+    {
+      "epoch": 5.943677361676766,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0003245097844442237,
+      "loss": 0.4709,
+      "step": 119670
+    },
+    {
+      "epoch": 5.944174033972385,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003244700506605742,
+      "loss": 0.4979,
+      "step": 119680
+    },
+    {
+      "epoch": 5.944670706268004,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0003244303168769246,
+      "loss": 0.4862,
+      "step": 119690
+    },
+    {
+      "epoch": 5.945167378563624,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003243905830932751,
+      "loss": 0.4885,
+      "step": 119700
+    },
+    {
+      "epoch": 5.945664050859243,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00032435084930962555,
+      "loss": 0.5369,
+      "step": 119710
+    },
+    {
+      "epoch": 5.946160723154862,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00032431111552597597,
+      "loss": 0.5043,
+      "step": 119720
+    },
+    {
+      "epoch": 5.946657395450481,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003242713817423264,
+      "loss": 0.5042,
+      "step": 119730
+    },
+    {
+      "epoch": 5.9471540677461014,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003242316479586769,
+      "loss": 0.4988,
+      "step": 119740
+    },
+    {
+      "epoch": 5.947650740041721,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00032419191417502733,
+      "loss": 0.4771,
+      "step": 119750
+    },
+    {
+      "epoch": 5.94814741233734,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003241521803913778,
+      "loss": 0.4992,
+      "step": 119760
+    },
+    {
+      "epoch": 5.948644084632959,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0003241124466077282,
+      "loss": 0.4915,
+      "step": 119770
+    },
+    {
+      "epoch": 5.949140756928578,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003240727128240787,
+      "loss": 0.4938,
+      "step": 119780
+    },
+    {
+      "epoch": 5.949637429224198,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00032403297904042916,
+      "loss": 0.5259,
+      "step": 119790
+    },
+    {
+      "epoch": 5.950134101519817,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0003239932452567796,
+      "loss": 0.5263,
+      "step": 119800
+    },
+    {
+      "epoch": 5.950630773815437,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00032395351147313,
+      "loss": 0.5156,
+      "step": 119810
+    },
+    {
+      "epoch": 5.951127446111056,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003239137776894805,
+      "loss": 0.4934,
+      "step": 119820
+    },
+    {
+      "epoch": 5.951624118406675,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00032387404390583094,
+      "loss": 0.4919,
+      "step": 119830
+    },
+    {
+      "epoch": 5.952120790702295,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0003238343101221814,
+      "loss": 0.482,
+      "step": 119840
+    },
+    {
+      "epoch": 5.952617462997914,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00032379457633853183,
+      "loss": 0.4674,
+      "step": 119850
+    },
+    {
+      "epoch": 5.953114135293533,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003237548425548823,
+      "loss": 0.5062,
+      "step": 119860
+    },
+    {
+      "epoch": 5.953610807589152,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00032371510877123277,
+      "loss": 0.492,
+      "step": 119870
+    },
+    {
+      "epoch": 5.954107479884772,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003236753749875832,
+      "loss": 0.5158,
+      "step": 119880
+    },
+    {
+      "epoch": 5.954604152180392,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00032363564120393366,
+      "loss": 0.51,
+      "step": 119890
+    },
+    {
+      "epoch": 5.955100824476011,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00032359590742028413,
+      "loss": 0.5204,
+      "step": 119900
+    },
+    {
+      "epoch": 5.95559749677163,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00032355617363663455,
+      "loss": 0.4954,
+      "step": 119910
+    },
+    {
+      "epoch": 5.956094169067249,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.000323516439852985,
+      "loss": 0.4786,
+      "step": 119920
+    },
+    {
+      "epoch": 5.9565908413628685,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0003234767060693355,
+      "loss": 0.4845,
+      "step": 119930
+    },
+    {
+      "epoch": 5.957087513658488,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0003234369722856859,
+      "loss": 0.5167,
+      "step": 119940
+    },
+    {
+      "epoch": 5.957584185954108,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003233972385020364,
+      "loss": 0.5018,
+      "step": 119950
+    },
+    {
+      "epoch": 5.958080858249727,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0003233575047183868,
+      "loss": 0.4911,
+      "step": 119960
+    },
+    {
+      "epoch": 5.958577530545346,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0003233177709347373,
+      "loss": 0.5116,
+      "step": 119970
+    },
+    {
+      "epoch": 5.959074202840966,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00032327803715108774,
+      "loss": 0.4855,
+      "step": 119980
+    },
+    {
+      "epoch": 5.959570875136585,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00032323830336743816,
+      "loss": 0.5309,
+      "step": 119990
+    },
+    {
+      "epoch": 5.960067547432204,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0003231985695837886,
+      "loss": 0.4953,
+      "step": 120000
+    },
+    {
+      "epoch": 5.960564219727823,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003231588358001391,
+      "loss": 0.5211,
+      "step": 120010
+    },
+    {
+      "epoch": 5.961060892023443,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0003231191020164895,
+      "loss": 0.5121,
+      "step": 120020
+    },
+    {
+      "epoch": 5.961557564319063,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00032307936823284,
+      "loss": 0.4955,
+      "step": 120030
+    },
+    {
+      "epoch": 5.962054236614682,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00032303963444919046,
+      "loss": 0.5018,
+      "step": 120040
+    },
+    {
+      "epoch": 5.962550908910301,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00032299990066554093,
+      "loss": 0.5069,
+      "step": 120050
+    },
+    {
+      "epoch": 5.96304758120592,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00032296016688189135,
+      "loss": 0.4962,
+      "step": 120060
+    },
+    {
+      "epoch": 5.9635442535015395,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00032292043309824176,
+      "loss": 0.4756,
+      "step": 120070
+    },
+    {
+      "epoch": 5.964040925797159,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00032288069931459224,
+      "loss": 0.4822,
+      "step": 120080
+    },
+    {
+      "epoch": 5.964537598092779,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0003228409655309427,
+      "loss": 0.5071,
+      "step": 120090
+    },
+    {
+      "epoch": 5.965034270388398,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0003228012317472931,
+      "loss": 0.4711,
+      "step": 120100
+    },
+    {
+      "epoch": 5.965530942684017,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003227614979636436,
+      "loss": 0.4887,
+      "step": 120110
+    },
+    {
+      "epoch": 5.9660276149796365,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00032272176417999407,
+      "loss": 0.5109,
+      "step": 120120
+    },
+    {
+      "epoch": 5.966524287275256,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00032268203039634454,
+      "loss": 0.5055,
+      "step": 120130
+    },
+    {
+      "epoch": 5.967020959570875,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00032264229661269495,
+      "loss": 0.5203,
+      "step": 120140
+    },
+    {
+      "epoch": 5.967517631866494,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00032260256282904537,
+      "loss": 0.5225,
+      "step": 120150
+    },
+    {
+      "epoch": 5.968014304162114,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0003225628290453959,
+      "loss": 0.5164,
+      "step": 120160
+    },
+    {
+      "epoch": 5.9685109764577335,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003225230952617463,
+      "loss": 0.4974,
+      "step": 120170
+    },
+    {
+      "epoch": 5.969007648753353,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00032248336147809673,
+      "loss": 0.5121,
+      "step": 120180
+    },
+    {
+      "epoch": 5.969504321048972,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003224436276944472,
+      "loss": 0.5109,
+      "step": 120190
+    },
+    {
+      "epoch": 5.970000993344591,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003224038939107977,
+      "loss": 0.5188,
+      "step": 120200
+    },
+    {
+      "epoch": 5.9704976656402104,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00032236416012714815,
+      "loss": 0.4893,
+      "step": 120210
+    },
+    {
+      "epoch": 5.97099433793583,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00032232442634349856,
+      "loss": 0.5092,
+      "step": 120220
+    },
+    {
+      "epoch": 5.971491010231449,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00032228469255984903,
+      "loss": 0.496,
+      "step": 120230
+    },
+    {
+      "epoch": 5.971987682527069,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0003222449587761995,
+      "loss": 0.5161,
+      "step": 120240
+    },
+    {
+      "epoch": 5.972484354822688,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0003222052249925499,
+      "loss": 0.531,
+      "step": 120250
+    },
+    {
+      "epoch": 5.9729810271183075,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0003221654912089004,
+      "loss": 0.4973,
+      "step": 120260
+    },
+    {
+      "epoch": 5.973477699413927,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00032212575742525087,
+      "loss": 0.5332,
+      "step": 120270
+    },
+    {
+      "epoch": 5.973974371709546,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003220860236416013,
+      "loss": 0.5034,
+      "step": 120280
+    },
+    {
+      "epoch": 5.974471044005165,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00032204628985795175,
+      "loss": 0.5177,
+      "step": 120290
+    },
+    {
+      "epoch": 5.974967716300784,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00032200655607430217,
+      "loss": 0.4775,
+      "step": 120300
+    },
+    {
+      "epoch": 5.975464388596404,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00032196682229065264,
+      "loss": 0.4971,
+      "step": 120310
+    },
+    {
+      "epoch": 5.975961060892024,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003219270885070031,
+      "loss": 0.5205,
+      "step": 120320
+    },
+    {
+      "epoch": 5.976457733187643,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00032188735472335353,
+      "loss": 0.507,
+      "step": 120330
+    },
+    {
+      "epoch": 5.976954405483262,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.000321847620939704,
+      "loss": 0.5493,
+      "step": 120340
+    },
+    {
+      "epoch": 5.977451077778881,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003218078871560545,
+      "loss": 0.4735,
+      "step": 120350
+    },
+    {
+      "epoch": 5.977947750074501,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0003217681533724049,
+      "loss": 0.5099,
+      "step": 120360
+    },
+    {
+      "epoch": 5.97844442237012,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00032172841958875536,
+      "loss": 0.486,
+      "step": 120370
+    },
+    {
+      "epoch": 5.978941094665739,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003216886858051058,
+      "loss": 0.5223,
+      "step": 120380
+    },
+    {
+      "epoch": 5.979437766961359,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00032164895202145625,
+      "loss": 0.489,
+      "step": 120390
+    },
+    {
+      "epoch": 5.979934439256978,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003216092182378067,
+      "loss": 0.4881,
+      "step": 120400
+    },
+    {
+      "epoch": 5.980431111552598,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00032156948445415714,
+      "loss": 0.4873,
+      "step": 120410
+    },
+    {
+      "epoch": 5.980927783848217,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003215297506705076,
+      "loss": 0.4935,
+      "step": 120420
+    },
+    {
+      "epoch": 5.981424456143836,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003214900168868581,
+      "loss": 0.5151,
+      "step": 120430
+    },
+    {
+      "epoch": 5.981921128439455,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003214502831032085,
+      "loss": 0.4887,
+      "step": 120440
+    },
+    {
+      "epoch": 5.982417800735075,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00032141054931955897,
+      "loss": 0.4957,
+      "step": 120450
+    },
+    {
+      "epoch": 5.982914473030695,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00032137081553590944,
+      "loss": 0.5141,
+      "step": 120460
+    },
+    {
+      "epoch": 5.983411145326314,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00032133108175225986,
+      "loss": 0.4862,
+      "step": 120470
+    },
+    {
+      "epoch": 5.983907817621933,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00032129134796861033,
+      "loss": 0.5018,
+      "step": 120480
+    },
+    {
+      "epoch": 5.984404489917552,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00032125161418496075,
+      "loss": 0.5064,
+      "step": 120490
+    },
+    {
+      "epoch": 5.984901162213172,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003212118804013113,
+      "loss": 0.4906,
+      "step": 120500
+    },
+    {
+      "epoch": 5.985397834508791,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003211721466176617,
+      "loss": 0.4987,
+      "step": 120510
+    },
+    {
+      "epoch": 5.98589450680441,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003211324128340121,
+      "loss": 0.4978,
+      "step": 120520
+    },
+    {
+      "epoch": 5.98639117910003,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003210926790503626,
+      "loss": 0.4796,
+      "step": 120530
+    },
+    {
+      "epoch": 5.986887851395649,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00032105294526671305,
+      "loss": 0.5211,
+      "step": 120540
+    },
+    {
+      "epoch": 5.987384523691269,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.00032101321148306347,
+      "loss": 0.5118,
+      "step": 120550
+    },
+    {
+      "epoch": 5.987881195986888,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00032097347769941394,
+      "loss": 0.5214,
+      "step": 120560
+    },
+    {
+      "epoch": 5.988377868282507,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003209337439157644,
+      "loss": 0.5044,
+      "step": 120570
+    },
+    {
+      "epoch": 5.988874540578126,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003208940101321149,
+      "loss": 0.4716,
+      "step": 120580
+    },
+    {
+      "epoch": 5.9893712128737455,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003208542763484653,
+      "loss": 0.5025,
+      "step": 120590
+    },
+    {
+      "epoch": 5.989867885169366,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003208145425648157,
+      "loss": 0.4836,
+      "step": 120600
+    },
+    {
+      "epoch": 5.990364557464985,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003207748087811662,
+      "loss": 0.5115,
+      "step": 120610
+    },
+    {
+      "epoch": 5.990861229760604,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00032073507499751666,
+      "loss": 0.4903,
+      "step": 120620
+    },
+    {
+      "epoch": 5.991357902056223,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0003206953412138671,
+      "loss": 0.4831,
+      "step": 120630
+    },
+    {
+      "epoch": 5.9918545743518425,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00032065560743021755,
+      "loss": 0.4961,
+      "step": 120640
+    },
+    {
+      "epoch": 5.992351246647462,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.000320615873646568,
+      "loss": 0.4973,
+      "step": 120650
+    },
+    {
+      "epoch": 5.992847918943081,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0003205761398629185,
+      "loss": 0.4923,
+      "step": 120660
+    },
+    {
+      "epoch": 5.993344591238701,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003205364060792689,
+      "loss": 0.5084,
+      "step": 120670
+    },
+    {
+      "epoch": 5.99384126353432,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0003204966722956193,
+      "loss": 0.5122,
+      "step": 120680
+    },
+    {
+      "epoch": 5.99433793582994,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00032045693851196985,
+      "loss": 0.4817,
+      "step": 120690
+    },
+    {
+      "epoch": 5.994834608125559,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00032041720472832027,
+      "loss": 0.5426,
+      "step": 120700
+    },
+    {
+      "epoch": 5.995331280421178,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00032037747094467074,
+      "loss": 0.514,
+      "step": 120710
+    },
+    {
+      "epoch": 5.995827952716797,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00032033773716102116,
+      "loss": 0.4534,
+      "step": 120720
+    },
+    {
+      "epoch": 5.9963246250124165,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003202980033773716,
+      "loss": 0.4894,
+      "step": 120730
+    },
+    {
+      "epoch": 5.996821297308037,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0003202582695937221,
+      "loss": 0.495,
+      "step": 120740
+    },
+    {
+      "epoch": 5.997317969603656,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003202185358100725,
+      "loss": 0.481,
+      "step": 120750
+    },
+    {
+      "epoch": 5.997814641899275,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.000320178802026423,
+      "loss": 0.4966,
+      "step": 120760
+    },
+    {
+      "epoch": 5.998311314194894,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00032013906824277346,
+      "loss": 0.5028,
+      "step": 120770
+    },
+    {
+      "epoch": 5.9988079864905135,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003200993344591239,
+      "loss": 0.4975,
+      "step": 120780
+    },
+    {
+      "epoch": 5.999304658786133,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00032005960067547435,
+      "loss": 0.4972,
+      "step": 120790
+    },
+    {
+      "epoch": 5.999801331081752,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003200198668918248,
+      "loss": 0.5059,
+      "step": 120800
+    },
+    {
+      "epoch": 6.000298003377371,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00031998013310817523,
+      "loss": 0.4947,
+      "step": 120810
+    },
+    {
+      "epoch": 6.000794675672991,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003199403993245257,
+      "loss": 0.4986,
+      "step": 120820
+    },
+    {
+      "epoch": 6.0012913479686105,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0003199006655408761,
+      "loss": 0.4952,
+      "step": 120830
+    },
+    {
+      "epoch": 6.00178802026423,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0003198609317572266,
+      "loss": 0.4735,
+      "step": 120840
+    },
+    {
+      "epoch": 6.002284692559849,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00031982119797357707,
+      "loss": 0.5047,
+      "step": 120850
+    },
+    {
+      "epoch": 6.002781364855468,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.0003197814641899275,
+      "loss": 0.5082,
+      "step": 120860
+    },
+    {
+      "epoch": 6.003278037151087,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00031974173040627795,
+      "loss": 0.5107,
+      "step": 120870
+    },
+    {
+      "epoch": 6.003774709446707,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003197019966226284,
+      "loss": 0.488,
+      "step": 120880
+    },
+    {
+      "epoch": 6.004271381742327,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00031966226283897884,
+      "loss": 0.4976,
+      "step": 120890
+    },
+    {
+      "epoch": 6.004768054037946,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003196225290553293,
+      "loss": 0.4974,
+      "step": 120900
+    },
+    {
+      "epoch": 6.005264726333565,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00031958279527167973,
+      "loss": 0.4583,
+      "step": 120910
+    },
+    {
+      "epoch": 6.0057613986291845,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003195430614880302,
+      "loss": 0.4876,
+      "step": 120920
+    },
+    {
+      "epoch": 6.006258070924804,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0003195033277043807,
+      "loss": 0.4885,
+      "step": 120930
+    },
+    {
+      "epoch": 6.006754743220423,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003194635939207311,
+      "loss": 0.5095,
+      "step": 120940
+    },
+    {
+      "epoch": 6.007251415516042,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00031942386013708156,
+      "loss": 0.4792,
+      "step": 120950
+    },
+    {
+      "epoch": 6.007748087811662,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00031938412635343203,
+      "loss": 0.4897,
+      "step": 120960
+    },
+    {
+      "epoch": 6.0082447601072815,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00031934439256978245,
+      "loss": 0.4957,
+      "step": 120970
+    },
+    {
+      "epoch": 6.008741432402901,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0003193046587861329,
+      "loss": 0.4716,
+      "step": 120980
+    },
+    {
+      "epoch": 6.00923810469852,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0003192649250024834,
+      "loss": 0.4969,
+      "step": 120990
+    },
+    {
+      "epoch": 6.009734776994139,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0003192251912188338,
+      "loss": 0.4757,
+      "step": 121000
+    },
+    {
+      "epoch": 6.010231449289758,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003191854574351843,
+      "loss": 0.4886,
+      "step": 121010
+    },
+    {
+      "epoch": 6.010728121585378,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003191457236515347,
+      "loss": 0.4558,
+      "step": 121020
+    },
+    {
+      "epoch": 6.011224793880998,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003191059898678852,
+      "loss": 0.476,
+      "step": 121030
+    },
+    {
+      "epoch": 6.011721466176617,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00031906625608423564,
+      "loss": 0.5165,
+      "step": 121040
+    },
+    {
+      "epoch": 6.012218138472236,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00031902652230058606,
+      "loss": 0.493,
+      "step": 121050
+    },
+    {
+      "epoch": 6.012714810767855,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00031898678851693653,
+      "loss": 0.4721,
+      "step": 121060
+    },
+    {
+      "epoch": 6.013211483063475,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000318947054733287,
+      "loss": 0.5085,
+      "step": 121070
+    },
+    {
+      "epoch": 6.013708155359094,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0003189073209496374,
+      "loss": 0.4958,
+      "step": 121080
+    },
+    {
+      "epoch": 6.014204827654713,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0003188675871659879,
+      "loss": 0.4813,
+      "step": 121090
+    },
+    {
+      "epoch": 6.014701499950333,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00031882785338233836,
+      "loss": 0.4742,
+      "step": 121100
+    },
+    {
+      "epoch": 6.015198172245952,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00031878811959868883,
+      "loss": 0.484,
+      "step": 121110
+    },
+    {
+      "epoch": 6.015694844541572,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00031874838581503925,
+      "loss": 0.4848,
+      "step": 121120
+    },
+    {
+      "epoch": 6.016191516837191,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00031870865203138967,
+      "loss": 0.4943,
+      "step": 121130
+    },
+    {
+      "epoch": 6.01668818913281,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003186689182477402,
+      "loss": 0.4791,
+      "step": 121140
+    },
+    {
+      "epoch": 6.017184861428429,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003186291844640906,
+      "loss": 0.4831,
+      "step": 121150
+    },
+    {
+      "epoch": 6.017681533724049,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0003185894506804411,
+      "loss": 0.49,
+      "step": 121160
+    },
+    {
+      "epoch": 6.018178206019668,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0003185497168967915,
+      "loss": 0.5079,
+      "step": 121170
+    },
+    {
+      "epoch": 6.018674878315288,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00031850998311314197,
+      "loss": 0.4653,
+      "step": 121180
+    },
+    {
+      "epoch": 6.019171550610907,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00031847024932949244,
+      "loss": 0.479,
+      "step": 121190
+    },
+    {
+      "epoch": 6.019668222906526,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00031843051554584286,
+      "loss": 0.5137,
+      "step": 121200
+    },
+    {
+      "epoch": 6.020164895202146,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003183907817621933,
+      "loss": 0.4862,
+      "step": 121210
+    },
+    {
+      "epoch": 6.020661567497765,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0003183510479785438,
+      "loss": 0.5049,
+      "step": 121220
+    },
+    {
+      "epoch": 6.021158239793384,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003183113141948942,
+      "loss": 0.4808,
+      "step": 121230
+    },
+    {
+      "epoch": 6.021654912089003,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003182715804112447,
+      "loss": 0.5062,
+      "step": 121240
+    },
+    {
+      "epoch": 6.022151584384623,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0003182318466275951,
+      "loss": 0.4616,
+      "step": 121250
+    },
+    {
+      "epoch": 6.022648256680243,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003181921128439456,
+      "loss": 0.4979,
+      "step": 121260
+    },
+    {
+      "epoch": 6.023144928975862,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00031815237906029605,
+      "loss": 0.479,
+      "step": 121270
+    },
+    {
+      "epoch": 6.023641601271481,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00031811264527664647,
+      "loss": 0.5044,
+      "step": 121280
+    },
+    {
+      "epoch": 6.0241382735671,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00031807291149299694,
+      "loss": 0.4953,
+      "step": 121290
+    },
+    {
+      "epoch": 6.0246349458627195,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003180331777093474,
+      "loss": 0.477,
+      "step": 121300
+    },
+    {
+      "epoch": 6.025131618158339,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0003179934439256978,
+      "loss": 0.4868,
+      "step": 121310
+    },
+    {
+      "epoch": 6.025628290453959,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003179537101420483,
+      "loss": 0.4859,
+      "step": 121320
+    },
+    {
+      "epoch": 6.026124962749578,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00031791397635839877,
+      "loss": 0.4923,
+      "step": 121330
+    },
+    {
+      "epoch": 6.026621635045197,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003178742425747492,
+      "loss": 0.5309,
+      "step": 121340
+    },
+    {
+      "epoch": 6.0271183073408165,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00031783450879109966,
+      "loss": 0.4882,
+      "step": 121350
+    },
+    {
+      "epoch": 6.027614979636436,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003177947750074501,
+      "loss": 0.5061,
+      "step": 121360
+    },
+    {
+      "epoch": 6.028111651932055,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00031775504122380055,
+      "loss": 0.4982,
+      "step": 121370
+    },
+    {
+      "epoch": 6.028608324227674,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.000317715307440151,
+      "loss": 0.4912,
+      "step": 121380
+    },
+    {
+      "epoch": 6.029104996523294,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00031767557365650143,
+      "loss": 0.4962,
+      "step": 121390
+    },
+    {
+      "epoch": 6.029601668818914,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0003176358398728519,
+      "loss": 0.5142,
+      "step": 121400
+    },
+    {
+      "epoch": 6.030098341114533,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0003175961060892024,
+      "loss": 0.5126,
+      "step": 121410
+    },
+    {
+      "epoch": 6.030595013410152,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0003175563723055528,
+      "loss": 0.4588,
+      "step": 121420
+    },
+    {
+      "epoch": 6.031091685705771,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00031751663852190327,
+      "loss": 0.4973,
+      "step": 121430
+    },
+    {
+      "epoch": 6.0315883580013905,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00031747690473825374,
+      "loss": 0.504,
+      "step": 121440
+    },
+    {
+      "epoch": 6.03208503029701,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.00031743717095460415,
+      "loss": 0.5137,
+      "step": 121450
+    },
+    {
+      "epoch": 6.03258170259263,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0003173974371709546,
+      "loss": 0.4801,
+      "step": 121460
+    },
+    {
+      "epoch": 6.033078374888249,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00031735770338730504,
+      "loss": 0.4453,
+      "step": 121470
+    },
+    {
+      "epoch": 6.033575047183868,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003173179696036555,
+      "loss": 0.5028,
+      "step": 121480
+    },
+    {
+      "epoch": 6.0340717194794875,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.000317278235820006,
+      "loss": 0.5093,
+      "step": 121490
+    },
+    {
+      "epoch": 6.034568391775107,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003172385020363564,
+      "loss": 0.4924,
+      "step": 121500
+    },
+    {
+      "epoch": 6.035065064070726,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0003171987682527069,
+      "loss": 0.5078,
+      "step": 121510
+    },
+    {
+      "epoch": 6.035561736366345,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00031715903446905735,
+      "loss": 0.4967,
+      "step": 121520
+    },
+    {
+      "epoch": 6.036058408661964,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0003171193006854078,
+      "loss": 0.4631,
+      "step": 121530
+    },
+    {
+      "epoch": 6.0365550809575845,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.00031707956690175823,
+      "loss": 0.5028,
+      "step": 121540
+    },
+    {
+      "epoch": 6.037051753253204,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00031703983311810865,
+      "loss": 0.4688,
+      "step": 121550
+    },
+    {
+      "epoch": 6.037548425548823,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0003170000993344592,
+      "loss": 0.4739,
+      "step": 121560
+    },
+    {
+      "epoch": 6.038045097844442,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003169603655508096,
+      "loss": 0.5197,
+      "step": 121570
+    },
+    {
+      "epoch": 6.038541770140061,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00031692063176716,
+      "loss": 0.5127,
+      "step": 121580
+    },
+    {
+      "epoch": 6.039038442435681,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0003168808979835105,
+      "loss": 0.4882,
+      "step": 121590
+    },
+    {
+      "epoch": 6.0395351147313,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00031684116419986095,
+      "loss": 0.4512,
+      "step": 121600
+    },
+    {
+      "epoch": 6.04003178702692,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0003168014304162114,
+      "loss": 0.508,
+      "step": 121610
+    },
+    {
+      "epoch": 6.040528459322539,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00031676169663256184,
+      "loss": 0.4672,
+      "step": 121620
+    },
+    {
+      "epoch": 6.0410251316181585,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0003167219628489123,
+      "loss": 0.5086,
+      "step": 121630
+    },
+    {
+      "epoch": 6.041521803913778,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0003166822290652628,
+      "loss": 0.4855,
+      "step": 121640
+    },
+    {
+      "epoch": 6.042018476209397,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003166424952816132,
+      "loss": 0.4929,
+      "step": 121650
+    },
+    {
+      "epoch": 6.042515148505016,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0003166027614979636,
+      "loss": 0.5249,
+      "step": 121660
+    },
+    {
+      "epoch": 6.043011820800635,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00031656302771431415,
+      "loss": 0.5053,
+      "step": 121670
+    },
+    {
+      "epoch": 6.0435084930962555,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00031652329393066456,
+      "loss": 0.4981,
+      "step": 121680
+    },
+    {
+      "epoch": 6.044005165391875,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00031648356014701503,
+      "loss": 0.5165,
+      "step": 121690
+    },
+    {
+      "epoch": 6.044501837687494,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00031644382636336545,
+      "loss": 0.4961,
+      "step": 121700
+    },
+    {
+      "epoch": 6.044998509983113,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0003164040925797159,
+      "loss": 0.4909,
+      "step": 121710
+    },
+    {
+      "epoch": 6.045495182278732,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003163643587960664,
+      "loss": 0.4829,
+      "step": 121720
+    },
+    {
+      "epoch": 6.045991854574352,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0003163246250124168,
+      "loss": 0.4891,
+      "step": 121730
+    },
+    {
+      "epoch": 6.046488526869971,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003162848912287673,
+      "loss": 0.4948,
+      "step": 121740
+    },
+    {
+      "epoch": 6.046985199165591,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00031624515744511775,
+      "loss": 0.5221,
+      "step": 121750
+    },
+    {
+      "epoch": 6.04748187146121,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00031620542366146817,
+      "loss": 0.4994,
+      "step": 121760
+    },
+    {
+      "epoch": 6.047978543756829,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00031616568987781864,
+      "loss": 0.5257,
+      "step": 121770
+    },
+    {
+      "epoch": 6.048475216052449,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00031612595609416906,
+      "loss": 0.4944,
+      "step": 121780
+    },
+    {
+      "epoch": 6.048971888348068,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00031608622231051953,
+      "loss": 0.5183,
+      "step": 121790
+    },
+    {
+      "epoch": 6.049468560643687,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.00031604648852687,
+      "loss": 0.4723,
+      "step": 121800
+    },
+    {
+      "epoch": 6.049965232939306,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0003160067547432204,
+      "loss": 0.4915,
+      "step": 121810
+    },
+    {
+      "epoch": 6.0504619052349256,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0003159670209595709,
+      "loss": 0.4774,
+      "step": 121820
+    },
+    {
+      "epoch": 6.050958577530546,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00031592728717592136,
+      "loss": 0.5147,
+      "step": 121830
+    },
+    {
+      "epoch": 6.051455249826165,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0003158875533922718,
+      "loss": 0.4558,
+      "step": 121840
+    },
+    {
+      "epoch": 6.051951922121784,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00031584781960862225,
+      "loss": 0.456,
+      "step": 121850
+    },
+    {
+      "epoch": 6.052448594417403,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003158080858249727,
+      "loss": 0.4759,
+      "step": 121860
+    },
+    {
+      "epoch": 6.052945266713023,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00031576835204132314,
+      "loss": 0.4621,
+      "step": 121870
+    },
+    {
+      "epoch": 6.053441939008642,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003157286182576736,
+      "loss": 0.5008,
+      "step": 121880
+    },
+    {
+      "epoch": 6.053938611304261,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.000315688884474024,
+      "loss": 0.5374,
+      "step": 121890
+    },
+    {
+      "epoch": 6.054435283599881,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003156491506903745,
+      "loss": 0.4992,
+      "step": 121900
+    },
+    {
+      "epoch": 6.0549319558955,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00031560941690672497,
+      "loss": 0.5023,
+      "step": 121910
+    },
+    {
+      "epoch": 6.05542862819112,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003155696831230754,
+      "loss": 0.5027,
+      "step": 121920
+    },
+    {
+      "epoch": 6.055925300486739,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00031552994933942586,
+      "loss": 0.4602,
+      "step": 121930
+    },
+    {
+      "epoch": 6.056421972782358,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00031549021555577633,
+      "loss": 0.4808,
+      "step": 121940
+    },
+    {
+      "epoch": 6.056918645077977,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00031545048177212675,
+      "loss": 0.4805,
+      "step": 121950
+    },
+    {
+      "epoch": 6.0574153173735965,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003154107479884772,
+      "loss": 0.5113,
+      "step": 121960
+    },
+    {
+      "epoch": 6.057911989669217,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0003153710142048277,
+      "loss": 0.4748,
+      "step": 121970
+    },
+    {
+      "epoch": 6.058408661964836,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00031533128042117816,
+      "loss": 0.4888,
+      "step": 121980
+    },
+    {
+      "epoch": 6.058905334260455,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0003152915466375286,
+      "loss": 0.4952,
+      "step": 121990
+    },
+    {
+      "epoch": 6.059402006556074,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.000315251812853879,
+      "loss": 0.4945,
+      "step": 122000
+    },
+    {
+      "epoch": 6.0598986788516935,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00031521207907022947,
+      "loss": 0.4795,
+      "step": 122010
+    },
+    {
+      "epoch": 6.060395351147313,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00031517234528657994,
+      "loss": 0.4664,
+      "step": 122020
+    },
+    {
+      "epoch": 6.060892023442932,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00031513261150293035,
+      "loss": 0.4937,
+      "step": 122030
+    },
+    {
+      "epoch": 6.061388695738552,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0003150928777192808,
+      "loss": 0.5135,
+      "step": 122040
+    },
+    {
+      "epoch": 6.061885368034171,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003150531439356313,
+      "loss": 0.4935,
+      "step": 122050
+    },
+    {
+      "epoch": 6.0623820403297906,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00031501341015198177,
+      "loss": 0.4765,
+      "step": 122060
+    },
+    {
+      "epoch": 6.06287871262541,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0003149736763683322,
+      "loss": 0.4895,
+      "step": 122070
+    },
+    {
+      "epoch": 6.063375384921029,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0003149339425846826,
+      "loss": 0.4835,
+      "step": 122080
+    },
+    {
+      "epoch": 6.063872057216648,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00031489420880103313,
+      "loss": 0.4845,
+      "step": 122090
+    },
+    {
+      "epoch": 6.0643687295122675,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00031485447501738355,
+      "loss": 0.4877,
+      "step": 122100
+    },
+    {
+      "epoch": 6.064865401807888,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00031481474123373396,
+      "loss": 0.5008,
+      "step": 122110
+    },
+    {
+      "epoch": 6.065362074103507,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00031477500745008443,
+      "loss": 0.5,
+      "step": 122120
+    },
+    {
+      "epoch": 6.065858746399126,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003147352736664349,
+      "loss": 0.49,
+      "step": 122130
+    },
+    {
+      "epoch": 6.066355418694745,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003146955398827854,
+      "loss": 0.4981,
+      "step": 122140
+    },
+    {
+      "epoch": 6.0668520909903645,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003146558060991358,
+      "loss": 0.4632,
+      "step": 122150
+    },
+    {
+      "epoch": 6.067348763285984,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.00031461607231548627,
+      "loss": 0.4868,
+      "step": 122160
+    },
+    {
+      "epoch": 6.067845435581603,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00031457633853183674,
+      "loss": 0.4793,
+      "step": 122170
+    },
+    {
+      "epoch": 6.068342107877223,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00031453660474818715,
+      "loss": 0.5122,
+      "step": 122180
+    },
+    {
+      "epoch": 6.068838780172842,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00031449687096453757,
+      "loss": 0.4774,
+      "step": 122190
+    },
+    {
+      "epoch": 6.0693354524684615,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003144571371808881,
+      "loss": 0.4966,
+      "step": 122200
+    },
+    {
+      "epoch": 6.069832124764081,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003144174033972385,
+      "loss": 0.4972,
+      "step": 122210
+    },
+    {
+      "epoch": 6.0703287970597,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.000314377669613589,
+      "loss": 0.5048,
+      "step": 122220
+    },
+    {
+      "epoch": 6.070825469355319,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003143379358299394,
+      "loss": 0.4965,
+      "step": 122230
+    },
+    {
+      "epoch": 6.071322141650938,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0003142982020462899,
+      "loss": 0.489,
+      "step": 122240
+    },
+    {
+      "epoch": 6.071818813946558,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00031425846826264035,
+      "loss": 0.4829,
+      "step": 122250
+    },
+    {
+      "epoch": 6.072315486242178,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00031421873447899076,
+      "loss": 0.4837,
+      "step": 122260
+    },
+    {
+      "epoch": 6.072812158537797,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00031417900069534123,
+      "loss": 0.4843,
+      "step": 122270
+    },
+    {
+      "epoch": 6.073308830833416,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0003141392669116917,
+      "loss": 0.4852,
+      "step": 122280
+    },
+    {
+      "epoch": 6.073805503129035,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003140995331280421,
+      "loss": 0.502,
+      "step": 122290
+    },
+    {
+      "epoch": 6.074302175424655,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0003140597993443926,
+      "loss": 0.4878,
+      "step": 122300
+    },
+    {
+      "epoch": 6.074798847720274,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.000314020065560743,
+      "loss": 0.4589,
+      "step": 122310
+    },
+    {
+      "epoch": 6.075295520015893,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0003139803317770935,
+      "loss": 0.4846,
+      "step": 122320
+    },
+    {
+      "epoch": 6.075792192311513,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00031394059799344395,
+      "loss": 0.4871,
+      "step": 122330
+    },
+    {
+      "epoch": 6.0762888646071325,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00031390086420979437,
+      "loss": 0.4678,
+      "step": 122340
+    },
+    {
+      "epoch": 6.076785536902752,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00031386113042614484,
+      "loss": 0.4907,
+      "step": 122350
+    },
+    {
+      "epoch": 6.077282209198371,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003138213966424953,
+      "loss": 0.4916,
+      "step": 122360
+    },
+    {
+      "epoch": 6.07777888149399,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00031378166285884573,
+      "loss": 0.482,
+      "step": 122370
+    },
+    {
+      "epoch": 6.078275553789609,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003137419290751962,
+      "loss": 0.518,
+      "step": 122380
+    },
+    {
+      "epoch": 6.078772226085229,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003137021952915467,
+      "loss": 0.5041,
+      "step": 122390
+    },
+    {
+      "epoch": 6.079268898380849,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003136624615078971,
+      "loss": 0.4746,
+      "step": 122400
+    },
+    {
+      "epoch": 6.079765570676468,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00031362272772424756,
+      "loss": 0.5145,
+      "step": 122410
+    },
+    {
+      "epoch": 6.080262242972087,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.000313582993940598,
+      "loss": 0.4585,
+      "step": 122420
+    },
+    {
+      "epoch": 6.080758915267706,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003135432601569485,
+      "loss": 0.4651,
+      "step": 122430
+    },
+    {
+      "epoch": 6.081255587563326,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0003135035263732989,
+      "loss": 0.4972,
+      "step": 122440
+    },
+    {
+      "epoch": 6.081752259858945,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00031346379258964934,
+      "loss": 0.4813,
+      "step": 122450
+    },
+    {
+      "epoch": 6.082248932154564,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003134240588059998,
+      "loss": 0.5177,
+      "step": 122460
+    },
+    {
+      "epoch": 6.082745604450184,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0003133843250223503,
+      "loss": 0.4936,
+      "step": 122470
+    },
+    {
+      "epoch": 6.083242276745803,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0003133445912387007,
+      "loss": 0.4782,
+      "step": 122480
+    },
+    {
+      "epoch": 6.083738949041423,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00031330485745505117,
+      "loss": 0.4816,
+      "step": 122490
+    },
+    {
+      "epoch": 6.084235621337042,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00031326512367140164,
+      "loss": 0.4851,
+      "step": 122500
+    },
+    {
+      "epoch": 6.084732293632661,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003132253898877521,
+      "loss": 0.5047,
+      "step": 122510
+    },
+    {
+      "epoch": 6.08522896592828,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00031318565610410253,
+      "loss": 0.4938,
+      "step": 122520
+    },
+    {
+      "epoch": 6.0857256382239,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00031314592232045295,
+      "loss": 0.4875,
+      "step": 122530
+    },
+    {
+      "epoch": 6.086222310519519,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00031310618853680347,
+      "loss": 0.5155,
+      "step": 122540
+    },
+    {
+      "epoch": 6.086718982815139,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003130664547531539,
+      "loss": 0.4958,
+      "step": 122550
+    },
+    {
+      "epoch": 6.087215655110758,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003130267209695043,
+      "loss": 0.4881,
+      "step": 122560
+    },
+    {
+      "epoch": 6.087712327406377,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0003129869871858548,
+      "loss": 0.4943,
+      "step": 122570
+    },
+    {
+      "epoch": 6.088208999701997,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00031294725340220525,
+      "loss": 0.5165,
+      "step": 122580
+    },
+    {
+      "epoch": 6.088705671997616,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0003129075196185557,
+      "loss": 0.5117,
+      "step": 122590
+    },
+    {
+      "epoch": 6.089202344293235,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00031286778583490614,
+      "loss": 0.5006,
+      "step": 122600
+    },
+    {
+      "epoch": 6.089699016588854,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00031282805205125656,
+      "loss": 0.4859,
+      "step": 122610
+    },
+    {
+      "epoch": 6.090195688884474,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0003127883182676071,
+      "loss": 0.522,
+      "step": 122620
+    },
+    {
+      "epoch": 6.090692361180094,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003127485844839575,
+      "loss": 0.5042,
+      "step": 122630
+    },
+    {
+      "epoch": 6.091189033475713,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0003127088507003079,
+      "loss": 0.4801,
+      "step": 122640
+    },
+    {
+      "epoch": 6.091685705771332,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003126691169166584,
+      "loss": 0.4733,
+      "step": 122650
+    },
+    {
+      "epoch": 6.092182378066951,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00031262938313300886,
+      "loss": 0.497,
+      "step": 122660
+    },
+    {
+      "epoch": 6.0926790503625705,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00031258964934935933,
+      "loss": 0.4845,
+      "step": 122670
+    },
+    {
+      "epoch": 6.09317572265819,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00031254991556570975,
+      "loss": 0.4707,
+      "step": 122680
+    },
+    {
+      "epoch": 6.09367239495381,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003125101817820602,
+      "loss": 0.4533,
+      "step": 122690
+    },
+    {
+      "epoch": 6.094169067249429,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003124704479984107,
+      "loss": 0.5105,
+      "step": 122700
+    },
+    {
+      "epoch": 6.094665739545048,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0003124307142147611,
+      "loss": 0.4662,
+      "step": 122710
+    },
+    {
+      "epoch": 6.0951624118406675,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003123909804311115,
+      "loss": 0.474,
+      "step": 122720
+    },
+    {
+      "epoch": 6.095659084136287,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00031235124664746205,
+      "loss": 0.4727,
+      "step": 122730
+    },
+    {
+      "epoch": 6.096155756431906,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00031231151286381247,
+      "loss": 0.4855,
+      "step": 122740
+    },
+    {
+      "epoch": 6.096652428727525,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00031227177908016294,
+      "loss": 0.4932,
+      "step": 122750
+    },
+    {
+      "epoch": 6.097149101023145,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00031223204529651335,
+      "loss": 0.4787,
+      "step": 122760
+    },
+    {
+      "epoch": 6.097645773318765,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003121923115128638,
+      "loss": 0.4811,
+      "step": 122770
+    },
+    {
+      "epoch": 6.098142445614384,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0003121525777292143,
+      "loss": 0.5072,
+      "step": 122780
+    },
+    {
+      "epoch": 6.098639117910003,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003121128439455647,
+      "loss": 0.4753,
+      "step": 122790
+    },
+    {
+      "epoch": 6.099135790205622,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0003120731101619152,
+      "loss": 0.4942,
+      "step": 122800
+    },
+    {
+      "epoch": 6.0996324625012415,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00031203337637826566,
+      "loss": 0.5047,
+      "step": 122810
+    },
+    {
+      "epoch": 6.100129134796861,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003119936425946161,
+      "loss": 0.4998,
+      "step": 122820
+    },
+    {
+      "epoch": 6.100625807092481,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00031195390881096655,
+      "loss": 0.5084,
+      "step": 122830
+    },
+    {
+      "epoch": 6.1011224793881,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.000311914175027317,
+      "loss": 0.4893,
+      "step": 122840
+    },
+    {
+      "epoch": 6.101619151683719,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00031187444124366743,
+      "loss": 0.4915,
+      "step": 122850
+    },
+    {
+      "epoch": 6.1021158239793385,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0003118347074600179,
+      "loss": 0.4839,
+      "step": 122860
+    },
+    {
+      "epoch": 6.102612496274958,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003117949736763683,
+      "loss": 0.4837,
+      "step": 122870
+    },
+    {
+      "epoch": 6.103109168570577,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0003117552398927188,
+      "loss": 0.5083,
+      "step": 122880
+    },
+    {
+      "epoch": 6.103605840866196,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00031171550610906927,
+      "loss": 0.477,
+      "step": 122890
+    },
+    {
+      "epoch": 6.104102513161815,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003116757723254197,
+      "loss": 0.4669,
+      "step": 122900
+    },
+    {
+      "epoch": 6.1045991854574355,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00031163603854177015,
+      "loss": 0.5053,
+      "step": 122910
+    },
+    {
+      "epoch": 6.105095857753055,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003115963047581206,
+      "loss": 0.4852,
+      "step": 122920
+    },
+    {
+      "epoch": 6.105592530048674,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.00031155657097447104,
+      "loss": 0.4933,
+      "step": 122930
+    },
+    {
+      "epoch": 6.106089202344293,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0003115168371908215,
+      "loss": 0.4907,
+      "step": 122940
+    },
+    {
+      "epoch": 6.106585874639912,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00031147710340717193,
+      "loss": 0.473,
+      "step": 122950
+    },
+    {
+      "epoch": 6.107082546935532,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00031143736962352246,
+      "loss": 0.4875,
+      "step": 122960
+    },
+    {
+      "epoch": 6.107579219231151,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003113976358398729,
+      "loss": 0.4991,
+      "step": 122970
+    },
+    {
+      "epoch": 6.108075891526771,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0003113579020562233,
+      "loss": 0.5012,
+      "step": 122980
+    },
+    {
+      "epoch": 6.10857256382239,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00031131816827257376,
+      "loss": 0.4928,
+      "step": 122990
+    },
+    {
+      "epoch": 6.1090692361180094,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00031127843448892423,
+      "loss": 0.5107,
+      "step": 123000
+    },
+    {
+      "epoch": 6.109565908413629,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00031123870070527465,
+      "loss": 0.5096,
+      "step": 123010
+    },
+    {
+      "epoch": 6.110062580709248,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0003111989669216251,
+      "loss": 0.5067,
+      "step": 123020
+    },
+    {
+      "epoch": 6.110559253004867,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003111592331379756,
+      "loss": 0.4906,
+      "step": 123030
+    },
+    {
+      "epoch": 6.111055925300486,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00031111949935432606,
+      "loss": 0.5247,
+      "step": 123040
+    },
+    {
+      "epoch": 6.1115525975961065,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0003110797655706765,
+      "loss": 0.4742,
+      "step": 123050
+    },
+    {
+      "epoch": 6.112049269891726,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003110400317870269,
+      "loss": 0.5073,
+      "step": 123060
+    },
+    {
+      "epoch": 6.112545942187345,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003110002980033774,
+      "loss": 0.5044,
+      "step": 123070
+    },
+    {
+      "epoch": 6.113042614482964,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00031096056421972784,
+      "loss": 0.4577,
+      "step": 123080
+    },
+    {
+      "epoch": 6.113539286778583,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00031092083043607826,
+      "loss": 0.5072,
+      "step": 123090
+    },
+    {
+      "epoch": 6.114035959074203,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00031088109665242873,
+      "loss": 0.4861,
+      "step": 123100
+    },
+    {
+      "epoch": 6.114532631369822,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0003108413628687792,
+      "loss": 0.4976,
+      "step": 123110
+    },
+    {
+      "epoch": 6.115029303665442,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.00031080162908512967,
+      "loss": 0.5007,
+      "step": 123120
+    },
+    {
+      "epoch": 6.115525975961061,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003107618953014801,
+      "loss": 0.508,
+      "step": 123130
+    },
+    {
+      "epoch": 6.11602264825668,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00031072216151783056,
+      "loss": 0.4942,
+      "step": 123140
+    },
+    {
+      "epoch": 6.1165193205523,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00031068242773418103,
+      "loss": 0.4963,
+      "step": 123150
+    },
+    {
+      "epoch": 6.117015992847919,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00031064269395053145,
+      "loss": 0.5308,
+      "step": 123160
+    },
+    {
+      "epoch": 6.117512665143538,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00031060296016688187,
+      "loss": 0.513,
+      "step": 123170
+    },
+    {
+      "epoch": 6.118009337439157,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00031056322638323234,
+      "loss": 0.5011,
+      "step": 123180
+    },
+    {
+      "epoch": 6.118506009734777,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003105234925995828,
+      "loss": 0.4966,
+      "step": 123190
+    },
+    {
+      "epoch": 6.119002682030397,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003104837588159333,
+      "loss": 0.5091,
+      "step": 123200
+    },
+    {
+      "epoch": 6.119499354326016,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0003104440250322837,
+      "loss": 0.5286,
+      "step": 123210
+    },
+    {
+      "epoch": 6.119996026621635,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00031040429124863417,
+      "loss": 0.5122,
+      "step": 123220
+    },
+    {
+      "epoch": 6.120492698917254,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00031036455746498464,
+      "loss": 0.4795,
+      "step": 123230
+    },
+    {
+      "epoch": 6.120989371212874,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00031032482368133506,
+      "loss": 0.4663,
+      "step": 123240
+    },
+    {
+      "epoch": 6.121486043508493,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00031028508989768553,
+      "loss": 0.5169,
+      "step": 123250
+    },
+    {
+      "epoch": 6.121982715804112,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.000310245356114036,
+      "loss": 0.5197,
+      "step": 123260
+    },
+    {
+      "epoch": 6.122479388099732,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0003102056223303864,
+      "loss": 0.4805,
+      "step": 123270
+    },
+    {
+      "epoch": 6.122976060395351,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0003101658885467369,
+      "loss": 0.478,
+      "step": 123280
+    },
+    {
+      "epoch": 6.123472732690971,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003101261547630873,
+      "loss": 0.507,
+      "step": 123290
+    },
+    {
+      "epoch": 6.12396940498659,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0003100864209794378,
+      "loss": 0.4978,
+      "step": 123300
+    },
+    {
+      "epoch": 6.124466077282209,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00031004668719578825,
+      "loss": 0.4718,
+      "step": 123310
+    },
+    {
+      "epoch": 6.124962749577828,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00031000695341213867,
+      "loss": 0.4664,
+      "step": 123320
+    },
+    {
+      "epoch": 6.1254594218734475,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00030996721962848914,
+      "loss": 0.4814,
+      "step": 123330
+    },
+    {
+      "epoch": 6.125956094169068,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003099274858448396,
+      "loss": 0.5186,
+      "step": 123340
+    },
+    {
+      "epoch": 6.126452766464687,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00030988775206119,
+      "loss": 0.4886,
+      "step": 123350
+    },
+    {
+      "epoch": 6.126949438760306,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0003098480182775405,
+      "loss": 0.514,
+      "step": 123360
+    },
+    {
+      "epoch": 6.127446111055925,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00030980828449389097,
+      "loss": 0.4787,
+      "step": 123370
+    },
+    {
+      "epoch": 6.1279427833515445,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003097685507102414,
+      "loss": 0.4772,
+      "step": 123380
+    },
+    {
+      "epoch": 6.128439455647164,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00030972881692659186,
+      "loss": 0.4977,
+      "step": 123390
+    },
+    {
+      "epoch": 6.128936127942783,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0003096890831429423,
+      "loss": 0.4828,
+      "step": 123400
+    },
+    {
+      "epoch": 6.129432800238403,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003096493493592928,
+      "loss": 0.5129,
+      "step": 123410
+    },
+    {
+      "epoch": 6.129929472534022,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003096096155756432,
+      "loss": 0.4896,
+      "step": 123420
+    },
+    {
+      "epoch": 6.1304261448296415,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00030956988179199363,
+      "loss": 0.5021,
+      "step": 123430
+    },
+    {
+      "epoch": 6.130922817125261,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003095301480083441,
+      "loss": 0.5188,
+      "step": 123440
+    },
+    {
+      "epoch": 6.13141948942088,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003094904142246946,
+      "loss": 0.4963,
+      "step": 123450
+    },
+    {
+      "epoch": 6.131916161716499,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000309450680441045,
+      "loss": 0.4983,
+      "step": 123460
+    },
+    {
+      "epoch": 6.1324128340121185,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00030941094665739547,
+      "loss": 0.4857,
+      "step": 123470
+    },
+    {
+      "epoch": 6.132909506307739,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0003093712128737459,
+      "loss": 0.4902,
+      "step": 123480
+    },
+    {
+      "epoch": 6.133406178603358,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0003093314790900964,
+      "loss": 0.4763,
+      "step": 123490
+    },
+    {
+      "epoch": 6.133902850898977,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0003092917453064468,
+      "loss": 0.4937,
+      "step": 123500
+    },
+    {
+      "epoch": 6.134399523194596,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00030925201152279724,
+      "loss": 0.4926,
+      "step": 123510
+    },
+    {
+      "epoch": 6.1348961954902155,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003092122777391477,
+      "loss": 0.5156,
+      "step": 123520
+    },
+    {
+      "epoch": 6.135392867785835,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0003091725439554982,
+      "loss": 0.4528,
+      "step": 123530
+    },
+    {
+      "epoch": 6.135889540081454,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003091328101718486,
+      "loss": 0.4661,
+      "step": 123540
+    },
+    {
+      "epoch": 6.136386212377074,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0003090930763881991,
+      "loss": 0.4997,
+      "step": 123550
+    },
+    {
+      "epoch": 6.136882884672693,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00030905334260454954,
+      "loss": 0.5024,
+      "step": 123560
+    },
+    {
+      "epoch": 6.1373795569683125,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0003090136088209,
+      "loss": 0.5092,
+      "step": 123570
+    },
+    {
+      "epoch": 6.137876229263932,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00030897387503725043,
+      "loss": 0.462,
+      "step": 123580
+    },
+    {
+      "epoch": 6.138372901559551,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00030893414125360085,
+      "loss": 0.5407,
+      "step": 123590
+    },
+    {
+      "epoch": 6.13886957385517,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0003088944074699514,
+      "loss": 0.4835,
+      "step": 123600
+    },
+    {
+      "epoch": 6.139366246150789,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0003088546736863018,
+      "loss": 0.5255,
+      "step": 123610
+    },
+    {
+      "epoch": 6.139862918446409,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0003088149399026522,
+      "loss": 0.4959,
+      "step": 123620
+    },
+    {
+      "epoch": 6.140359590742029,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003087752061190027,
+      "loss": 0.4974,
+      "step": 123630
+    },
+    {
+      "epoch": 6.140856263037648,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00030873547233535315,
+      "loss": 0.4779,
+      "step": 123640
+    },
+    {
+      "epoch": 6.141352935333267,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0003086957385517036,
+      "loss": 0.51,
+      "step": 123650
+    },
+    {
+      "epoch": 6.141849607628886,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00030865600476805404,
+      "loss": 0.4807,
+      "step": 123660
+    },
+    {
+      "epoch": 6.142346279924506,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0003086162709844045,
+      "loss": 0.482,
+      "step": 123670
+    },
+    {
+      "epoch": 6.142842952220125,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.000308576537200755,
+      "loss": 0.4833,
+      "step": 123680
+    },
+    {
+      "epoch": 6.143339624515744,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0003085368034171054,
+      "loss": 0.4663,
+      "step": 123690
+    },
+    {
+      "epoch": 6.143836296811364,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00030849706963345587,
+      "loss": 0.5121,
+      "step": 123700
+    },
+    {
+      "epoch": 6.1443329691069835,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0003084573358498063,
+      "loss": 0.5233,
+      "step": 123710
+    },
+    {
+      "epoch": 6.144829641402603,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00030841760206615676,
+      "loss": 0.4788,
+      "step": 123720
+    },
+    {
+      "epoch": 6.145326313698222,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00030837786828250723,
+      "loss": 0.5001,
+      "step": 123730
+    },
+    {
+      "epoch": 6.145822985993841,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00030833813449885765,
+      "loss": 0.4717,
+      "step": 123740
+    },
+    {
+      "epoch": 6.14631965828946,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003082984007152081,
+      "loss": 0.4938,
+      "step": 123750
+    },
+    {
+      "epoch": 6.14681633058508,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003082586669315586,
+      "loss": 0.4876,
+      "step": 123760
+    },
+    {
+      "epoch": 6.1473130028807,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.000308218933147909,
+      "loss": 0.5093,
+      "step": 123770
+    },
+    {
+      "epoch": 6.147809675176319,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003081791993642595,
+      "loss": 0.4692,
+      "step": 123780
+    },
+    {
+      "epoch": 6.148306347471938,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00030813946558060995,
+      "loss": 0.5277,
+      "step": 123790
+    },
+    {
+      "epoch": 6.148803019767557,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00030809973179696037,
+      "loss": 0.4778,
+      "step": 123800
+    },
+    {
+      "epoch": 6.149299692063177,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00030805999801331084,
+      "loss": 0.4829,
+      "step": 123810
+    },
+    {
+      "epoch": 6.149796364358796,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00030802026422966126,
+      "loss": 0.5135,
+      "step": 123820
+    },
+    {
+      "epoch": 6.150293036654415,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00030798053044601173,
+      "loss": 0.4854,
+      "step": 123830
+    },
+    {
+      "epoch": 6.150789708950035,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003079407966623622,
+      "loss": 0.4766,
+      "step": 123840
+    },
+    {
+      "epoch": 6.151286381245654,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003079010628787126,
+      "loss": 0.4788,
+      "step": 123850
+    },
+    {
+      "epoch": 6.151783053541274,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0003078613290950631,
+      "loss": 0.4611,
+      "step": 123860
+    },
+    {
+      "epoch": 6.152279725836893,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00030782159531141356,
+      "loss": 0.4839,
+      "step": 123870
+    },
+    {
+      "epoch": 6.152776398132512,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.000307781861527764,
+      "loss": 0.4894,
+      "step": 123880
+    },
+    {
+      "epoch": 6.153273070428131,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00030774212774411445,
+      "loss": 0.5046,
+      "step": 123890
+    },
+    {
+      "epoch": 6.1537697427237505,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003077023939604649,
+      "loss": 0.4849,
+      "step": 123900
+    },
+    {
+      "epoch": 6.15426641501937,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00030766266017681534,
+      "loss": 0.4949,
+      "step": 123910
+    },
+    {
+      "epoch": 6.15476308731499,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003076229263931658,
+      "loss": 0.4976,
+      "step": 123920
+    },
+    {
+      "epoch": 6.155259759610609,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0003075831926095162,
+      "loss": 0.4847,
+      "step": 123930
+    },
+    {
+      "epoch": 6.155756431906228,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00030754345882586675,
+      "loss": 0.479,
+      "step": 123940
+    },
+    {
+      "epoch": 6.156253104201848,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00030750372504221717,
+      "loss": 0.5073,
+      "step": 123950
+    },
+    {
+      "epoch": 6.156749776497467,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003074639912585676,
+      "loss": 0.5041,
+      "step": 123960
+    },
+    {
+      "epoch": 6.157246448793086,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00030742425747491806,
+      "loss": 0.4983,
+      "step": 123970
+    },
+    {
+      "epoch": 6.157743121088705,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00030738452369126853,
+      "loss": 0.4878,
+      "step": 123980
+    },
+    {
+      "epoch": 6.158239793384325,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00030734478990761895,
+      "loss": 0.4951,
+      "step": 123990
+    },
+    {
+      "epoch": 6.158736465679945,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0003073050561239694,
+      "loss": 0.515,
+      "step": 124000
+    },
+    {
+      "epoch": 6.159233137975564,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00030726532234031983,
+      "loss": 0.4662,
+      "step": 124010
+    },
+    {
+      "epoch": 6.159729810271183,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00030722558855667036,
+      "loss": 0.4846,
+      "step": 124020
+    },
+    {
+      "epoch": 6.160226482566802,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0003071858547730208,
+      "loss": 0.4826,
+      "step": 124030
+    },
+    {
+      "epoch": 6.1607231548624215,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003071461209893712,
+      "loss": 0.5076,
+      "step": 124040
+    },
+    {
+      "epoch": 6.161219827158041,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00030710638720572167,
+      "loss": 0.4848,
+      "step": 124050
+    },
+    {
+      "epoch": 6.161716499453661,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00030706665342207214,
+      "loss": 0.4963,
+      "step": 124060
+    },
+    {
+      "epoch": 6.16221317174928,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0003070269196384226,
+      "loss": 0.4908,
+      "step": 124070
+    },
+    {
+      "epoch": 6.162709844044899,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.000306987185854773,
+      "loss": 0.5055,
+      "step": 124080
+    },
+    {
+      "epoch": 6.1632065163405185,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0003069474520711235,
+      "loss": 0.5082,
+      "step": 124090
+    },
+    {
+      "epoch": 6.163703188636138,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00030690771828747397,
+      "loss": 0.5017,
+      "step": 124100
+    },
+    {
+      "epoch": 6.164199860931757,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003068679845038244,
+      "loss": 0.5244,
+      "step": 124110
+    },
+    {
+      "epoch": 6.164696533227376,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003068282507201748,
+      "loss": 0.4966,
+      "step": 124120
+    },
+    {
+      "epoch": 6.165193205522996,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00030678851693652533,
+      "loss": 0.4702,
+      "step": 124130
+    },
+    {
+      "epoch": 6.1656898778186155,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00030674878315287575,
+      "loss": 0.4807,
+      "step": 124140
+    },
+    {
+      "epoch": 6.166186550114235,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003067090493692262,
+      "loss": 0.4972,
+      "step": 124150
+    },
+    {
+      "epoch": 6.166683222409854,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00030666931558557663,
+      "loss": 0.4697,
+      "step": 124160
+    },
+    {
+      "epoch": 6.167179894705473,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0003066295818019271,
+      "loss": 0.4771,
+      "step": 124170
+    },
+    {
+      "epoch": 6.1676765670010925,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0003065898480182776,
+      "loss": 0.4711,
+      "step": 124180
+    },
+    {
+      "epoch": 6.168173239296712,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.000306550114234628,
+      "loss": 0.4897,
+      "step": 124190
+    },
+    {
+      "epoch": 6.168669911592332,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00030651038045097846,
+      "loss": 0.4892,
+      "step": 124200
+    },
+    {
+      "epoch": 6.169166583887951,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00030647064666732894,
+      "loss": 0.5014,
+      "step": 124210
+    },
+    {
+      "epoch": 6.16966325618357,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00030643091288367935,
+      "loss": 0.5003,
+      "step": 124220
+    },
+    {
+      "epoch": 6.1701599284791895,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0003063911791000298,
+      "loss": 0.4752,
+      "step": 124230
+    },
+    {
+      "epoch": 6.170656600774809,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003063514453163803,
+      "loss": 0.5083,
+      "step": 124240
+    },
+    {
+      "epoch": 6.171153273070428,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0003063117115327307,
+      "loss": 0.487,
+      "step": 124250
+    },
+    {
+      "epoch": 6.171649945366047,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003062719777490812,
+      "loss": 0.4781,
+      "step": 124260
+    },
+    {
+      "epoch": 6.172146617661667,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003062322439654316,
+      "loss": 0.5144,
+      "step": 124270
+    },
+    {
+      "epoch": 6.1726432899572865,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0003061925101817821,
+      "loss": 0.4547,
+      "step": 124280
+    },
+    {
+      "epoch": 6.173139962252906,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00030615277639813254,
+      "loss": 0.5286,
+      "step": 124290
+    },
+    {
+      "epoch": 6.173636634548525,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00030611304261448296,
+      "loss": 0.4818,
+      "step": 124300
+    },
+    {
+      "epoch": 6.174133306844144,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00030607330883083343,
+      "loss": 0.4961,
+      "step": 124310
+    },
+    {
+      "epoch": 6.174629979139763,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0003060335750471839,
+      "loss": 0.4753,
+      "step": 124320
+    },
+    {
+      "epoch": 6.175126651435383,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0003059938412635343,
+      "loss": 0.4897,
+      "step": 124330
+    },
+    {
+      "epoch": 6.175623323731002,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0003059541074798848,
+      "loss": 0.4628,
+      "step": 124340
+    },
+    {
+      "epoch": 6.176119996026622,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003059143736962352,
+      "loss": 0.5122,
+      "step": 124350
+    },
+    {
+      "epoch": 6.176616668322241,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0003058746399125857,
+      "loss": 0.4718,
+      "step": 124360
+    },
+    {
+      "epoch": 6.17711334061786,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00030583490612893615,
+      "loss": 0.495,
+      "step": 124370
+    },
+    {
+      "epoch": 6.17761001291348,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00030579517234528657,
+      "loss": 0.4725,
+      "step": 124380
+    },
+    {
+      "epoch": 6.178106685209099,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00030575543856163704,
+      "loss": 0.4962,
+      "step": 124390
+    },
+    {
+      "epoch": 6.178603357504718,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003057157047779875,
+      "loss": 0.4841,
+      "step": 124400
+    },
+    {
+      "epoch": 6.179100029800337,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.00030567597099433793,
+      "loss": 0.4938,
+      "step": 124410
+    },
+    {
+      "epoch": 6.1795967020959575,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003056362372106884,
+      "loss": 0.4639,
+      "step": 124420
+    },
+    {
+      "epoch": 6.180093374391577,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00030559650342703887,
+      "loss": 0.4921,
+      "step": 124430
+    },
+    {
+      "epoch": 6.180590046687196,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003055567696433893,
+      "loss": 0.4631,
+      "step": 124440
+    },
+    {
+      "epoch": 6.181086718982815,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00030551703585973976,
+      "loss": 0.4965,
+      "step": 124450
+    },
+    {
+      "epoch": 6.181583391278434,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003054773020760902,
+      "loss": 0.4815,
+      "step": 124460
+    },
+    {
+      "epoch": 6.182080063574054,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003054375682924407,
+      "loss": 0.4847,
+      "step": 124470
+    },
+    {
+      "epoch": 6.182576735869673,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003053978345087911,
+      "loss": 0.4787,
+      "step": 124480
+    },
+    {
+      "epoch": 6.183073408165293,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00030535810072514154,
+      "loss": 0.4774,
+      "step": 124490
+    },
+    {
+      "epoch": 6.183570080460912,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.000305318366941492,
+      "loss": 0.4949,
+      "step": 124500
+    },
+    {
+      "epoch": 6.184066752756531,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003052786331578425,
+      "loss": 0.4976,
+      "step": 124510
+    },
+    {
+      "epoch": 6.184563425052151,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00030523889937419295,
+      "loss": 0.4635,
+      "step": 124520
+    },
+    {
+      "epoch": 6.18506009734777,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00030519916559054337,
+      "loss": 0.4866,
+      "step": 124530
+    },
+    {
+      "epoch": 6.185556769643389,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00030515943180689384,
+      "loss": 0.4859,
+      "step": 124540
+    },
+    {
+      "epoch": 6.186053441939008,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0003051196980232443,
+      "loss": 0.4943,
+      "step": 124550
+    },
+    {
+      "epoch": 6.186550114234628,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00030507996423959473,
+      "loss": 0.4914,
+      "step": 124560
+    },
+    {
+      "epoch": 6.187046786530248,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.00030504023045594515,
+      "loss": 0.5256,
+      "step": 124570
+    },
+    {
+      "epoch": 6.187543458825867,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003050004966722956,
+      "loss": 0.5115,
+      "step": 124580
+    },
+    {
+      "epoch": 6.188040131121486,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003049607628886461,
+      "loss": 0.5121,
+      "step": 124590
+    },
+    {
+      "epoch": 6.188536803417105,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00030492102910499656,
+      "loss": 0.4957,
+      "step": 124600
+    },
+    {
+      "epoch": 6.1890334757127246,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.000304881295321347,
+      "loss": 0.5048,
+      "step": 124610
+    },
+    {
+      "epoch": 6.189530148008344,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00030484156153769745,
+      "loss": 0.4986,
+      "step": 124620
+    },
+    {
+      "epoch": 6.190026820303963,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0003048018277540479,
+      "loss": 0.5007,
+      "step": 124630
+    },
+    {
+      "epoch": 6.190523492599583,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00030476209397039834,
+      "loss": 0.4803,
+      "step": 124640
+    },
+    {
+      "epoch": 6.191020164895202,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00030472236018674875,
+      "loss": 0.5185,
+      "step": 124650
+    },
+    {
+      "epoch": 6.191516837190822,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003046826264030993,
+      "loss": 0.4786,
+      "step": 124660
+    },
+    {
+      "epoch": 6.192013509486441,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0003046428926194497,
+      "loss": 0.513,
+      "step": 124670
+    },
+    {
+      "epoch": 6.19251018178206,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00030460315883580017,
+      "loss": 0.4815,
+      "step": 124680
+    },
+    {
+      "epoch": 6.193006854077679,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0003045634250521506,
+      "loss": 0.5076,
+      "step": 124690
+    },
+    {
+      "epoch": 6.1935035263732985,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00030452369126850106,
+      "loss": 0.4995,
+      "step": 124700
+    },
+    {
+      "epoch": 6.194000198668919,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00030448395748485153,
+      "loss": 0.4794,
+      "step": 124710
+    },
+    {
+      "epoch": 6.194496870964538,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00030444422370120195,
+      "loss": 0.5052,
+      "step": 124720
+    },
+    {
+      "epoch": 6.194993543260157,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0003044044899175524,
+      "loss": 0.4816,
+      "step": 124730
+    },
+    {
+      "epoch": 6.195490215555776,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003043647561339029,
+      "loss": 0.4861,
+      "step": 124740
+    },
+    {
+      "epoch": 6.1959868878513955,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003043250223502533,
+      "loss": 0.4902,
+      "step": 124750
+    },
+    {
+      "epoch": 6.196483560147015,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0003042852885666038,
+      "loss": 0.4454,
+      "step": 124760
+    },
+    {
+      "epoch": 6.196980232442634,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00030424555478295425,
+      "loss": 0.5016,
+      "step": 124770
+    },
+    {
+      "epoch": 6.197476904738254,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00030420582099930467,
+      "loss": 0.5039,
+      "step": 124780
+    },
+    {
+      "epoch": 6.197973577033873,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00030416608721565514,
+      "loss": 0.5065,
+      "step": 124790
+    },
+    {
+      "epoch": 6.1984702493294925,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00030412635343200555,
+      "loss": 0.4974,
+      "step": 124800
+    },
+    {
+      "epoch": 6.198966921625112,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.000304086619648356,
+      "loss": 0.4879,
+      "step": 124810
+    },
+    {
+      "epoch": 6.199463593920731,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003040468858647065,
+      "loss": 0.5344,
+      "step": 124820
+    },
+    {
+      "epoch": 6.19996026621635,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003040071520810569,
+      "loss": 0.4836,
+      "step": 124830
+    },
+    {
+      "epoch": 6.200456938511969,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003039674182974074,
+      "loss": 0.4812,
+      "step": 124840
+    },
+    {
+      "epoch": 6.2009536108075896,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00030392768451375786,
+      "loss": 0.4752,
+      "step": 124850
+    },
+    {
+      "epoch": 6.201450283103209,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003038879507301083,
+      "loss": 0.4966,
+      "step": 124860
+    },
+    {
+      "epoch": 6.201946955398828,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00030384821694645874,
+      "loss": 0.5055,
+      "step": 124870
+    },
+    {
+      "epoch": 6.202443627694447,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00030380848316280916,
+      "loss": 0.491,
+      "step": 124880
+    },
+    {
+      "epoch": 6.2029402999900665,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00030376874937915963,
+      "loss": 0.4983,
+      "step": 124890
+    },
+    {
+      "epoch": 6.203436972285686,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0003037290155955101,
+      "loss": 0.4944,
+      "step": 124900
+    },
+    {
+      "epoch": 6.203933644581305,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0003036892818118605,
+      "loss": 0.4925,
+      "step": 124910
+    },
+    {
+      "epoch": 6.204430316876925,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.000303649548028211,
+      "loss": 0.49,
+      "step": 124920
+    },
+    {
+      "epoch": 6.204926989172544,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00030360981424456146,
+      "loss": 0.4816,
+      "step": 124930
+    },
+    {
+      "epoch": 6.2054236614681635,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003035700804609119,
+      "loss": 0.4919,
+      "step": 124940
+    },
+    {
+      "epoch": 6.205920333763783,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00030353034667726235,
+      "loss": 0.4759,
+      "step": 124950
+    },
+    {
+      "epoch": 6.206417006059402,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0003034906128936128,
+      "loss": 0.5108,
+      "step": 124960
+    },
+    {
+      "epoch": 6.206913678355021,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0003034508791099633,
+      "loss": 0.4897,
+      "step": 124970
+    },
+    {
+      "epoch": 6.20741035065064,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0003034111453263137,
+      "loss": 0.4982,
+      "step": 124980
+    },
+    {
+      "epoch": 6.2079070229462605,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00030337141154266413,
+      "loss": 0.5029,
+      "step": 124990
+    },
+    {
+      "epoch": 6.20840369524188,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00030333167775901466,
+      "loss": 0.4894,
+      "step": 125000
+    },
+    {
+      "epoch": 6.208900367537499,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00030329194397536507,
+      "loss": 0.4799,
+      "step": 125010
+    },
+    {
+      "epoch": 6.209397039833118,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0003032522101917155,
+      "loss": 0.5078,
+      "step": 125020
+    },
+    {
+      "epoch": 6.209893712128737,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00030321247640806596,
+      "loss": 0.5166,
+      "step": 125030
+    },
+    {
+      "epoch": 6.210390384424357,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00030317274262441643,
+      "loss": 0.4705,
+      "step": 125040
+    },
+    {
+      "epoch": 6.210887056719976,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003031330088407669,
+      "loss": 0.4962,
+      "step": 125050
+    },
+    {
+      "epoch": 6.211383729015595,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0003030932750571173,
+      "loss": 0.4978,
+      "step": 125060
+    },
+    {
+      "epoch": 6.211880401311215,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003030535412734678,
+      "loss": 0.4915,
+      "step": 125070
+    },
+    {
+      "epoch": 6.212377073606834,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00030301380748981826,
+      "loss": 0.463,
+      "step": 125080
+    },
+    {
+      "epoch": 6.212873745902454,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0003029740737061687,
+      "loss": 0.4876,
+      "step": 125090
+    },
+    {
+      "epoch": 6.213370418198073,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0003029343399225191,
+      "loss": 0.5183,
+      "step": 125100
+    },
+    {
+      "epoch": 6.213867090493692,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0003028946061388696,
+      "loss": 0.5035,
+      "step": 125110
+    },
+    {
+      "epoch": 6.214363762789311,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00030285487235522004,
+      "loss": 0.5088,
+      "step": 125120
+    },
+    {
+      "epoch": 6.214860435084931,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003028151385715705,
+      "loss": 0.4658,
+      "step": 125130
+    },
+    {
+      "epoch": 6.215357107380551,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00030277540478792093,
+      "loss": 0.4993,
+      "step": 125140
+    },
+    {
+      "epoch": 6.21585377967617,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0003027356710042714,
+      "loss": 0.5285,
+      "step": 125150
+    },
+    {
+      "epoch": 6.216350451971789,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00030269593722062187,
+      "loss": 0.4791,
+      "step": 125160
+    },
+    {
+      "epoch": 6.216847124267408,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0003026562034369723,
+      "loss": 0.5056,
+      "step": 125170
+    },
+    {
+      "epoch": 6.217343796563028,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0003026164696533227,
+      "loss": 0.5113,
+      "step": 125180
+    },
+    {
+      "epoch": 6.217840468858647,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00030257673586967323,
+      "loss": 0.4772,
+      "step": 125190
+    },
+    {
+      "epoch": 6.218337141154266,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00030253700208602365,
+      "loss": 0.4725,
+      "step": 125200
+    },
+    {
+      "epoch": 6.218833813449886,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0003024972683023741,
+      "loss": 0.4514,
+      "step": 125210
+    },
+    {
+      "epoch": 6.219330485745505,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00030245753451872454,
+      "loss": 0.4987,
+      "step": 125220
+    },
+    {
+      "epoch": 6.219827158041125,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.000302417800735075,
+      "loss": 0.5278,
+      "step": 125230
+    },
+    {
+      "epoch": 6.220323830336744,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0003023780669514255,
+      "loss": 0.4924,
+      "step": 125240
+    },
+    {
+      "epoch": 6.220820502632363,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0003023383331677759,
+      "loss": 0.4994,
+      "step": 125250
+    },
+    {
+      "epoch": 6.221317174927982,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00030229859938412637,
+      "loss": 0.5085,
+      "step": 125260
+    },
+    {
+      "epoch": 6.2218138472236015,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00030225886560047684,
+      "loss": 0.5046,
+      "step": 125270
+    },
+    {
+      "epoch": 6.222310519519221,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00030221913181682726,
+      "loss": 0.5063,
+      "step": 125280
+    },
+    {
+      "epoch": 6.222807191814841,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00030217939803317773,
+      "loss": 0.5053,
+      "step": 125290
+    },
+    {
+      "epoch": 6.22330386411046,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003021396642495282,
+      "loss": 0.4956,
+      "step": 125300
+    },
+    {
+      "epoch": 6.223800536406079,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0003020999304658786,
+      "loss": 0.523,
+      "step": 125310
+    },
+    {
+      "epoch": 6.224297208701699,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003020601966822291,
+      "loss": 0.4867,
+      "step": 125320
+    },
+    {
+      "epoch": 6.224793880997318,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0003020204628985795,
+      "loss": 0.4856,
+      "step": 125330
+    },
+    {
+      "epoch": 6.225290553292937,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00030198072911493003,
+      "loss": 0.4829,
+      "step": 125340
+    },
+    {
+      "epoch": 6.225787225588556,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00030194099533128045,
+      "loss": 0.5264,
+      "step": 125350
+    },
+    {
+      "epoch": 6.226283897884176,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00030190126154763087,
+      "loss": 0.4902,
+      "step": 125360
+    },
+    {
+      "epoch": 6.226780570179796,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00030186152776398134,
+      "loss": 0.5243,
+      "step": 125370
+    },
+    {
+      "epoch": 6.227277242475415,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003018217939803318,
+      "loss": 0.4956,
+      "step": 125380
+    },
+    {
+      "epoch": 6.227773914771034,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003017820601966822,
+      "loss": 0.5035,
+      "step": 125390
+    },
+    {
+      "epoch": 6.228270587066653,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0003017423264130327,
+      "loss": 0.5026,
+      "step": 125400
+    },
+    {
+      "epoch": 6.2287672593622725,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0003017025926293831,
+      "loss": 0.4802,
+      "step": 125410
+    },
+    {
+      "epoch": 6.229263931657892,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00030166285884573364,
+      "loss": 0.5156,
+      "step": 125420
+    },
+    {
+      "epoch": 6.229760603953512,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00030162312506208406,
+      "loss": 0.4948,
+      "step": 125430
+    },
+    {
+      "epoch": 6.230257276249131,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003015833912784345,
+      "loss": 0.4874,
+      "step": 125440
+    },
+    {
+      "epoch": 6.23075394854475,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00030154365749478494,
+      "loss": 0.4725,
+      "step": 125450
+    },
+    {
+      "epoch": 6.2312506208403695,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0003015039237111354,
+      "loss": 0.4897,
+      "step": 125460
+    },
+    {
+      "epoch": 6.231747293135989,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00030146418992748583,
+      "loss": 0.5026,
+      "step": 125470
+    },
+    {
+      "epoch": 6.232243965431608,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0003014244561438363,
+      "loss": 0.4734,
+      "step": 125480
+    },
+    {
+      "epoch": 6.232740637727227,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0003013847223601868,
+      "loss": 0.4932,
+      "step": 125490
+    },
+    {
+      "epoch": 6.233237310022847,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00030134498857653725,
+      "loss": 0.4852,
+      "step": 125500
+    },
+    {
+      "epoch": 6.2337339823184665,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00030130525479288766,
+      "loss": 0.4757,
+      "step": 125510
+    },
+    {
+      "epoch": 6.234230654614086,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0003012655210092381,
+      "loss": 0.4989,
+      "step": 125520
+    },
+    {
+      "epoch": 6.234727326909705,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003012257872255886,
+      "loss": 0.4904,
+      "step": 125530
+    },
+    {
+      "epoch": 6.235223999205324,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.000301186053441939,
+      "loss": 0.5209,
+      "step": 125540
+    },
+    {
+      "epoch": 6.2357206715009434,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00030114631965828944,
+      "loss": 0.4679,
+      "step": 125550
+    },
+    {
+      "epoch": 6.236217343796563,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0003011065858746399,
+      "loss": 0.5565,
+      "step": 125560
+    },
+    {
+      "epoch": 6.236714016092183,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0003010668520909904,
+      "loss": 0.4529,
+      "step": 125570
+    },
+    {
+      "epoch": 6.237210688387802,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00030102711830734086,
+      "loss": 0.4735,
+      "step": 125580
+    },
+    {
+      "epoch": 6.237707360683421,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00030098738452369127,
+      "loss": 0.4933,
+      "step": 125590
+    },
+    {
+      "epoch": 6.2382040329790405,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00030094765074004174,
+      "loss": 0.4917,
+      "step": 125600
+    },
+    {
+      "epoch": 6.23870070527466,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0003009079169563922,
+      "loss": 0.4819,
+      "step": 125610
+    },
+    {
+      "epoch": 6.239197377570279,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00030086818317274263,
+      "loss": 0.4896,
+      "step": 125620
+    },
+    {
+      "epoch": 6.239694049865898,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00030082844938909305,
+      "loss": 0.5149,
+      "step": 125630
+    },
+    {
+      "epoch": 6.240190722161518,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0003007887156054436,
+      "loss": 0.5049,
+      "step": 125640
+    },
+    {
+      "epoch": 6.2406873944571375,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.000300748981821794,
+      "loss": 0.4964,
+      "step": 125650
+    },
+    {
+      "epoch": 6.241184066752757,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00030070924803814446,
+      "loss": 0.5195,
+      "step": 125660
+    },
+    {
+      "epoch": 6.241680739048376,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0003006695142544949,
+      "loss": 0.4992,
+      "step": 125670
+    },
+    {
+      "epoch": 6.242177411343995,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00030062978047084535,
+      "loss": 0.4925,
+      "step": 125680
+    },
+    {
+      "epoch": 6.242674083639614,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0003005900466871958,
+      "loss": 0.4941,
+      "step": 125690
+    },
+    {
+      "epoch": 6.243170755935234,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00030055031290354624,
+      "loss": 0.5038,
+      "step": 125700
+    },
+    {
+      "epoch": 6.243667428230853,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00030051057911989666,
+      "loss": 0.4964,
+      "step": 125710
+    },
+    {
+      "epoch": 6.244164100526473,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0003004708453362472,
+      "loss": 0.4944,
+      "step": 125720
+    },
+    {
+      "epoch": 6.244660772822092,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0003004311115525976,
+      "loss": 0.455,
+      "step": 125730
+    },
+    {
+      "epoch": 6.245157445117711,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00030039137776894807,
+      "loss": 0.502,
+      "step": 125740
+    },
+    {
+      "epoch": 6.245654117413331,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0003003516439852985,
+      "loss": 0.4861,
+      "step": 125750
+    },
+    {
+      "epoch": 6.24615078970895,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00030031191020164896,
+      "loss": 0.5074,
+      "step": 125760
+    },
+    {
+      "epoch": 6.246647462004569,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00030027217641799943,
+      "loss": 0.5044,
+      "step": 125770
+    },
+    {
+      "epoch": 6.247144134300188,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00030023244263434985,
+      "loss": 0.504,
+      "step": 125780
+    },
+    {
+      "epoch": 6.2476408065958084,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0003001927088507003,
+      "loss": 0.5104,
+      "step": 125790
+    },
+    {
+      "epoch": 6.248137478891428,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0003001529750670508,
+      "loss": 0.4601,
+      "step": 125800
+    },
+    {
+      "epoch": 6.248634151187047,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0003001132412834012,
+      "loss": 0.5146,
+      "step": 125810
+    },
+    {
+      "epoch": 6.249130823482666,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0003000735074997517,
+      "loss": 0.4963,
+      "step": 125820
+    },
+    {
+      "epoch": 6.249627495778285,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00030003377371610215,
+      "loss": 0.4802,
+      "step": 125830
+    },
+    {
+      "epoch": 6.250124168073905,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00029999403993245257,
+      "loss": 0.489,
+      "step": 125840
+    },
+    {
+      "epoch": 6.250620840369524,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00029995430614880304,
+      "loss": 0.4896,
+      "step": 125850
+    },
+    {
+      "epoch": 6.251117512665144,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00029991457236515346,
+      "loss": 0.4813,
+      "step": 125860
+    },
+    {
+      "epoch": 6.251614184960763,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.000299874838581504,
+      "loss": 0.5038,
+      "step": 125870
+    },
+    {
+      "epoch": 6.252110857256382,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002998351047978544,
+      "loss": 0.5025,
+      "step": 125880
+    },
+    {
+      "epoch": 6.252607529552002,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002997953710142048,
+      "loss": 0.4924,
+      "step": 125890
+    },
+    {
+      "epoch": 6.253104201847621,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002997556372305553,
+      "loss": 0.5108,
+      "step": 125900
+    },
+    {
+      "epoch": 6.25360087414324,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00029971590344690576,
+      "loss": 0.5069,
+      "step": 125910
+    },
+    {
+      "epoch": 6.254097546438859,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0002996761696632562,
+      "loss": 0.4908,
+      "step": 125920
+    },
+    {
+      "epoch": 6.254594218734479,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00029963643587960665,
+      "loss": 0.5153,
+      "step": 125930
+    },
+    {
+      "epoch": 6.255090891030099,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002995967020959571,
+      "loss": 0.4955,
+      "step": 125940
+    },
+    {
+      "epoch": 6.255587563325718,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002995569683123076,
+      "loss": 0.5118,
+      "step": 125950
+    },
+    {
+      "epoch": 6.256084235621337,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.000299517234528658,
+      "loss": 0.5032,
+      "step": 125960
+    },
+    {
+      "epoch": 6.256580907916956,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002994775007450084,
+      "loss": 0.5198,
+      "step": 125970
+    },
+    {
+      "epoch": 6.2570775802125755,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002994377669613589,
+      "loss": 0.4657,
+      "step": 125980
+    },
+    {
+      "epoch": 6.257574252508195,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00029939803317770937,
+      "loss": 0.5155,
+      "step": 125990
+    },
+    {
+      "epoch": 6.258070924803814,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0002993582993940598,
+      "loss": 0.4896,
+      "step": 126000
+    },
+    {
+      "epoch": 6.258567597099434,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00029931856561041026,
+      "loss": 0.4847,
+      "step": 126010
+    },
+    {
+      "epoch": 6.259064269395053,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00029927883182676073,
+      "loss": 0.5099,
+      "step": 126020
+    },
+    {
+      "epoch": 6.259560941690673,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002992390980431112,
+      "loss": 0.503,
+      "step": 126030
+    },
+    {
+      "epoch": 6.260057613986292,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002991993642594616,
+      "loss": 0.5044,
+      "step": 126040
+    },
+    {
+      "epoch": 6.260554286281911,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00029915963047581203,
+      "loss": 0.5047,
+      "step": 126050
+    },
+    {
+      "epoch": 6.26105095857753,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00029911989669216256,
+      "loss": 0.4773,
+      "step": 126060
+    },
+    {
+      "epoch": 6.2615476308731495,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000299080162908513,
+      "loss": 0.4588,
+      "step": 126070
+    },
+    {
+      "epoch": 6.26204430316877,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002990404291248634,
+      "loss": 0.4935,
+      "step": 126080
+    },
+    {
+      "epoch": 6.262540975464389,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00029900069534121386,
+      "loss": 0.5197,
+      "step": 126090
+    },
+    {
+      "epoch": 6.263037647760008,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00029896096155756434,
+      "loss": 0.4806,
+      "step": 126100
+    },
+    {
+      "epoch": 6.263534320055627,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002989212277739148,
+      "loss": 0.4985,
+      "step": 126110
+    },
+    {
+      "epoch": 6.2640309923512465,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002988814939902652,
+      "loss": 0.4776,
+      "step": 126120
+    },
+    {
+      "epoch": 6.264527664646866,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002988417602066157,
+      "loss": 0.4892,
+      "step": 126130
+    },
+    {
+      "epoch": 6.265024336942485,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00029880202642296617,
+      "loss": 0.5194,
+      "step": 126140
+    },
+    {
+      "epoch": 6.265521009238105,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002987622926393166,
+      "loss": 0.5437,
+      "step": 126150
+    },
+    {
+      "epoch": 6.266017681533724,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00029872255885566706,
+      "loss": 0.48,
+      "step": 126160
+    },
+    {
+      "epoch": 6.2665143538293435,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00029868282507201753,
+      "loss": 0.4789,
+      "step": 126170
+    },
+    {
+      "epoch": 6.267011026124963,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00029864309128836794,
+      "loss": 0.4868,
+      "step": 126180
+    },
+    {
+      "epoch": 6.267507698420582,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002986033575047184,
+      "loss": 0.5062,
+      "step": 126190
+    },
+    {
+      "epoch": 6.268004370716201,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00029856362372106883,
+      "loss": 0.4907,
+      "step": 126200
+    },
+    {
+      "epoch": 6.26850104301182,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002985238899374193,
+      "loss": 0.5097,
+      "step": 126210
+    },
+    {
+      "epoch": 6.2689977153074405,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002984841561537698,
+      "loss": 0.4833,
+      "step": 126220
+    },
+    {
+      "epoch": 6.26949438760306,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0002984444223701202,
+      "loss": 0.4974,
+      "step": 126230
+    },
+    {
+      "epoch": 6.269991059898679,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00029840468858647066,
+      "loss": 0.5146,
+      "step": 126240
+    },
+    {
+      "epoch": 6.270487732194298,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00029836495480282114,
+      "loss": 0.4795,
+      "step": 126250
+    },
+    {
+      "epoch": 6.2709844044899175,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00029832522101917155,
+      "loss": 0.4976,
+      "step": 126260
+    },
+    {
+      "epoch": 6.271481076785537,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.000298285487235522,
+      "loss": 0.5039,
+      "step": 126270
+    },
+    {
+      "epoch": 6.271977749081156,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00029824575345187244,
+      "loss": 0.488,
+      "step": 126280
+    },
+    {
+      "epoch": 6.272474421376776,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0002982060196682229,
+      "loss": 0.4881,
+      "step": 126290
+    },
+    {
+      "epoch": 6.272971093672395,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002981662858845734,
+      "loss": 0.52,
+      "step": 126300
+    },
+    {
+      "epoch": 6.2734677659680145,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0002981265521009238,
+      "loss": 0.5048,
+      "step": 126310
+    },
+    {
+      "epoch": 6.273964438263634,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00029808681831727427,
+      "loss": 0.4607,
+      "step": 126320
+    },
+    {
+      "epoch": 6.274461110559253,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00029804708453362474,
+      "loss": 0.5297,
+      "step": 126330
+    },
+    {
+      "epoch": 6.274957782854872,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00029800735074997516,
+      "loss": 0.5118,
+      "step": 126340
+    },
+    {
+      "epoch": 6.275454455150491,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00029796761696632563,
+      "loss": 0.4898,
+      "step": 126350
+    },
+    {
+      "epoch": 6.2759511274461115,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002979278831826761,
+      "loss": 0.4915,
+      "step": 126360
+    },
+    {
+      "epoch": 6.276447799741731,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0002978881493990265,
+      "loss": 0.4895,
+      "step": 126370
+    },
+    {
+      "epoch": 6.27694447203735,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.000297848415615377,
+      "loss": 0.5334,
+      "step": 126380
+    },
+    {
+      "epoch": 6.277441144332969,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002978086818317274,
+      "loss": 0.4788,
+      "step": 126390
+    },
+    {
+      "epoch": 6.277937816628588,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00029776894804807793,
+      "loss": 0.4646,
+      "step": 126400
+    },
+    {
+      "epoch": 6.278434488924208,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00029772921426442835,
+      "loss": 0.5157,
+      "step": 126410
+    },
+    {
+      "epoch": 6.278931161219827,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00029768948048077877,
+      "loss": 0.4983,
+      "step": 126420
+    },
+    {
+      "epoch": 6.279427833515447,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00029764974669712924,
+      "loss": 0.4527,
+      "step": 126430
+    },
+    {
+      "epoch": 6.279924505811066,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0002976100129134797,
+      "loss": 0.5077,
+      "step": 126440
+    },
+    {
+      "epoch": 6.280421178106685,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00029757027912983013,
+      "loss": 0.5107,
+      "step": 126450
+    },
+    {
+      "epoch": 6.280917850402305,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002975305453461806,
+      "loss": 0.4965,
+      "step": 126460
+    },
+    {
+      "epoch": 6.281414522697924,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00029749081156253107,
+      "loss": 0.4856,
+      "step": 126470
+    },
+    {
+      "epoch": 6.281911194993543,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00029745107777888154,
+      "loss": 0.5022,
+      "step": 126480
+    },
+    {
+      "epoch": 6.282407867289162,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00029741134399523196,
+      "loss": 0.4864,
+      "step": 126490
+    },
+    {
+      "epoch": 6.282904539584782,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002973716102115824,
+      "loss": 0.5043,
+      "step": 126500
+    },
+    {
+      "epoch": 6.283401211880402,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002973318764279329,
+      "loss": 0.4887,
+      "step": 126510
+    },
+    {
+      "epoch": 6.283897884176021,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002972921426442833,
+      "loss": 0.4909,
+      "step": 126520
+    },
+    {
+      "epoch": 6.28439455647164,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00029725240886063374,
+      "loss": 0.4986,
+      "step": 126530
+    },
+    {
+      "epoch": 6.284891228767259,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0002972126750769842,
+      "loss": 0.4875,
+      "step": 126540
+    },
+    {
+      "epoch": 6.285387901062879,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002971729412933347,
+      "loss": 0.4904,
+      "step": 126550
+    },
+    {
+      "epoch": 6.285884573358498,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00029713320750968515,
+      "loss": 0.5244,
+      "step": 126560
+    },
+    {
+      "epoch": 6.286381245654117,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00029709347372603557,
+      "loss": 0.4744,
+      "step": 126570
+    },
+    {
+      "epoch": 6.286877917949737,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.000297053739942386,
+      "loss": 0.4744,
+      "step": 126580
+    },
+    {
+      "epoch": 6.287374590245356,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0002970140061587365,
+      "loss": 0.4994,
+      "step": 126590
+    },
+    {
+      "epoch": 6.287871262540976,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00029697427237508693,
+      "loss": 0.4812,
+      "step": 126600
+    },
+    {
+      "epoch": 6.288367934836595,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0002969345385914374,
+      "loss": 0.488,
+      "step": 126610
+    },
+    {
+      "epoch": 6.288864607132214,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002968948048077878,
+      "loss": 0.5159,
+      "step": 126620
+    },
+    {
+      "epoch": 6.289361279427833,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0002968550710241383,
+      "loss": 0.4736,
+      "step": 126630
+    },
+    {
+      "epoch": 6.2898579517234525,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00029681533724048876,
+      "loss": 0.4954,
+      "step": 126640
+    },
+    {
+      "epoch": 6.290354624019072,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002967756034568392,
+      "loss": 0.5119,
+      "step": 126650
+    },
+    {
+      "epoch": 6.290851296314692,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00029673586967318965,
+      "loss": 0.5072,
+      "step": 126660
+    },
+    {
+      "epoch": 6.291347968610311,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002966961358895401,
+      "loss": 0.4805,
+      "step": 126670
+    },
+    {
+      "epoch": 6.29184464090593,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00029665640210589054,
+      "loss": 0.4895,
+      "step": 126680
+    },
+    {
+      "epoch": 6.2923413132015495,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.000296616668322241,
+      "loss": 0.5036,
+      "step": 126690
+    },
+    {
+      "epoch": 6.292837985497169,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0002965769345385915,
+      "loss": 0.4635,
+      "step": 126700
+    },
+    {
+      "epoch": 6.293334657792788,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002965372007549419,
+      "loss": 0.5062,
+      "step": 126710
+    },
+    {
+      "epoch": 6.293831330088407,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00029649746697129237,
+      "loss": 0.4842,
+      "step": 126720
+    },
+    {
+      "epoch": 6.294328002384027,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002964577331876428,
+      "loss": 0.5248,
+      "step": 126730
+    },
+    {
+      "epoch": 6.294824674679647,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00029641799940399326,
+      "loss": 0.4795,
+      "step": 126740
+    },
+    {
+      "epoch": 6.295321346975266,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00029637826562034373,
+      "loss": 0.5147,
+      "step": 126750
+    },
+    {
+      "epoch": 6.295818019270885,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00029633853183669414,
+      "loss": 0.4838,
+      "step": 126760
+    },
+    {
+      "epoch": 6.296314691566504,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002962987980530446,
+      "loss": 0.5255,
+      "step": 126770
+    },
+    {
+      "epoch": 6.2968113638621235,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002962590642693951,
+      "loss": 0.5057,
+      "step": 126780
+    },
+    {
+      "epoch": 6.297308036157743,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002962193304857455,
+      "loss": 0.496,
+      "step": 126790
+    },
+    {
+      "epoch": 6.297804708453363,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.000296179596702096,
+      "loss": 0.4741,
+      "step": 126800
+    },
+    {
+      "epoch": 6.298301380748982,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00029613986291844645,
+      "loss": 0.4893,
+      "step": 126810
+    },
+    {
+      "epoch": 6.298798053044601,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00029610012913479686,
+      "loss": 0.483,
+      "step": 126820
+    },
+    {
+      "epoch": 6.2992947253402205,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00029606039535114734,
+      "loss": 0.5172,
+      "step": 126830
+    },
+    {
+      "epoch": 6.29979139763584,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00029602066156749775,
+      "loss": 0.5366,
+      "step": 126840
+    },
+    {
+      "epoch": 6.300288069931459,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0002959809277838482,
+      "loss": 0.5062,
+      "step": 126850
+    },
+    {
+      "epoch": 6.300784742227078,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0002959411940001987,
+      "loss": 0.4965,
+      "step": 126860
+    },
+    {
+      "epoch": 6.301281414522698,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0002959014602165491,
+      "loss": 0.4638,
+      "step": 126870
+    },
+    {
+      "epoch": 6.3017780868183175,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0002958617264328996,
+      "loss": 0.4834,
+      "step": 126880
+    },
+    {
+      "epoch": 6.302274759113937,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00029582199264925006,
+      "loss": 0.4711,
+      "step": 126890
+    },
+    {
+      "epoch": 6.302771431409556,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00029578225886560047,
+      "loss": 0.5093,
+      "step": 126900
+    },
+    {
+      "epoch": 6.303268103705175,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00029574252508195094,
+      "loss": 0.496,
+      "step": 126910
+    },
+    {
+      "epoch": 6.303764776000794,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00029570279129830136,
+      "loss": 0.5124,
+      "step": 126920
+    },
+    {
+      "epoch": 6.304261448296414,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002956630575146519,
+      "loss": 0.4961,
+      "step": 126930
+    },
+    {
+      "epoch": 6.304758120592034,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0002956233237310023,
+      "loss": 0.4907,
+      "step": 126940
+    },
+    {
+      "epoch": 6.305254792887653,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0002955835899473527,
+      "loss": 0.4881,
+      "step": 126950
+    },
+    {
+      "epoch": 6.305751465183272,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002955438561637032,
+      "loss": 0.4682,
+      "step": 126960
+    },
+    {
+      "epoch": 6.3062481374788915,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00029550412238005366,
+      "loss": 0.5269,
+      "step": 126970
+    },
+    {
+      "epoch": 6.306744809774511,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002954643885964041,
+      "loss": 0.513,
+      "step": 126980
+    },
+    {
+      "epoch": 6.30724148207013,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00029542465481275455,
+      "loss": 0.4934,
+      "step": 126990
+    },
+    {
+      "epoch": 6.307738154365749,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.000295384921029105,
+      "loss": 0.4866,
+      "step": 127000
+    },
+    {
+      "epoch": 6.308234826661369,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002953451872454555,
+      "loss": 0.4747,
+      "step": 127010
+    },
+    {
+      "epoch": 6.3087314989569885,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0002953054534618059,
+      "loss": 0.5089,
+      "step": 127020
+    },
+    {
+      "epoch": 6.309228171252608,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00029526571967815633,
+      "loss": 0.5182,
+      "step": 127030
+    },
+    {
+      "epoch": 6.309724843548227,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00029522598589450685,
+      "loss": 0.4992,
+      "step": 127040
+    },
+    {
+      "epoch": 6.310221515843846,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00029518625211085727,
+      "loss": 0.4945,
+      "step": 127050
+    },
+    {
+      "epoch": 6.310718188139465,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00029514651832720774,
+      "loss": 0.4966,
+      "step": 127060
+    },
+    {
+      "epoch": 6.311214860435085,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00029510678454355816,
+      "loss": 0.5039,
+      "step": 127070
+    },
+    {
+      "epoch": 6.311711532730705,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00029506705075990863,
+      "loss": 0.501,
+      "step": 127080
+    },
+    {
+      "epoch": 6.312208205026324,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002950273169762591,
+      "loss": 0.5039,
+      "step": 127090
+    },
+    {
+      "epoch": 6.312704877321943,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002949875831926095,
+      "loss": 0.4992,
+      "step": 127100
+    },
+    {
+      "epoch": 6.313201549617562,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00029494784940895994,
+      "loss": 0.5168,
+      "step": 127110
+    },
+    {
+      "epoch": 6.313698221913182,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00029490811562531046,
+      "loss": 0.5004,
+      "step": 127120
+    },
+    {
+      "epoch": 6.314194894208801,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002948683818416609,
+      "loss": 0.4948,
+      "step": 127130
+    },
+    {
+      "epoch": 6.31469156650442,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00029482864805801135,
+      "loss": 0.4946,
+      "step": 127140
+    },
+    {
+      "epoch": 6.315188238800039,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00029478891427436177,
+      "loss": 0.5133,
+      "step": 127150
+    },
+    {
+      "epoch": 6.315684911095659,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00029474918049071224,
+      "loss": 0.4776,
+      "step": 127160
+    },
+    {
+      "epoch": 6.316181583391279,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0002947094467070627,
+      "loss": 0.5288,
+      "step": 127170
+    },
+    {
+      "epoch": 6.316678255686898,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00029466971292341313,
+      "loss": 0.4827,
+      "step": 127180
+    },
+    {
+      "epoch": 6.317174927982517,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002946299791397636,
+      "loss": 0.504,
+      "step": 127190
+    },
+    {
+      "epoch": 6.317671600278136,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00029459024535611407,
+      "loss": 0.5046,
+      "step": 127200
+    },
+    {
+      "epoch": 6.318168272573756,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002945505115724645,
+      "loss": 0.4915,
+      "step": 127210
+    },
+    {
+      "epoch": 6.318664944869375,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00029451077778881496,
+      "loss": 0.5133,
+      "step": 127220
+    },
+    {
+      "epoch": 6.319161617164995,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00029447104400516543,
+      "loss": 0.5189,
+      "step": 127230
+    },
+    {
+      "epoch": 6.319658289460614,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00029443131022151585,
+      "loss": 0.5,
+      "step": 127240
+    },
+    {
+      "epoch": 6.320154961756233,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002943915764378663,
+      "loss": 0.4652,
+      "step": 127250
+    },
+    {
+      "epoch": 6.320651634051853,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00029435184265421674,
+      "loss": 0.4974,
+      "step": 127260
+    },
+    {
+      "epoch": 6.321148306347472,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002943121088705672,
+      "loss": 0.4879,
+      "step": 127270
+    },
+    {
+      "epoch": 6.321644978643091,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002942723750869177,
+      "loss": 0.5183,
+      "step": 127280
+    },
+    {
+      "epoch": 6.32214165093871,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002942326413032681,
+      "loss": 0.4927,
+      "step": 127290
+    },
+    {
+      "epoch": 6.32263832323433,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00029419290751961857,
+      "loss": 0.5134,
+      "step": 127300
+    },
+    {
+      "epoch": 6.32313499552995,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00029415317373596904,
+      "loss": 0.4883,
+      "step": 127310
+    },
+    {
+      "epoch": 6.323631667825569,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00029411343995231946,
+      "loss": 0.4833,
+      "step": 127320
+    },
+    {
+      "epoch": 6.324128340121188,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00029407370616866993,
+      "loss": 0.504,
+      "step": 127330
+    },
+    {
+      "epoch": 6.324625012416807,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0002940339723850204,
+      "loss": 0.5042,
+      "step": 127340
+    },
+    {
+      "epoch": 6.3251216847124265,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002939942386013708,
+      "loss": 0.5016,
+      "step": 127350
+    },
+    {
+      "epoch": 6.325618357008046,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002939545048177213,
+      "loss": 0.4883,
+      "step": 127360
+    },
+    {
+      "epoch": 6.326115029303665,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002939147710340717,
+      "loss": 0.481,
+      "step": 127370
+    },
+    {
+      "epoch": 6.326611701599285,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002938750372504222,
+      "loss": 0.5074,
+      "step": 127380
+    },
+    {
+      "epoch": 6.327108373894904,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.00029383530346677265,
+      "loss": 0.5119,
+      "step": 127390
+    },
+    {
+      "epoch": 6.3276050461905236,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00029379556968312306,
+      "loss": 0.4492,
+      "step": 127400
+    },
+    {
+      "epoch": 6.328101718486143,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00029375583589947354,
+      "loss": 0.4809,
+      "step": 127410
+    },
+    {
+      "epoch": 6.328598390781762,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.000293716102115824,
+      "loss": 0.4887,
+      "step": 127420
+    },
+    {
+      "epoch": 6.329095063077381,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002936763683321745,
+      "loss": 0.5163,
+      "step": 127430
+    },
+    {
+      "epoch": 6.3295917353730005,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002936366345485249,
+      "loss": 0.5008,
+      "step": 127440
+    },
+    {
+      "epoch": 6.330088407668621,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002935969007648753,
+      "loss": 0.5162,
+      "step": 127450
+    },
+    {
+      "epoch": 6.33058507996424,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00029355716698122584,
+      "loss": 0.4824,
+      "step": 127460
+    },
+    {
+      "epoch": 6.331081752259859,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00029351743319757626,
+      "loss": 0.4889,
+      "step": 127470
+    },
+    {
+      "epoch": 6.331578424555478,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00029347769941392667,
+      "loss": 0.4926,
+      "step": 127480
+    },
+    {
+      "epoch": 6.3320750968510975,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00029343796563027714,
+      "loss": 0.4883,
+      "step": 127490
+    },
+    {
+      "epoch": 6.332571769146717,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002933982318466276,
+      "loss": 0.4851,
+      "step": 127500
+    },
+    {
+      "epoch": 6.333068441442336,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0002933584980629781,
+      "loss": 0.5089,
+      "step": 127510
+    },
+    {
+      "epoch": 6.333565113737956,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002933187642793285,
+      "loss": 0.4786,
+      "step": 127520
+    },
+    {
+      "epoch": 6.334061786033575,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.000293279030495679,
+      "loss": 0.5079,
+      "step": 127530
+    },
+    {
+      "epoch": 6.3345584583291945,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00029323929671202945,
+      "loss": 0.4996,
+      "step": 127540
+    },
+    {
+      "epoch": 6.335055130624814,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00029319956292837986,
+      "loss": 0.4921,
+      "step": 127550
+    },
+    {
+      "epoch": 6.335551802920433,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002931598291447303,
+      "loss": 0.5025,
+      "step": 127560
+    },
+    {
+      "epoch": 6.336048475216052,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002931200953610808,
+      "loss": 0.4897,
+      "step": 127570
+    },
+    {
+      "epoch": 6.336545147511671,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002930803615774312,
+      "loss": 0.481,
+      "step": 127580
+    },
+    {
+      "epoch": 6.3370418198072915,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002930406277937817,
+      "loss": 0.5139,
+      "step": 127590
+    },
+    {
+      "epoch": 6.337538492102911,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0002930008940101321,
+      "loss": 0.4962,
+      "step": 127600
+    },
+    {
+      "epoch": 6.33803516439853,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002929611602264826,
+      "loss": 0.5171,
+      "step": 127610
+    },
+    {
+      "epoch": 6.338531836694149,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00029292142644283305,
+      "loss": 0.4813,
+      "step": 127620
+    },
+    {
+      "epoch": 6.339028508989768,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00029288169265918347,
+      "loss": 0.4879,
+      "step": 127630
+    },
+    {
+      "epoch": 6.339525181285388,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00029284195887553394,
+      "loss": 0.4723,
+      "step": 127640
+    },
+    {
+      "epoch": 6.340021853581007,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002928022250918844,
+      "loss": 0.4918,
+      "step": 127650
+    },
+    {
+      "epoch": 6.340518525876627,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00029276249130823483,
+      "loss": 0.4824,
+      "step": 127660
+    },
+    {
+      "epoch": 6.341015198172246,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0002927227575245853,
+      "loss": 0.4864,
+      "step": 127670
+    },
+    {
+      "epoch": 6.3415118704678655,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002926830237409357,
+      "loss": 0.5111,
+      "step": 127680
+    },
+    {
+      "epoch": 6.342008542763485,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002926432899572862,
+      "loss": 0.4874,
+      "step": 127690
+    },
+    {
+      "epoch": 6.342505215059104,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00029260355617363666,
+      "loss": 0.4807,
+      "step": 127700
+    },
+    {
+      "epoch": 6.343001887354723,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002925638223899871,
+      "loss": 0.504,
+      "step": 127710
+    },
+    {
+      "epoch": 6.343498559650342,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00029252408860633755,
+      "loss": 0.511,
+      "step": 127720
+    },
+    {
+      "epoch": 6.3439952319459625,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.000292484354822688,
+      "loss": 0.4594,
+      "step": 127730
+    },
+    {
+      "epoch": 6.344491904241582,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00029244462103903844,
+      "loss": 0.501,
+      "step": 127740
+    },
+    {
+      "epoch": 6.344988576537201,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0002924048872553889,
+      "loss": 0.504,
+      "step": 127750
+    },
+    {
+      "epoch": 6.34548524883282,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002923651534717394,
+      "loss": 0.4805,
+      "step": 127760
+    },
+    {
+      "epoch": 6.345981921128439,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002923254196880898,
+      "loss": 0.5137,
+      "step": 127770
+    },
+    {
+      "epoch": 6.346478593424059,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00029228568590444027,
+      "loss": 0.4669,
+      "step": 127780
+    },
+    {
+      "epoch": 6.346975265719678,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002922459521207907,
+      "loss": 0.4747,
+      "step": 127790
+    },
+    {
+      "epoch": 6.347471938015298,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00029220621833714116,
+      "loss": 0.5053,
+      "step": 127800
+    },
+    {
+      "epoch": 6.347968610310917,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00029216648455349163,
+      "loss": 0.4982,
+      "step": 127810
+    },
+    {
+      "epoch": 6.348465282606536,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00029212675076984205,
+      "loss": 0.505,
+      "step": 127820
+    },
+    {
+      "epoch": 6.348961954902156,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002920870169861925,
+      "loss": 0.5074,
+      "step": 127830
+    },
+    {
+      "epoch": 6.349458627197775,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.000292047283202543,
+      "loss": 0.4657,
+      "step": 127840
+    },
+    {
+      "epoch": 6.349955299493394,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002920075494188934,
+      "loss": 0.5303,
+      "step": 127850
+    },
+    {
+      "epoch": 6.350451971789013,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002919678156352439,
+      "loss": 0.4858,
+      "step": 127860
+    },
+    {
+      "epoch": 6.3509486440846326,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00029192808185159435,
+      "loss": 0.4904,
+      "step": 127870
+    },
+    {
+      "epoch": 6.351445316380253,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0002918883480679448,
+      "loss": 0.4802,
+      "step": 127880
+    },
+    {
+      "epoch": 6.351941988675872,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00029184861428429524,
+      "loss": 0.4675,
+      "step": 127890
+    },
+    {
+      "epoch": 6.352438660971491,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00029180888050064566,
+      "loss": 0.509,
+      "step": 127900
+    },
+    {
+      "epoch": 6.35293533326711,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002917691467169962,
+      "loss": 0.5031,
+      "step": 127910
+    },
+    {
+      "epoch": 6.35343200556273,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002917294129333466,
+      "loss": 0.4812,
+      "step": 127920
+    },
+    {
+      "epoch": 6.353928677858349,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.000291689679149697,
+      "loss": 0.4554,
+      "step": 127930
+    },
+    {
+      "epoch": 6.354425350153968,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002916499453660475,
+      "loss": 0.5126,
+      "step": 127940
+    },
+    {
+      "epoch": 6.354922022449588,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00029161021158239796,
+      "loss": 0.4959,
+      "step": 127950
+    },
+    {
+      "epoch": 6.355418694745207,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00029157047779874843,
+      "loss": 0.5187,
+      "step": 127960
+    },
+    {
+      "epoch": 6.355915367040827,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00029153074401509885,
+      "loss": 0.505,
+      "step": 127970
+    },
+    {
+      "epoch": 6.356412039336446,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00029149101023144926,
+      "loss": 0.4917,
+      "step": 127980
+    },
+    {
+      "epoch": 6.356908711632065,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002914512764477998,
+      "loss": 0.467,
+      "step": 127990
+    },
+    {
+      "epoch": 6.357405383927684,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002914115426641502,
+      "loss": 0.4777,
+      "step": 128000
+    },
+    {
+      "epoch": 6.3579020562233035,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0002913718088805006,
+      "loss": 0.4678,
+      "step": 128010
+    },
+    {
+      "epoch": 6.358398728518924,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002913320750968511,
+      "loss": 0.522,
+      "step": 128020
+    },
+    {
+      "epoch": 6.358895400814543,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00029129234131320157,
+      "loss": 0.4893,
+      "step": 128030
+    },
+    {
+      "epoch": 6.359392073110162,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00029125260752955204,
+      "loss": 0.5169,
+      "step": 128040
+    },
+    {
+      "epoch": 6.359888745405781,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00029121287374590246,
+      "loss": 0.4913,
+      "step": 128050
+    },
+    {
+      "epoch": 6.3603854177014005,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00029117313996225293,
+      "loss": 0.4992,
+      "step": 128060
+    },
+    {
+      "epoch": 6.36088208999702,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002911334061786034,
+      "loss": 0.5027,
+      "step": 128070
+    },
+    {
+      "epoch": 6.361378762292639,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0002910936723949538,
+      "loss": 0.4832,
+      "step": 128080
+    },
+    {
+      "epoch": 6.361875434588258,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00029105393861130423,
+      "loss": 0.4816,
+      "step": 128090
+    },
+    {
+      "epoch": 6.362372106883878,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00029101420482765476,
+      "loss": 0.4841,
+      "step": 128100
+    },
+    {
+      "epoch": 6.362868779179498,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002909744710440052,
+      "loss": 0.4955,
+      "step": 128110
+    },
+    {
+      "epoch": 6.363365451475117,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00029093473726035565,
+      "loss": 0.5074,
+      "step": 128120
+    },
+    {
+      "epoch": 6.363862123770736,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00029089500347670606,
+      "loss": 0.4901,
+      "step": 128130
+    },
+    {
+      "epoch": 6.364358796066355,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00029085526969305654,
+      "loss": 0.509,
+      "step": 128140
+    },
+    {
+      "epoch": 6.3648554683619745,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000290815535909407,
+      "loss": 0.4937,
+      "step": 128150
+    },
+    {
+      "epoch": 6.365352140657594,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002907758021257574,
+      "loss": 0.5029,
+      "step": 128160
+    },
+    {
+      "epoch": 6.365848812953214,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0002907360683421079,
+      "loss": 0.4996,
+      "step": 128170
+    },
+    {
+      "epoch": 6.366345485248833,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00029069633455845837,
+      "loss": 0.4869,
+      "step": 128180
+    },
+    {
+      "epoch": 6.366842157544452,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0002906566007748088,
+      "loss": 0.5017,
+      "step": 128190
+    },
+    {
+      "epoch": 6.3673388298400715,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.00029061686699115926,
+      "loss": 0.4763,
+      "step": 128200
+    },
+    {
+      "epoch": 6.367835502135691,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0002905771332075097,
+      "loss": 0.5134,
+      "step": 128210
+    },
+    {
+      "epoch": 6.36833217443131,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00029053739942386014,
+      "loss": 0.4816,
+      "step": 128220
+    },
+    {
+      "epoch": 6.368828846726929,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002904976656402106,
+      "loss": 0.4838,
+      "step": 128230
+    },
+    {
+      "epoch": 6.369325519022549,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00029045793185656103,
+      "loss": 0.4561,
+      "step": 128240
+    },
+    {
+      "epoch": 6.3698221913181685,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002904181980729115,
+      "loss": 0.5005,
+      "step": 128250
+    },
+    {
+      "epoch": 6.370318863613788,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.000290378464289262,
+      "loss": 0.5194,
+      "step": 128260
+    },
+    {
+      "epoch": 6.370815535909407,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0002903387305056124,
+      "loss": 0.4942,
+      "step": 128270
+    },
+    {
+      "epoch": 6.371312208205026,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00029029899672196286,
+      "loss": 0.4992,
+      "step": 128280
+    },
+    {
+      "epoch": 6.371808880500645,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00029025926293831333,
+      "loss": 0.5089,
+      "step": 128290
+    },
+    {
+      "epoch": 6.372305552796265,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00029021952915466375,
+      "loss": 0.511,
+      "step": 128300
+    },
+    {
+      "epoch": 6.372802225091885,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002901797953710142,
+      "loss": 0.5151,
+      "step": 128310
+    },
+    {
+      "epoch": 6.373298897387504,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00029014006158736464,
+      "loss": 0.495,
+      "step": 128320
+    },
+    {
+      "epoch": 6.373795569683123,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00029010032780371517,
+      "loss": 0.4658,
+      "step": 128330
+    },
+    {
+      "epoch": 6.3742922419787424,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002900605940200656,
+      "loss": 0.4733,
+      "step": 128340
+    },
+    {
+      "epoch": 6.374788914274362,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.000290020860236416,
+      "loss": 0.5039,
+      "step": 128350
+    },
+    {
+      "epoch": 6.375285586569981,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00028998112645276647,
+      "loss": 0.4856,
+      "step": 128360
+    },
+    {
+      "epoch": 6.3757822588656,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00028994139266911694,
+      "loss": 0.4915,
+      "step": 128370
+    },
+    {
+      "epoch": 6.37627893116122,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.00028990165888546736,
+      "loss": 0.5009,
+      "step": 128380
+    },
+    {
+      "epoch": 6.3767756034568395,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00028986192510181783,
+      "loss": 0.5039,
+      "step": 128390
+    },
+    {
+      "epoch": 6.377272275752459,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0002898221913181683,
+      "loss": 0.5007,
+      "step": 128400
+    },
+    {
+      "epoch": 6.377768948048078,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002897824575345188,
+      "loss": 0.4989,
+      "step": 128410
+    },
+    {
+      "epoch": 6.378265620343697,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002897427237508692,
+      "loss": 0.4918,
+      "step": 128420
+    },
+    {
+      "epoch": 6.378762292639316,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0002897029899672196,
+      "loss": 0.5297,
+      "step": 128430
+    },
+    {
+      "epoch": 6.379258964934936,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00028966325618357013,
+      "loss": 0.5047,
+      "step": 128440
+    },
+    {
+      "epoch": 6.379755637230556,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00028962352239992055,
+      "loss": 0.4742,
+      "step": 128450
+    },
+    {
+      "epoch": 6.380252309526175,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00028958378861627097,
+      "loss": 0.486,
+      "step": 128460
+    },
+    {
+      "epoch": 6.380748981821794,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00028954405483262144,
+      "loss": 0.5336,
+      "step": 128470
+    },
+    {
+      "epoch": 6.381245654117413,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002895043210489719,
+      "loss": 0.5085,
+      "step": 128480
+    },
+    {
+      "epoch": 6.381742326413033,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0002894645872653224,
+      "loss": 0.4967,
+      "step": 128490
+    },
+    {
+      "epoch": 6.382238998708652,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002894248534816728,
+      "loss": 0.4979,
+      "step": 128500
+    },
+    {
+      "epoch": 6.382735671004271,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00028938511969802327,
+      "loss": 0.4953,
+      "step": 128510
+    },
+    {
+      "epoch": 6.383232343299891,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00028934538591437374,
+      "loss": 0.4948,
+      "step": 128520
+    },
+    {
+      "epoch": 6.38372901559551,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00028930565213072416,
+      "loss": 0.517,
+      "step": 128530
+    },
+    {
+      "epoch": 6.38422568789113,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0002892659183470746,
+      "loss": 0.506,
+      "step": 128540
+    },
+    {
+      "epoch": 6.384722360186749,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00028922618456342505,
+      "loss": 0.4918,
+      "step": 128550
+    },
+    {
+      "epoch": 6.385219032482368,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002891864507797755,
+      "loss": 0.5063,
+      "step": 128560
+    },
+    {
+      "epoch": 6.385715704777987,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.000289146716996126,
+      "loss": 0.5084,
+      "step": 128570
+    },
+    {
+      "epoch": 6.386212377073607,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002891069832124764,
+      "loss": 0.4974,
+      "step": 128580
+    },
+    {
+      "epoch": 6.386709049369226,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002890672494288269,
+      "loss": 0.4743,
+      "step": 128590
+    },
+    {
+      "epoch": 6.387205721664846,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00028902751564517735,
+      "loss": 0.4642,
+      "step": 128600
+    },
+    {
+      "epoch": 6.387702393960465,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00028898778186152777,
+      "loss": 0.4721,
+      "step": 128610
+    },
+    {
+      "epoch": 6.388199066256084,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002889480480778782,
+      "loss": 0.4628,
+      "step": 128620
+    },
+    {
+      "epoch": 6.388695738551704,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002889083142942287,
+      "loss": 0.4595,
+      "step": 128630
+    },
+    {
+      "epoch": 6.389192410847323,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00028886858051057913,
+      "loss": 0.5141,
+      "step": 128640
+    },
+    {
+      "epoch": 6.389689083142942,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0002888288467269296,
+      "loss": 0.5169,
+      "step": 128650
+    },
+    {
+      "epoch": 6.390185755438561,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00028878911294328,
+      "loss": 0.489,
+      "step": 128660
+    },
+    {
+      "epoch": 6.390682427734181,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002887493791596305,
+      "loss": 0.4895,
+      "step": 128670
+    },
+    {
+      "epoch": 6.391179100029801,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00028870964537598096,
+      "loss": 0.4744,
+      "step": 128680
+    },
+    {
+      "epoch": 6.39167577232542,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0002886699115923314,
+      "loss": 0.464,
+      "step": 128690
+    },
+    {
+      "epoch": 6.392172444621039,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00028863017780868185,
+      "loss": 0.496,
+      "step": 128700
+    },
+    {
+      "epoch": 6.392669116916658,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002885904440250323,
+      "loss": 0.4764,
+      "step": 128710
+    },
+    {
+      "epoch": 6.3931657892122775,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00028855071024138274,
+      "loss": 0.4833,
+      "step": 128720
+    },
+    {
+      "epoch": 6.393662461507897,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002885109764577332,
+      "loss": 0.4572,
+      "step": 128730
+    },
+    {
+      "epoch": 6.394159133803516,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0002884712426740837,
+      "loss": 0.4848,
+      "step": 128740
+    },
+    {
+      "epoch": 6.394655806099136,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0002884315088904341,
+      "loss": 0.4934,
+      "step": 128750
+    },
+    {
+      "epoch": 6.395152478394755,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00028839177510678457,
+      "loss": 0.5331,
+      "step": 128760
+    },
+    {
+      "epoch": 6.3956491506903745,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.000288352041323135,
+      "loss": 0.4802,
+      "step": 128770
+    },
+    {
+      "epoch": 6.396145822985994,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00028831230753948546,
+      "loss": 0.5062,
+      "step": 128780
+    },
+    {
+      "epoch": 6.396642495281613,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002882725737558359,
+      "loss": 0.5086,
+      "step": 128790
+    },
+    {
+      "epoch": 6.397139167577232,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00028823283997218634,
+      "loss": 0.5039,
+      "step": 128800
+    },
+    {
+      "epoch": 6.3976358398728514,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002881931061885368,
+      "loss": 0.5243,
+      "step": 128810
+    },
+    {
+      "epoch": 6.398132512168472,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002881533724048873,
+      "loss": 0.4958,
+      "step": 128820
+    },
+    {
+      "epoch": 6.398629184464091,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0002881136386212377,
+      "loss": 0.4818,
+      "step": 128830
+    },
+    {
+      "epoch": 6.39912585675971,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002880739048375882,
+      "loss": 0.4959,
+      "step": 128840
+    },
+    {
+      "epoch": 6.399622529055329,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002880341710539386,
+      "loss": 0.5146,
+      "step": 128850
+    },
+    {
+      "epoch": 6.4001192013509485,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002879944372702891,
+      "loss": 0.5259,
+      "step": 128860
+    },
+    {
+      "epoch": 6.400615873646568,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00028795470348663953,
+      "loss": 0.4883,
+      "step": 128870
+    },
+    {
+      "epoch": 6.401112545942187,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00028791496970298995,
+      "loss": 0.4779,
+      "step": 128880
+    },
+    {
+      "epoch": 6.401609218237807,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002878752359193404,
+      "loss": 0.5104,
+      "step": 128890
+    },
+    {
+      "epoch": 6.402105890533426,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002878355021356909,
+      "loss": 0.5174,
+      "step": 128900
+    },
+    {
+      "epoch": 6.4026025628290455,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002877957683520413,
+      "loss": 0.4908,
+      "step": 128910
+    },
+    {
+      "epoch": 6.403099235124665,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002877560345683918,
+      "loss": 0.4861,
+      "step": 128920
+    },
+    {
+      "epoch": 6.403595907420284,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00028771630078474225,
+      "loss": 0.4865,
+      "step": 128930
+    },
+    {
+      "epoch": 6.404092579715903,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002876765670010927,
+      "loss": 0.5096,
+      "step": 128940
+    },
+    {
+      "epoch": 6.404589252011522,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00028763683321744314,
+      "loss": 0.5094,
+      "step": 128950
+    },
+    {
+      "epoch": 6.4050859243071425,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00028759709943379356,
+      "loss": 0.4718,
+      "step": 128960
+    },
+    {
+      "epoch": 6.405582596602762,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002875573656501441,
+      "loss": 0.512,
+      "step": 128970
+    },
+    {
+      "epoch": 6.406079268898381,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002875176318664945,
+      "loss": 0.497,
+      "step": 128980
+    },
+    {
+      "epoch": 6.406575941194,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0002874778980828449,
+      "loss": 0.4937,
+      "step": 128990
+    },
+    {
+      "epoch": 6.407072613489619,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002874381642991954,
+      "loss": 0.5073,
+      "step": 129000
+    },
+    {
+      "epoch": 6.407569285785239,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00028739843051554586,
+      "loss": 0.5271,
+      "step": 129010
+    },
+    {
+      "epoch": 6.408065958080858,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00028735869673189633,
+      "loss": 0.4926,
+      "step": 129020
+    },
+    {
+      "epoch": 6.408562630376478,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00028731896294824675,
+      "loss": 0.4877,
+      "step": 129030
+    },
+    {
+      "epoch": 6.409059302672097,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002872792291645972,
+      "loss": 0.4803,
+      "step": 129040
+    },
+    {
+      "epoch": 6.4095559749677165,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0002872394953809477,
+      "loss": 0.476,
+      "step": 129050
+    },
+    {
+      "epoch": 6.410052647263336,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002871997615972981,
+      "loss": 0.5281,
+      "step": 129060
+    },
+    {
+      "epoch": 6.410549319558955,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00028716002781364853,
+      "loss": 0.4813,
+      "step": 129070
+    },
+    {
+      "epoch": 6.411045991854574,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.000287120294029999,
+      "loss": 0.4843,
+      "step": 129080
+    },
+    {
+      "epoch": 6.411542664150193,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00028708056024634947,
+      "loss": 0.4956,
+      "step": 129090
+    },
+    {
+      "epoch": 6.4120393364458135,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00028704082646269994,
+      "loss": 0.4909,
+      "step": 129100
+    },
+    {
+      "epoch": 6.412536008741433,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00028700109267905036,
+      "loss": 0.4595,
+      "step": 129110
+    },
+    {
+      "epoch": 6.413032681037052,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00028696135889540083,
+      "loss": 0.4983,
+      "step": 129120
+    },
+    {
+      "epoch": 6.413529353332671,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0002869216251117513,
+      "loss": 0.476,
+      "step": 129130
+    },
+    {
+      "epoch": 6.41402602562829,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002868818913281017,
+      "loss": 0.4897,
+      "step": 129140
+    },
+    {
+      "epoch": 6.41452269792391,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002868421575444522,
+      "loss": 0.4898,
+      "step": 129150
+    },
+    {
+      "epoch": 6.415019370219529,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00028680242376080266,
+      "loss": 0.4681,
+      "step": 129160
+    },
+    {
+      "epoch": 6.415516042515149,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002867626899771531,
+      "loss": 0.4966,
+      "step": 129170
+    },
+    {
+      "epoch": 6.416012714810768,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00028672295619350355,
+      "loss": 0.5034,
+      "step": 129180
+    },
+    {
+      "epoch": 6.416509387106387,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00028668322240985397,
+      "loss": 0.4797,
+      "step": 129190
+    },
+    {
+      "epoch": 6.417006059402007,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00028664348862620444,
+      "loss": 0.5159,
+      "step": 129200
+    },
+    {
+      "epoch": 6.417502731697626,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002866037548425549,
+      "loss": 0.4948,
+      "step": 129210
+    },
+    {
+      "epoch": 6.417999403993245,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00028656402105890533,
+      "loss": 0.513,
+      "step": 129220
+    },
+    {
+      "epoch": 6.418496076288864,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002865242872752558,
+      "loss": 0.4909,
+      "step": 129230
+    },
+    {
+      "epoch": 6.4189927485844835,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00028648455349160627,
+      "loss": 0.4891,
+      "step": 129240
+    },
+    {
+      "epoch": 6.419489420880104,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002864448197079567,
+      "loss": 0.4745,
+      "step": 129250
+    },
+    {
+      "epoch": 6.419986093175723,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00028640508592430716,
+      "loss": 0.4804,
+      "step": 129260
+    },
+    {
+      "epoch": 6.420482765471342,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00028636535214065763,
+      "loss": 0.4977,
+      "step": 129270
+    },
+    {
+      "epoch": 6.420979437766961,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00028632561835700805,
+      "loss": 0.5057,
+      "step": 129280
+    },
+    {
+      "epoch": 6.421476110062581,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002862858845733585,
+      "loss": 0.5113,
+      "step": 129290
+    },
+    {
+      "epoch": 6.4219727823582,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00028624615078970894,
+      "loss": 0.4901,
+      "step": 129300
+    },
+    {
+      "epoch": 6.422469454653819,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00028620641700605946,
+      "loss": 0.5311,
+      "step": 129310
+    },
+    {
+      "epoch": 6.422966126949439,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002861666832224099,
+      "loss": 0.4552,
+      "step": 129320
+    },
+    {
+      "epoch": 6.423462799245058,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002861269494387603,
+      "loss": 0.4717,
+      "step": 129330
+    },
+    {
+      "epoch": 6.423959471540678,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00028608721565511077,
+      "loss": 0.4843,
+      "step": 129340
+    },
+    {
+      "epoch": 6.424456143836297,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00028604748187146124,
+      "loss": 0.5129,
+      "step": 129350
+    },
+    {
+      "epoch": 6.424952816131916,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00028600774808781166,
+      "loss": 0.4949,
+      "step": 129360
+    },
+    {
+      "epoch": 6.425449488427535,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002859680143041621,
+      "loss": 0.4898,
+      "step": 129370
+    },
+    {
+      "epoch": 6.4259461607231545,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00028592828052051254,
+      "loss": 0.4962,
+      "step": 129380
+    },
+    {
+      "epoch": 6.426442833018775,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00028588854673686307,
+      "loss": 0.4867,
+      "step": 129390
+    },
+    {
+      "epoch": 6.426939505314394,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002858488129532135,
+      "loss": 0.5046,
+      "step": 129400
+    },
+    {
+      "epoch": 6.427436177610013,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002858090791695639,
+      "loss": 0.4614,
+      "step": 129410
+    },
+    {
+      "epoch": 6.427932849905632,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002857693453859144,
+      "loss": 0.4749,
+      "step": 129420
+    },
+    {
+      "epoch": 6.4284295222012515,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00028572961160226485,
+      "loss": 0.5002,
+      "step": 129430
+    },
+    {
+      "epoch": 6.428926194496871,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00028568987781861526,
+      "loss": 0.4991,
+      "step": 129440
+    },
+    {
+      "epoch": 6.42942286679249,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00028565014403496574,
+      "loss": 0.4621,
+      "step": 129450
+    },
+    {
+      "epoch": 6.429919539088109,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002856104102513162,
+      "loss": 0.4931,
+      "step": 129460
+    },
+    {
+      "epoch": 6.430416211383729,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002855706764676667,
+      "loss": 0.4974,
+      "step": 129470
+    },
+    {
+      "epoch": 6.4309128836793485,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0002855309426840171,
+      "loss": 0.4677,
+      "step": 129480
+    },
+    {
+      "epoch": 6.431409555974968,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002854912089003675,
+      "loss": 0.4817,
+      "step": 129490
+    },
+    {
+      "epoch": 6.431906228270587,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00028545147511671804,
+      "loss": 0.4852,
+      "step": 129500
+    },
+    {
+      "epoch": 6.432402900566206,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00028541174133306845,
+      "loss": 0.4982,
+      "step": 129510
+    },
+    {
+      "epoch": 6.4328995728618255,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00028537200754941887,
+      "loss": 0.5092,
+      "step": 129520
+    },
+    {
+      "epoch": 6.433396245157445,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00028533227376576934,
+      "loss": 0.476,
+      "step": 129530
+    },
+    {
+      "epoch": 6.433892917453065,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002852925399821198,
+      "loss": 0.4969,
+      "step": 129540
+    },
+    {
+      "epoch": 6.434389589748684,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002852528061984703,
+      "loss": 0.4964,
+      "step": 129550
+    },
+    {
+      "epoch": 6.434886262044303,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002852130724148207,
+      "loss": 0.484,
+      "step": 129560
+    },
+    {
+      "epoch": 6.4353829343399225,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002851733386311712,
+      "loss": 0.4956,
+      "step": 129570
+    },
+    {
+      "epoch": 6.435879606635542,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00028513360484752165,
+      "loss": 0.5129,
+      "step": 129580
+    },
+    {
+      "epoch": 6.436376278931161,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00028509387106387206,
+      "loss": 0.5242,
+      "step": 129590
+    },
+    {
+      "epoch": 6.43687295122678,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00028505413728022253,
+      "loss": 0.4968,
+      "step": 129600
+    },
+    {
+      "epoch": 6.4373696235224,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000285014403496573,
+      "loss": 0.4754,
+      "step": 129610
+    },
+    {
+      "epoch": 6.4378662958180195,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002849746697129234,
+      "loss": 0.5006,
+      "step": 129620
+    },
+    {
+      "epoch": 6.438362968113639,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002849349359292739,
+      "loss": 0.4809,
+      "step": 129630
+    },
+    {
+      "epoch": 6.438859640409258,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002848952021456243,
+      "loss": 0.5065,
+      "step": 129640
+    },
+    {
+      "epoch": 6.439356312704877,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002848554683619748,
+      "loss": 0.5027,
+      "step": 129650
+    },
+    {
+      "epoch": 6.439852985000496,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00028481573457832525,
+      "loss": 0.4714,
+      "step": 129660
+    },
+    {
+      "epoch": 6.440349657296116,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00028477600079467567,
+      "loss": 0.5001,
+      "step": 129670
+    },
+    {
+      "epoch": 6.440846329591736,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00028473626701102614,
+      "loss": 0.5064,
+      "step": 129680
+    },
+    {
+      "epoch": 6.441343001887355,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002846965332273766,
+      "loss": 0.4961,
+      "step": 129690
+    },
+    {
+      "epoch": 6.441839674182974,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00028465679944372703,
+      "loss": 0.5148,
+      "step": 129700
+    },
+    {
+      "epoch": 6.442336346478593,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002846170656600775,
+      "loss": 0.5455,
+      "step": 129710
+    },
+    {
+      "epoch": 6.442833018774213,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002845773318764279,
+      "loss": 0.4687,
+      "step": 129720
+    },
+    {
+      "epoch": 6.443329691069832,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0002845375980927784,
+      "loss": 0.4709,
+      "step": 129730
+    },
+    {
+      "epoch": 6.443826363365451,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00028449786430912886,
+      "loss": 0.4912,
+      "step": 129740
+    },
+    {
+      "epoch": 6.444323035661071,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002844581305254793,
+      "loss": 0.5387,
+      "step": 129750
+    },
+    {
+      "epoch": 6.4448197079566905,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00028441839674182975,
+      "loss": 0.4964,
+      "step": 129760
+    },
+    {
+      "epoch": 6.44531638025231,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002843786629581802,
+      "loss": 0.4962,
+      "step": 129770
+    },
+    {
+      "epoch": 6.445813052547929,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00028433892917453064,
+      "loss": 0.504,
+      "step": 129780
+    },
+    {
+      "epoch": 6.446309724843548,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0002842991953908811,
+      "loss": 0.5037,
+      "step": 129790
+    },
+    {
+      "epoch": 6.446806397139167,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002842594616072316,
+      "loss": 0.4845,
+      "step": 129800
+    },
+    {
+      "epoch": 6.447303069434787,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.000284219727823582,
+      "loss": 0.5116,
+      "step": 129810
+    },
+    {
+      "epoch": 6.447799741730407,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.00028417999403993247,
+      "loss": 0.4593,
+      "step": 129820
+    },
+    {
+      "epoch": 6.448296414026026,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002841402602562829,
+      "loss": 0.4668,
+      "step": 129830
+    },
+    {
+      "epoch": 6.448793086321645,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0002841005264726334,
+      "loss": 0.4476,
+      "step": 129840
+    },
+    {
+      "epoch": 6.449289758617264,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00028406079268898383,
+      "loss": 0.481,
+      "step": 129850
+    },
+    {
+      "epoch": 6.449786430912884,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00028402105890533425,
+      "loss": 0.5138,
+      "step": 129860
+    },
+    {
+      "epoch": 6.450283103208503,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002839813251216847,
+      "loss": 0.4492,
+      "step": 129870
+    },
+    {
+      "epoch": 6.450779775504122,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002839415913380352,
+      "loss": 0.4905,
+      "step": 129880
+    },
+    {
+      "epoch": 6.451276447799742,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002839018575543856,
+      "loss": 0.5089,
+      "step": 129890
+    },
+    {
+      "epoch": 6.451773120095361,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002838621237707361,
+      "loss": 0.4971,
+      "step": 129900
+    },
+    {
+      "epoch": 6.452269792390981,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00028382238998708655,
+      "loss": 0.5208,
+      "step": 129910
+    },
+    {
+      "epoch": 6.4527664646866,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.000283782656203437,
+      "loss": 0.4988,
+      "step": 129920
+    },
+    {
+      "epoch": 6.453263136982219,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00028374292241978744,
+      "loss": 0.4563,
+      "step": 129930
+    },
+    {
+      "epoch": 6.453759809277838,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00028370318863613786,
+      "loss": 0.5077,
+      "step": 129940
+    },
+    {
+      "epoch": 6.4542564815734575,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002836634548524883,
+      "loss": 0.4954,
+      "step": 129950
+    },
+    {
+      "epoch": 6.454753153869077,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002836237210688388,
+      "loss": 0.4795,
+      "step": 129960
+    },
+    {
+      "epoch": 6.455249826164697,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00028358398728518927,
+      "loss": 0.4732,
+      "step": 129970
+    },
+    {
+      "epoch": 6.455746498460316,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002835442535015397,
+      "loss": 0.4989,
+      "step": 129980
+    },
+    {
+      "epoch": 6.456243170755935,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00028350451971789016,
+      "loss": 0.4869,
+      "step": 129990
+    },
+    {
+      "epoch": 6.456739843051555,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00028346478593424063,
+      "loss": 0.488,
+      "step": 130000
+    },
+    {
+      "epoch": 6.457236515347174,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.00028342505215059105,
+      "loss": 0.4937,
+      "step": 130010
+    },
+    {
+      "epoch": 6.457733187642793,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00028338531836694146,
+      "loss": 0.5094,
+      "step": 130020
+    },
+    {
+      "epoch": 6.458229859938412,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.000283345584583292,
+      "loss": 0.4815,
+      "step": 130030
+    },
+    {
+      "epoch": 6.458726532234032,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002833058507996424,
+      "loss": 0.4699,
+      "step": 130040
+    },
+    {
+      "epoch": 6.459223204529652,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002832661170159929,
+      "loss": 0.4936,
+      "step": 130050
+    },
+    {
+      "epoch": 6.459719876825271,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002832263832323433,
+      "loss": 0.5095,
+      "step": 130060
+    },
+    {
+      "epoch": 6.46021654912089,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00028318664944869377,
+      "loss": 0.4635,
+      "step": 130070
+    },
+    {
+      "epoch": 6.460713221416509,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00028314691566504424,
+      "loss": 0.5229,
+      "step": 130080
+    },
+    {
+      "epoch": 6.4612098937121285,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00028310718188139466,
+      "loss": 0.4524,
+      "step": 130090
+    },
+    {
+      "epoch": 6.461706566007748,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002830674480977451,
+      "loss": 0.4919,
+      "step": 130100
+    },
+    {
+      "epoch": 6.462203238303368,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002830277143140956,
+      "loss": 0.4804,
+      "step": 130110
+    },
+    {
+      "epoch": 6.462699910598987,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.000282987980530446,
+      "loss": 0.4807,
+      "step": 130120
+    },
+    {
+      "epoch": 6.463196582894606,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002829482467467965,
+      "loss": 0.4682,
+      "step": 130130
+    },
+    {
+      "epoch": 6.4636932551902255,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00028290851296314696,
+      "loss": 0.4976,
+      "step": 130140
+    },
+    {
+      "epoch": 6.464189927485845,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002828687791794974,
+      "loss": 0.5323,
+      "step": 130150
+    },
+    {
+      "epoch": 6.464686599781464,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00028282904539584785,
+      "loss": 0.483,
+      "step": 130160
+    },
+    {
+      "epoch": 6.465183272077083,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00028278931161219826,
+      "loss": 0.5056,
+      "step": 130170
+    },
+    {
+      "epoch": 6.465679944372702,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00028274957782854873,
+      "loss": 0.5006,
+      "step": 130180
+    },
+    {
+      "epoch": 6.4661766166683226,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002827098440448992,
+      "loss": 0.4996,
+      "step": 130190
+    },
+    {
+      "epoch": 6.466673288963942,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0002826701102612496,
+      "loss": 0.5152,
+      "step": 130200
+    },
+    {
+      "epoch": 6.467169961259561,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002826303764776001,
+      "loss": 0.5057,
+      "step": 130210
+    },
+    {
+      "epoch": 6.46766663355518,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00028259064269395057,
+      "loss": 0.5236,
+      "step": 130220
+    },
+    {
+      "epoch": 6.4681633058507995,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.000282550908910301,
+      "loss": 0.4987,
+      "step": 130230
+    },
+    {
+      "epoch": 6.468659978146419,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00028251117512665145,
+      "loss": 0.5053,
+      "step": 130240
+    },
+    {
+      "epoch": 6.469156650442038,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00028247144134300187,
+      "loss": 0.4914,
+      "step": 130250
+    },
+    {
+      "epoch": 6.469653322737658,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00028243170755935234,
+      "loss": 0.4776,
+      "step": 130260
+    },
+    {
+      "epoch": 6.470149995033277,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002823919737757028,
+      "loss": 0.4943,
+      "step": 130270
+    },
+    {
+      "epoch": 6.4706466673288965,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00028235223999205323,
+      "loss": 0.4928,
+      "step": 130280
+    },
+    {
+      "epoch": 6.471143339624516,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002823125062084037,
+      "loss": 0.4775,
+      "step": 130290
+    },
+    {
+      "epoch": 6.471640011920135,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002822727724247542,
+      "loss": 0.4926,
+      "step": 130300
+    },
+    {
+      "epoch": 6.472136684215754,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0002822330386411046,
+      "loss": 0.5075,
+      "step": 130310
+    },
+    {
+      "epoch": 6.472633356511373,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00028219330485745506,
+      "loss": 0.5062,
+      "step": 130320
+    },
+    {
+      "epoch": 6.4731300288069935,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00028215357107380553,
+      "loss": 0.4768,
+      "step": 130330
+    },
+    {
+      "epoch": 6.473626701102613,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00028211383729015595,
+      "loss": 0.4855,
+      "step": 130340
+    },
+    {
+      "epoch": 6.474123373398232,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002820741035065064,
+      "loss": 0.4859,
+      "step": 130350
+    },
+    {
+      "epoch": 6.474620045693851,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00028203436972285684,
+      "loss": 0.4693,
+      "step": 130360
+    },
+    {
+      "epoch": 6.47511671798947,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00028199463593920737,
+      "loss": 0.4863,
+      "step": 130370
+    },
+    {
+      "epoch": 6.47561339028509,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002819549021555578,
+      "loss": 0.5194,
+      "step": 130380
+    },
+    {
+      "epoch": 6.476110062580709,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002819151683719082,
+      "loss": 0.4944,
+      "step": 130390
+    },
+    {
+      "epoch": 6.476606734876329,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00028187543458825867,
+      "loss": 0.4976,
+      "step": 130400
+    },
+    {
+      "epoch": 6.477103407171948,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00028183570080460914,
+      "loss": 0.4932,
+      "step": 130410
+    },
+    {
+      "epoch": 6.477600079467567,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002817959670209596,
+      "loss": 0.4907,
+      "step": 130420
+    },
+    {
+      "epoch": 6.478096751763187,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00028175623323731003,
+      "loss": 0.4963,
+      "step": 130430
+    },
+    {
+      "epoch": 6.478593424058806,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002817164994536605,
+      "loss": 0.4813,
+      "step": 130440
+    },
+    {
+      "epoch": 6.479090096354425,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.000281676765670011,
+      "loss": 0.4851,
+      "step": 130450
+    },
+    {
+      "epoch": 6.479586768650044,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002816370318863614,
+      "loss": 0.4949,
+      "step": 130460
+    },
+    {
+      "epoch": 6.4800834409456645,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002815972981027118,
+      "loss": 0.5014,
+      "step": 130470
+    },
+    {
+      "epoch": 6.480580113241284,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002815575643190623,
+      "loss": 0.4666,
+      "step": 130480
+    },
+    {
+      "epoch": 6.481076785536903,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00028151783053541275,
+      "loss": 0.478,
+      "step": 130490
+    },
+    {
+      "epoch": 6.481573457832522,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0002814780967517632,
+      "loss": 0.4738,
+      "step": 130500
+    },
+    {
+      "epoch": 6.482070130128141,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00028143836296811364,
+      "loss": 0.4727,
+      "step": 130510
+    },
+    {
+      "epoch": 6.482566802423761,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002813986291844641,
+      "loss": 0.4706,
+      "step": 130520
+    },
+    {
+      "epoch": 6.48306347471938,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002813588954008146,
+      "loss": 0.5028,
+      "step": 130530
+    },
+    {
+      "epoch": 6.483560147015,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.000281319161617165,
+      "loss": 0.4949,
+      "step": 130540
+    },
+    {
+      "epoch": 6.484056819310619,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0002812794278335154,
+      "loss": 0.4685,
+      "step": 130550
+    },
+    {
+      "epoch": 6.484553491606238,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00028123969404986594,
+      "loss": 0.4842,
+      "step": 130560
+    },
+    {
+      "epoch": 6.485050163901858,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00028119996026621636,
+      "loss": 0.4985,
+      "step": 130570
+    },
+    {
+      "epoch": 6.485546836197477,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00028116022648256683,
+      "loss": 0.5261,
+      "step": 130580
+    },
+    {
+      "epoch": 6.486043508493096,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00028112049269891725,
+      "loss": 0.5217,
+      "step": 130590
+    },
+    {
+      "epoch": 6.486540180788715,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002810807589152677,
+      "loss": 0.4858,
+      "step": 130600
+    },
+    {
+      "epoch": 6.487036853084335,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002810410251316182,
+      "loss": 0.4955,
+      "step": 130610
+    },
+    {
+      "epoch": 6.487533525379955,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002810012913479686,
+      "loss": 0.5115,
+      "step": 130620
+    },
+    {
+      "epoch": 6.488030197675574,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002809615575643191,
+      "loss": 0.5005,
+      "step": 130630
+    },
+    {
+      "epoch": 6.488526869971193,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00028092182378066955,
+      "loss": 0.5045,
+      "step": 130640
+    },
+    {
+      "epoch": 6.489023542266812,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00028088208999701997,
+      "loss": 0.5191,
+      "step": 130650
+    },
+    {
+      "epoch": 6.4895202145624316,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00028084235621337044,
+      "loss": 0.5194,
+      "step": 130660
+    },
+    {
+      "epoch": 6.490016886858051,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002808026224297209,
+      "loss": 0.5083,
+      "step": 130670
+    },
+    {
+      "epoch": 6.49051355915367,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002807628886460713,
+      "loss": 0.4884,
+      "step": 130680
+    },
+    {
+      "epoch": 6.49101023144929,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002807231548624218,
+      "loss": 0.4933,
+      "step": 130690
+    },
+    {
+      "epoch": 6.491506903744909,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002806834210787722,
+      "loss": 0.5009,
+      "step": 130700
+    },
+    {
+      "epoch": 6.492003576040529,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0002806436872951227,
+      "loss": 0.4687,
+      "step": 130710
+    },
+    {
+      "epoch": 6.492500248336148,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00028060395351147316,
+      "loss": 0.485,
+      "step": 130720
+    },
+    {
+      "epoch": 6.492996920631767,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002805642197278236,
+      "loss": 0.4935,
+      "step": 130730
+    },
+    {
+      "epoch": 6.493493592927386,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00028052448594417405,
+      "loss": 0.502,
+      "step": 130740
+    },
+    {
+      "epoch": 6.4939902652230055,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002804847521605245,
+      "loss": 0.4907,
+      "step": 130750
+    },
+    {
+      "epoch": 6.494486937518626,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00028044501837687493,
+      "loss": 0.5057,
+      "step": 130760
+    },
+    {
+      "epoch": 6.494983609814245,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002804052845932254,
+      "loss": 0.5142,
+      "step": 130770
+    },
+    {
+      "epoch": 6.495480282109864,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0002803655508095758,
+      "loss": 0.4872,
+      "step": 130780
+    },
+    {
+      "epoch": 6.495976954405483,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00028032581702592635,
+      "loss": 0.4561,
+      "step": 130790
+    },
+    {
+      "epoch": 6.4964736267011025,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00028028608324227677,
+      "loss": 0.5,
+      "step": 130800
+    },
+    {
+      "epoch": 6.496970298996722,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002802463494586272,
+      "loss": 0.4935,
+      "step": 130810
+    },
+    {
+      "epoch": 6.497466971292341,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.00028020661567497765,
+      "loss": 0.4962,
+      "step": 130820
+    },
+    {
+      "epoch": 6.49796364358796,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002801668818913281,
+      "loss": 0.4927,
+      "step": 130830
+    },
+    {
+      "epoch": 6.49846031588358,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00028012714810767854,
+      "loss": 0.4761,
+      "step": 130840
+    },
+    {
+      "epoch": 6.4989569881791995,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000280087414324029,
+      "loss": 0.4948,
+      "step": 130850
+    },
+    {
+      "epoch": 6.499453660474819,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002800476805403795,
+      "loss": 0.4841,
+      "step": 130860
+    },
+    {
+      "epoch": 6.499950332770438,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00028000794675672996,
+      "loss": 0.5016,
+      "step": 130870
+    },
+    {
+      "epoch": 6.500447005066057,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0002799682129730804,
+      "loss": 0.5021,
+      "step": 130880
+    },
+    {
+      "epoch": 6.500943677361676,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002799284791894308,
+      "loss": 0.4649,
+      "step": 130890
+    },
+    {
+      "epoch": 6.501440349657296,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002798887454057813,
+      "loss": 0.527,
+      "step": 130900
+    },
+    {
+      "epoch": 6.501937021952916,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00027984901162213173,
+      "loss": 0.4981,
+      "step": 130910
+    },
+    {
+      "epoch": 6.502433694248535,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00027980927783848215,
+      "loss": 0.4976,
+      "step": 130920
+    },
+    {
+      "epoch": 6.502930366544154,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002797695440548326,
+      "loss": 0.4933,
+      "step": 130930
+    },
+    {
+      "epoch": 6.5034270388397735,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002797298102711831,
+      "loss": 0.5107,
+      "step": 130940
+    },
+    {
+      "epoch": 6.503923711135393,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00027969007648753357,
+      "loss": 0.5004,
+      "step": 130950
+    },
+    {
+      "epoch": 6.504420383431012,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.000279650342703884,
+      "loss": 0.4815,
+      "step": 130960
+    },
+    {
+      "epoch": 6.504917055726631,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00027961060892023445,
+      "loss": 0.4926,
+      "step": 130970
+    },
+    {
+      "epoch": 6.505413728022251,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002795708751365849,
+      "loss": 0.4914,
+      "step": 130980
+    },
+    {
+      "epoch": 6.5059104003178705,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.00027953114135293534,
+      "loss": 0.4831,
+      "step": 130990
+    },
+    {
+      "epoch": 6.50640707261349,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00027949140756928576,
+      "loss": 0.4801,
+      "step": 131000
+    },
+    {
+      "epoch": 6.506903744909109,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0002794516737856363,
+      "loss": 0.5064,
+      "step": 131010
+    },
+    {
+      "epoch": 6.507400417204728,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002794119400019867,
+      "loss": 0.4795,
+      "step": 131020
+    },
+    {
+      "epoch": 6.507897089500347,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002793722062183372,
+      "loss": 0.4814,
+      "step": 131030
+    },
+    {
+      "epoch": 6.508393761795967,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0002793324724346876,
+      "loss": 0.4836,
+      "step": 131040
+    },
+    {
+      "epoch": 6.508890434091587,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00027929273865103806,
+      "loss": 0.4854,
+      "step": 131050
+    },
+    {
+      "epoch": 6.509387106387206,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00027925300486738853,
+      "loss": 0.4736,
+      "step": 131060
+    },
+    {
+      "epoch": 6.509883778682825,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00027921327108373895,
+      "loss": 0.4926,
+      "step": 131070
+    },
+    {
+      "epoch": 6.510380450978444,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00027917353730008937,
+      "loss": 0.487,
+      "step": 131080
+    },
+    {
+      "epoch": 6.510877123274064,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002791338035164399,
+      "loss": 0.4976,
+      "step": 131090
+    },
+    {
+      "epoch": 6.511373795569683,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002790940697327903,
+      "loss": 0.4842,
+      "step": 131100
+    },
+    {
+      "epoch": 6.511870467865302,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002790543359491408,
+      "loss": 0.4664,
+      "step": 131110
+    },
+    {
+      "epoch": 6.512367140160922,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002790146021654912,
+      "loss": 0.4964,
+      "step": 131120
+    },
+    {
+      "epoch": 6.5128638124565414,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00027897486838184167,
+      "loss": 0.5052,
+      "step": 131130
+    },
+    {
+      "epoch": 6.513360484752161,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00027893513459819214,
+      "loss": 0.4974,
+      "step": 131140
+    },
+    {
+      "epoch": 6.51385715704778,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00027889540081454256,
+      "loss": 0.4822,
+      "step": 131150
+    },
+    {
+      "epoch": 6.514353829343399,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00027885566703089303,
+      "loss": 0.5037,
+      "step": 131160
+    },
+    {
+      "epoch": 6.514850501639018,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002788159332472435,
+      "loss": 0.4931,
+      "step": 131170
+    },
+    {
+      "epoch": 6.515347173934638,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002787761994635939,
+      "loss": 0.501,
+      "step": 131180
+    },
+    {
+      "epoch": 6.515843846230258,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002787364656799444,
+      "loss": 0.4913,
+      "step": 131190
+    },
+    {
+      "epoch": 6.516340518525877,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00027869673189629486,
+      "loss": 0.5102,
+      "step": 131200
+    },
+    {
+      "epoch": 6.516837190821496,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002786569981126453,
+      "loss": 0.5017,
+      "step": 131210
+    },
+    {
+      "epoch": 6.517333863117115,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00027861726432899575,
+      "loss": 0.5112,
+      "step": 131220
+    },
+    {
+      "epoch": 6.517830535412735,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00027857753054534617,
+      "loss": 0.4644,
+      "step": 131230
+    },
+    {
+      "epoch": 6.518327207708354,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002785377967616967,
+      "loss": 0.48,
+      "step": 131240
+    },
+    {
+      "epoch": 6.518823880003973,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002784980629780471,
+      "loss": 0.4718,
+      "step": 131250
+    },
+    {
+      "epoch": 6.519320552299593,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002784583291943975,
+      "loss": 0.4539,
+      "step": 131260
+    },
+    {
+      "epoch": 6.519817224595212,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.000278418595410748,
+      "loss": 0.5051,
+      "step": 131270
+    },
+    {
+      "epoch": 6.520313896890832,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00027837886162709847,
+      "loss": 0.5004,
+      "step": 131280
+    },
+    {
+      "epoch": 6.520810569186451,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002783391278434489,
+      "loss": 0.4924,
+      "step": 131290
+    },
+    {
+      "epoch": 6.52130724148207,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00027829939405979936,
+      "loss": 0.4745,
+      "step": 131300
+    },
+    {
+      "epoch": 6.521803913777689,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00027825966027614983,
+      "loss": 0.489,
+      "step": 131310
+    },
+    {
+      "epoch": 6.5223005860733085,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002782199264925003,
+      "loss": 0.5139,
+      "step": 131320
+    },
+    {
+      "epoch": 6.522797258368929,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002781801927088507,
+      "loss": 0.5139,
+      "step": 131330
+    },
+    {
+      "epoch": 6.523293930664548,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00027814045892520114,
+      "loss": 0.4829,
+      "step": 131340
+    },
+    {
+      "epoch": 6.523790602960167,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002781007251415516,
+      "loss": 0.4999,
+      "step": 131350
+    },
+    {
+      "epoch": 6.524287275255786,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002780609913579021,
+      "loss": 0.4658,
+      "step": 131360
+    },
+    {
+      "epoch": 6.524783947551406,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002780212575742525,
+      "loss": 0.4999,
+      "step": 131370
+    },
+    {
+      "epoch": 6.525280619847025,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00027798152379060297,
+      "loss": 0.4937,
+      "step": 131380
+    },
+    {
+      "epoch": 6.525777292142644,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00027794179000695344,
+      "loss": 0.4633,
+      "step": 131390
+    },
+    {
+      "epoch": 6.526273964438264,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002779020562233039,
+      "loss": 0.4726,
+      "step": 131400
+    },
+    {
+      "epoch": 6.526770636733883,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0002778623224396543,
+      "loss": 0.5023,
+      "step": 131410
+    },
+    {
+      "epoch": 6.527267309029503,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00027782258865600474,
+      "loss": 0.4984,
+      "step": 131420
+    },
+    {
+      "epoch": 6.527763981325122,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00027778285487235527,
+      "loss": 0.5,
+      "step": 131430
+    },
+    {
+      "epoch": 6.528260653620741,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002777431210887057,
+      "loss": 0.4708,
+      "step": 131440
+    },
+    {
+      "epoch": 6.52875732591636,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002777033873050561,
+      "loss": 0.4609,
+      "step": 131450
+    },
+    {
+      "epoch": 6.5292539982119795,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002776636535214066,
+      "loss": 0.4748,
+      "step": 131460
+    },
+    {
+      "epoch": 6.529750670507599,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00027762391973775705,
+      "loss": 0.5029,
+      "step": 131470
+    },
+    {
+      "epoch": 6.530247342803218,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002775841859541075,
+      "loss": 0.4988,
+      "step": 131480
+    },
+    {
+      "epoch": 6.530744015098838,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00027754445217045793,
+      "loss": 0.5001,
+      "step": 131490
+    },
+    {
+      "epoch": 6.531240687394457,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002775047183868084,
+      "loss": 0.4659,
+      "step": 131500
+    },
+    {
+      "epoch": 6.5317373596900765,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002774649846031589,
+      "loss": 0.4866,
+      "step": 131510
+    },
+    {
+      "epoch": 6.532234031985696,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002774252508195093,
+      "loss": 0.4882,
+      "step": 131520
+    },
+    {
+      "epoch": 6.532730704281315,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002773855170358597,
+      "loss": 0.4618,
+      "step": 131530
+    },
+    {
+      "epoch": 6.533227376576934,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00027734578325221024,
+      "loss": 0.4812,
+      "step": 131540
+    },
+    {
+      "epoch": 6.533724048872553,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00027730604946856065,
+      "loss": 0.4814,
+      "step": 131550
+    },
+    {
+      "epoch": 6.5342207211681735,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002772663156849111,
+      "loss": 0.4608,
+      "step": 131560
+    },
+    {
+      "epoch": 6.534717393463793,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00027722658190126154,
+      "loss": 0.4921,
+      "step": 131570
+    },
+    {
+      "epoch": 6.535214065759412,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.000277186848117612,
+      "loss": 0.482,
+      "step": 131580
+    },
+    {
+      "epoch": 6.535710738055031,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002771471143339625,
+      "loss": 0.4734,
+      "step": 131590
+    },
+    {
+      "epoch": 6.5362074103506504,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002771073805503129,
+      "loss": 0.4999,
+      "step": 131600
+    },
+    {
+      "epoch": 6.53670408264627,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002770676467666634,
+      "loss": 0.4949,
+      "step": 131610
+    },
+    {
+      "epoch": 6.537200754941889,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00027702791298301385,
+      "loss": 0.5104,
+      "step": 131620
+    },
+    {
+      "epoch": 6.537697427237509,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00027698817919936426,
+      "loss": 0.5106,
+      "step": 131630
+    },
+    {
+      "epoch": 6.538194099533128,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00027694844541571473,
+      "loss": 0.4902,
+      "step": 131640
+    },
+    {
+      "epoch": 6.5386907718287475,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00027690871163206515,
+      "loss": 0.4874,
+      "step": 131650
+    },
+    {
+      "epoch": 6.539187444124367,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002768689778484156,
+      "loss": 0.4876,
+      "step": 131660
+    },
+    {
+      "epoch": 6.539684116419986,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002768292440647661,
+      "loss": 0.5096,
+      "step": 131670
+    },
+    {
+      "epoch": 6.540180788715605,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002767895102811165,
+      "loss": 0.4685,
+      "step": 131680
+    },
+    {
+      "epoch": 6.540677461011224,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.000276749776497467,
+      "loss": 0.5019,
+      "step": 131690
+    },
+    {
+      "epoch": 6.5411741333068445,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00027671004271381745,
+      "loss": 0.5112,
+      "step": 131700
+    },
+    {
+      "epoch": 6.541670805602464,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00027667030893016787,
+      "loss": 0.4696,
+      "step": 131710
+    },
+    {
+      "epoch": 6.542167477898083,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00027663057514651834,
+      "loss": 0.5079,
+      "step": 131720
+    },
+    {
+      "epoch": 6.542664150193702,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002765908413628688,
+      "loss": 0.4758,
+      "step": 131730
+    },
+    {
+      "epoch": 6.543160822489321,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00027655110757921923,
+      "loss": 0.5078,
+      "step": 131740
+    },
+    {
+      "epoch": 6.543657494784941,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002765113737955697,
+      "loss": 0.4915,
+      "step": 131750
+    },
+    {
+      "epoch": 6.54415416708056,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002764716400119201,
+      "loss": 0.4816,
+      "step": 131760
+    },
+    {
+      "epoch": 6.54465083937618,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00027643190622827064,
+      "loss": 0.4976,
+      "step": 131770
+    },
+    {
+      "epoch": 6.545147511671799,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00027639217244462106,
+      "loss": 0.5056,
+      "step": 131780
+    },
+    {
+      "epoch": 6.545644183967418,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002763524386609715,
+      "loss": 0.4844,
+      "step": 131790
+    },
+    {
+      "epoch": 6.546140856263038,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00027631270487732195,
+      "loss": 0.5009,
+      "step": 131800
+    },
+    {
+      "epoch": 6.546637528558657,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002762729710936724,
+      "loss": 0.5147,
+      "step": 131810
+    },
+    {
+      "epoch": 6.547134200854276,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00027623323731002284,
+      "loss": 0.472,
+      "step": 131820
+    },
+    {
+      "epoch": 6.547630873149895,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002761935035263733,
+      "loss": 0.512,
+      "step": 131830
+    },
+    {
+      "epoch": 6.5481275454455155,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0002761537697427238,
+      "loss": 0.5148,
+      "step": 131840
+    },
+    {
+      "epoch": 6.548624217741135,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00027611403595907425,
+      "loss": 0.4977,
+      "step": 131850
+    },
+    {
+      "epoch": 6.549120890036754,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00027607430217542467,
+      "loss": 0.4998,
+      "step": 131860
+    },
+    {
+      "epoch": 6.549617562332373,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002760345683917751,
+      "loss": 0.5116,
+      "step": 131870
+    },
+    {
+      "epoch": 6.550114234627992,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00027599483460812556,
+      "loss": 0.5357,
+      "step": 131880
+    },
+    {
+      "epoch": 6.550610906923612,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00027595510082447603,
+      "loss": 0.4937,
+      "step": 131890
+    },
+    {
+      "epoch": 6.551107579219231,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00027591536704082645,
+      "loss": 0.5095,
+      "step": 131900
+    },
+    {
+      "epoch": 6.551604251514851,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002758756332571769,
+      "loss": 0.49,
+      "step": 131910
+    },
+    {
+      "epoch": 6.55210092381047,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002758358994735274,
+      "loss": 0.5151,
+      "step": 131920
+    },
+    {
+      "epoch": 6.552597596106089,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00027579616568987786,
+      "loss": 0.4589,
+      "step": 131930
+    },
+    {
+      "epoch": 6.553094268401709,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0002757564319062283,
+      "loss": 0.4864,
+      "step": 131940
+    },
+    {
+      "epoch": 6.553590940697328,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002757166981225787,
+      "loss": 0.4945,
+      "step": 131950
+    },
+    {
+      "epoch": 6.554087612992947,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0002756769643389292,
+      "loss": 0.4959,
+      "step": 131960
+    },
+    {
+      "epoch": 6.554584285288566,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00027563723055527964,
+      "loss": 0.4824,
+      "step": 131970
+    },
+    {
+      "epoch": 6.555080957584186,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00027559749677163005,
+      "loss": 0.5406,
+      "step": 131980
+    },
+    {
+      "epoch": 6.555577629879806,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002755577629879805,
+      "loss": 0.4884,
+      "step": 131990
+    },
+    {
+      "epoch": 6.556074302175425,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.000275518029204331,
+      "loss": 0.4944,
+      "step": 132000
+    },
+    {
+      "epoch": 6.556570974471044,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00027547829542068147,
+      "loss": 0.49,
+      "step": 132010
+    },
+    {
+      "epoch": 6.557067646766663,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002754385616370319,
+      "loss": 0.5078,
+      "step": 132020
+    },
+    {
+      "epoch": 6.5575643190622825,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00027539882785338236,
+      "loss": 0.4808,
+      "step": 132030
+    },
+    {
+      "epoch": 6.558060991357902,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00027535909406973283,
+      "loss": 0.5131,
+      "step": 132040
+    },
+    {
+      "epoch": 6.558557663653522,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00027531936028608325,
+      "loss": 0.5028,
+      "step": 132050
+    },
+    {
+      "epoch": 6.559054335949141,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002752796265024337,
+      "loss": 0.4891,
+      "step": 132060
+    },
+    {
+      "epoch": 6.55955100824476,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002752398927187842,
+      "loss": 0.4724,
+      "step": 132070
+    },
+    {
+      "epoch": 6.56004768054038,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002752001589351346,
+      "loss": 0.5097,
+      "step": 132080
+    },
+    {
+      "epoch": 6.560544352835999,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002751604251514851,
+      "loss": 0.4934,
+      "step": 132090
+    },
+    {
+      "epoch": 6.561041025131618,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002751206913678355,
+      "loss": 0.4868,
+      "step": 132100
+    },
+    {
+      "epoch": 6.561537697427237,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00027508095758418597,
+      "loss": 0.4855,
+      "step": 132110
+    },
+    {
+      "epoch": 6.5620343697228565,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00027504122380053644,
+      "loss": 0.4691,
+      "step": 132120
+    },
+    {
+      "epoch": 6.562531042018477,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00027500149001688685,
+      "loss": 0.4996,
+      "step": 132130
+    },
+    {
+      "epoch": 6.563027714314096,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002749617562332373,
+      "loss": 0.5151,
+      "step": 132140
+    },
+    {
+      "epoch": 6.563524386609715,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002749220224495878,
+      "loss": 0.4846,
+      "step": 132150
+    },
+    {
+      "epoch": 6.564021058905334,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002748822886659382,
+      "loss": 0.493,
+      "step": 132160
+    },
+    {
+      "epoch": 6.5645177312009535,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002748425548822887,
+      "loss": 0.4981,
+      "step": 132170
+    },
+    {
+      "epoch": 6.565014403496573,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002748028210986391,
+      "loss": 0.4957,
+      "step": 132180
+    },
+    {
+      "epoch": 6.565511075792192,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0002747630873149896,
+      "loss": 0.476,
+      "step": 132190
+    },
+    {
+      "epoch": 6.566007748087811,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00027472335353134005,
+      "loss": 0.4689,
+      "step": 132200
+    },
+    {
+      "epoch": 6.566504420383431,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00027468361974769046,
+      "loss": 0.5053,
+      "step": 132210
+    },
+    {
+      "epoch": 6.5670010926790505,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00027464388596404093,
+      "loss": 0.4794,
+      "step": 132220
+    },
+    {
+      "epoch": 6.56749776497467,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0002746041521803914,
+      "loss": 0.4824,
+      "step": 132230
+    },
+    {
+      "epoch": 6.567994437270289,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002745644183967418,
+      "loss": 0.5269,
+      "step": 132240
+    },
+    {
+      "epoch": 6.568491109565908,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002745246846130923,
+      "loss": 0.4838,
+      "step": 132250
+    },
+    {
+      "epoch": 6.568987781861527,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00027448495082944277,
+      "loss": 0.497,
+      "step": 132260
+    },
+    {
+      "epoch": 6.569484454157147,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002744452170457932,
+      "loss": 0.5044,
+      "step": 132270
+    },
+    {
+      "epoch": 6.569981126452767,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00027440548326214365,
+      "loss": 0.4951,
+      "step": 132280
+    },
+    {
+      "epoch": 6.570477798748386,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00027436574947849407,
+      "loss": 0.5129,
+      "step": 132290
+    },
+    {
+      "epoch": 6.570974471044005,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002743260156948446,
+      "loss": 0.4742,
+      "step": 132300
+    },
+    {
+      "epoch": 6.5714711433396245,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.000274286281911195,
+      "loss": 0.4861,
+      "step": 132310
+    },
+    {
+      "epoch": 6.571967815635244,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00027424654812754543,
+      "loss": 0.4985,
+      "step": 132320
+    },
+    {
+      "epoch": 6.572464487930863,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002742068143438959,
+      "loss": 0.4778,
+      "step": 132330
+    },
+    {
+      "epoch": 6.572961160226482,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002741670805602464,
+      "loss": 0.5032,
+      "step": 132340
+    },
+    {
+      "epoch": 6.573457832522102,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002741273467765968,
+      "loss": 0.4898,
+      "step": 132350
+    },
+    {
+      "epoch": 6.5739545048177215,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00027408761299294726,
+      "loss": 0.5188,
+      "step": 132360
+    },
+    {
+      "epoch": 6.574451177113341,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00027404787920929773,
+      "loss": 0.4875,
+      "step": 132370
+    },
+    {
+      "epoch": 6.57494784940896,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002740081454256482,
+      "loss": 0.5,
+      "step": 132380
+    },
+    {
+      "epoch": 6.575444521704579,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002739684116419986,
+      "loss": 0.4891,
+      "step": 132390
+    },
+    {
+      "epoch": 6.575941194000198,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00027392867785834904,
+      "loss": 0.4886,
+      "step": 132400
+    },
+    {
+      "epoch": 6.576437866295818,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00027388894407469956,
+      "loss": 0.4917,
+      "step": 132410
+    },
+    {
+      "epoch": 6.576934538591438,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00027384921029105,
+      "loss": 0.4932,
+      "step": 132420
+    },
+    {
+      "epoch": 6.577431210887057,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0002738094765074004,
+      "loss": 0.4747,
+      "step": 132430
+    },
+    {
+      "epoch": 6.577927883182676,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00027376974272375087,
+      "loss": 0.5104,
+      "step": 132440
+    },
+    {
+      "epoch": 6.578424555478295,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00027373000894010134,
+      "loss": 0.4977,
+      "step": 132450
+    },
+    {
+      "epoch": 6.578921227773915,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002736902751564518,
+      "loss": 0.4797,
+      "step": 132460
+    },
+    {
+      "epoch": 6.579417900069534,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.00027365054137280223,
+      "loss": 0.4796,
+      "step": 132470
+    },
+    {
+      "epoch": 6.579914572365153,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00027361080758915265,
+      "loss": 0.4642,
+      "step": 132480
+    },
+    {
+      "epoch": 6.580411244660773,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00027357107380550317,
+      "loss": 0.4925,
+      "step": 132490
+    },
+    {
+      "epoch": 6.580907916956392,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002735313400218536,
+      "loss": 0.486,
+      "step": 132500
+    },
+    {
+      "epoch": 6.581404589252012,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00027349160623820406,
+      "loss": 0.4796,
+      "step": 132510
+    },
+    {
+      "epoch": 6.581901261547631,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002734518724545545,
+      "loss": 0.512,
+      "step": 132520
+    },
+    {
+      "epoch": 6.58239793384325,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00027341213867090495,
+      "loss": 0.4844,
+      "step": 132530
+    },
+    {
+      "epoch": 6.582894606138869,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002733724048872554,
+      "loss": 0.4681,
+      "step": 132540
+    },
+    {
+      "epoch": 6.583391278434489,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00027333267110360584,
+      "loss": 0.4842,
+      "step": 132550
+    },
+    {
+      "epoch": 6.583887950730109,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002732929373199563,
+      "loss": 0.4881,
+      "step": 132560
+    },
+    {
+      "epoch": 6.584384623025728,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002732532035363068,
+      "loss": 0.5233,
+      "step": 132570
+    },
+    {
+      "epoch": 6.584881295321347,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002732134697526572,
+      "loss": 0.4756,
+      "step": 132580
+    },
+    {
+      "epoch": 6.585377967616966,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00027317373596900767,
+      "loss": 0.4881,
+      "step": 132590
+    },
+    {
+      "epoch": 6.585874639912586,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00027313400218535814,
+      "loss": 0.4969,
+      "step": 132600
+    },
+    {
+      "epoch": 6.586371312208205,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00027309426840170856,
+      "loss": 0.504,
+      "step": 132610
+    },
+    {
+      "epoch": 6.586867984503824,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00027305453461805903,
+      "loss": 0.4739,
+      "step": 132620
+    },
+    {
+      "epoch": 6.587364656799444,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00027301480083440945,
+      "loss": 0.5395,
+      "step": 132630
+    },
+    {
+      "epoch": 6.587861329095063,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002729750670507599,
+      "loss": 0.4975,
+      "step": 132640
+    },
+    {
+      "epoch": 6.588358001390683,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002729353332671104,
+      "loss": 0.4969,
+      "step": 132650
+    },
+    {
+      "epoch": 6.588854673686302,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002728955994834608,
+      "loss": 0.4937,
+      "step": 132660
+    },
+    {
+      "epoch": 6.589351345981921,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002728558656998113,
+      "loss": 0.4718,
+      "step": 132670
+    },
+    {
+      "epoch": 6.58984801827754,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00027281613191616175,
+      "loss": 0.4972,
+      "step": 132680
+    },
+    {
+      "epoch": 6.5903446905731595,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00027277639813251217,
+      "loss": 0.4969,
+      "step": 132690
+    },
+    {
+      "epoch": 6.59084136286878,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00027273666434886264,
+      "loss": 0.4912,
+      "step": 132700
+    },
+    {
+      "epoch": 6.591338035164399,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002726969305652131,
+      "loss": 0.5153,
+      "step": 132710
+    },
+    {
+      "epoch": 6.591834707460018,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002726571967815635,
+      "loss": 0.4904,
+      "step": 132720
+    },
+    {
+      "epoch": 6.592331379755637,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.000272617462997914,
+      "loss": 0.49,
+      "step": 132730
+    },
+    {
+      "epoch": 6.5928280520512565,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002725777292142644,
+      "loss": 0.4925,
+      "step": 132740
+    },
+    {
+      "epoch": 6.593324724346876,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002725379954306149,
+      "loss": 0.5258,
+      "step": 132750
+    },
+    {
+      "epoch": 6.593821396642495,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00027249826164696536,
+      "loss": 0.4999,
+      "step": 132760
+    },
+    {
+      "epoch": 6.594318068938115,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0002724585278633158,
+      "loss": 0.4972,
+      "step": 132770
+    },
+    {
+      "epoch": 6.594814741233734,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00027241879407966625,
+      "loss": 0.4713,
+      "step": 132780
+    },
+    {
+      "epoch": 6.595311413529354,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002723790602960167,
+      "loss": 0.49,
+      "step": 132790
+    },
+    {
+      "epoch": 6.595808085824973,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00027233932651236713,
+      "loss": 0.504,
+      "step": 132800
+    },
+    {
+      "epoch": 6.596304758120592,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0002722995927287176,
+      "loss": 0.4771,
+      "step": 132810
+    },
+    {
+      "epoch": 6.596801430416211,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.000272259858945068,
+      "loss": 0.4936,
+      "step": 132820
+    },
+    {
+      "epoch": 6.5972981027118305,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00027222012516141855,
+      "loss": 0.4815,
+      "step": 132830
+    },
+    {
+      "epoch": 6.59779477500745,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00027218039137776897,
+      "loss": 0.5094,
+      "step": 132840
+    },
+    {
+      "epoch": 6.59829144730307,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002721406575941194,
+      "loss": 0.5089,
+      "step": 132850
+    },
+    {
+      "epoch": 6.598788119598689,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00027210092381046985,
+      "loss": 0.4632,
+      "step": 132860
+    },
+    {
+      "epoch": 6.599284791894308,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002720611900268203,
+      "loss": 0.483,
+      "step": 132870
+    },
+    {
+      "epoch": 6.5997814641899275,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00027202145624317074,
+      "loss": 0.4891,
+      "step": 132880
+    },
+    {
+      "epoch": 6.600278136485547,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002719817224595212,
+      "loss": 0.4684,
+      "step": 132890
+    },
+    {
+      "epoch": 6.600774808781166,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002719419886758717,
+      "loss": 0.4608,
+      "step": 132900
+    },
+    {
+      "epoch": 6.601271481076785,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00027190225489222216,
+      "loss": 0.485,
+      "step": 132910
+    },
+    {
+      "epoch": 6.601768153372404,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002718625211085726,
+      "loss": 0.4507,
+      "step": 132920
+    },
+    {
+      "epoch": 6.6022648256680245,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.000271822787324923,
+      "loss": 0.4968,
+      "step": 132930
+    },
+    {
+      "epoch": 6.602761497963644,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002717830535412735,
+      "loss": 0.4773,
+      "step": 132940
+    },
+    {
+      "epoch": 6.603258170259263,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00027174331975762393,
+      "loss": 0.4796,
+      "step": 132950
+    },
+    {
+      "epoch": 6.603754842554882,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0002717035859739744,
+      "loss": 0.488,
+      "step": 132960
+    },
+    {
+      "epoch": 6.604251514850501,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0002716638521903248,
+      "loss": 0.4728,
+      "step": 132970
+    },
+    {
+      "epoch": 6.604748187146121,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002716241184066753,
+      "loss": 0.5108,
+      "step": 132980
+    },
+    {
+      "epoch": 6.60524485944174,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00027158438462302576,
+      "loss": 0.4891,
+      "step": 132990
+    },
+    {
+      "epoch": 6.60574153173736,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002715446508393762,
+      "loss": 0.512,
+      "step": 133000
+    },
+    {
+      "epoch": 6.606238204032979,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00027150491705572665,
+      "loss": 0.4805,
+      "step": 133010
+    },
+    {
+      "epoch": 6.6067348763285985,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0002714651832720771,
+      "loss": 0.4996,
+      "step": 133020
+    },
+    {
+      "epoch": 6.607231548624218,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00027142544948842754,
+      "loss": 0.5027,
+      "step": 133030
+    },
+    {
+      "epoch": 6.607728220919837,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.000271385715704778,
+      "loss": 0.53,
+      "step": 133040
+    },
+    {
+      "epoch": 6.608224893215456,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00027134598192112843,
+      "loss": 0.507,
+      "step": 133050
+    },
+    {
+      "epoch": 6.608721565511075,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002713062481374789,
+      "loss": 0.4711,
+      "step": 133060
+    },
+    {
+      "epoch": 6.6092182378066955,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00027126651435382937,
+      "loss": 0.4811,
+      "step": 133070
+    },
+    {
+      "epoch": 6.609714910102315,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002712267805701798,
+      "loss": 0.4837,
+      "step": 133080
+    },
+    {
+      "epoch": 6.610211582397934,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00027118704678653026,
+      "loss": 0.5085,
+      "step": 133090
+    },
+    {
+      "epoch": 6.610708254693553,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00027114731300288073,
+      "loss": 0.5287,
+      "step": 133100
+    },
+    {
+      "epoch": 6.611204926989172,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00027110757921923115,
+      "loss": 0.5062,
+      "step": 133110
+    },
+    {
+      "epoch": 6.611701599284792,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002710678454355816,
+      "loss": 0.4982,
+      "step": 133120
+    },
+    {
+      "epoch": 6.612198271580411,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002710281116519321,
+      "loss": 0.4991,
+      "step": 133130
+    },
+    {
+      "epoch": 6.612694943876031,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002709883778682825,
+      "loss": 0.5036,
+      "step": 133140
+    },
+    {
+      "epoch": 6.61319161617165,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.000270948644084633,
+      "loss": 0.4817,
+      "step": 133150
+    },
+    {
+      "epoch": 6.613688288467269,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002709089103009834,
+      "loss": 0.4849,
+      "step": 133160
+    },
+    {
+      "epoch": 6.614184960762889,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00027086917651733387,
+      "loss": 0.5049,
+      "step": 133170
+    },
+    {
+      "epoch": 6.614681633058508,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00027082944273368434,
+      "loss": 0.5086,
+      "step": 133180
+    },
+    {
+      "epoch": 6.615178305354127,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00027078970895003476,
+      "loss": 0.4887,
+      "step": 133190
+    },
+    {
+      "epoch": 6.615674977649746,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00027074997516638523,
+      "loss": 0.4756,
+      "step": 133200
+    },
+    {
+      "epoch": 6.616171649945366,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0002707102413827357,
+      "loss": 0.4732,
+      "step": 133210
+    },
+    {
+      "epoch": 6.616668322240986,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002706705075990861,
+      "loss": 0.4917,
+      "step": 133220
+    },
+    {
+      "epoch": 6.617164994536605,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0002706307738154366,
+      "loss": 0.4943,
+      "step": 133230
+    },
+    {
+      "epoch": 6.617661666832224,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00027059104003178706,
+      "loss": 0.4608,
+      "step": 133240
+    },
+    {
+      "epoch": 6.618158339127843,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002705513062481375,
+      "loss": 0.4674,
+      "step": 133250
+    },
+    {
+      "epoch": 6.618655011423463,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00027051157246448795,
+      "loss": 0.4938,
+      "step": 133260
+    },
+    {
+      "epoch": 6.619151683719082,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00027047183868083837,
+      "loss": 0.4672,
+      "step": 133270
+    },
+    {
+      "epoch": 6.619648356014702,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002704321048971889,
+      "loss": 0.5048,
+      "step": 133280
+    },
+    {
+      "epoch": 6.620145028310321,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0002703923711135393,
+      "loss": 0.4894,
+      "step": 133290
+    },
+    {
+      "epoch": 6.62064170060594,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0002703526373298897,
+      "loss": 0.4835,
+      "step": 133300
+    },
+    {
+      "epoch": 6.62113837290156,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002703129035462402,
+      "loss": 0.5134,
+      "step": 133310
+    },
+    {
+      "epoch": 6.621635045197179,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00027027316976259067,
+      "loss": 0.5127,
+      "step": 133320
+    },
+    {
+      "epoch": 6.622131717492798,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00027023343597894114,
+      "loss": 0.4695,
+      "step": 133330
+    },
+    {
+      "epoch": 6.622628389788417,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00027019370219529156,
+      "loss": 0.5054,
+      "step": 133340
+    },
+    {
+      "epoch": 6.623125062084037,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.000270153968411642,
+      "loss": 0.5493,
+      "step": 133350
+    },
+    {
+      "epoch": 6.623621734379657,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002701142346279925,
+      "loss": 0.4761,
+      "step": 133360
+    },
+    {
+      "epoch": 6.624118406675276,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002700745008443429,
+      "loss": 0.5189,
+      "step": 133370
+    },
+    {
+      "epoch": 6.624615078970895,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00027003476706069333,
+      "loss": 0.4894,
+      "step": 133380
+    },
+    {
+      "epoch": 6.625111751266514,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002699950332770438,
+      "loss": 0.5142,
+      "step": 133390
+    },
+    {
+      "epoch": 6.6256084235621335,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002699552994933943,
+      "loss": 0.4884,
+      "step": 133400
+    },
+    {
+      "epoch": 6.626105095857753,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00026991556570974475,
+      "loss": 0.4813,
+      "step": 133410
+    },
+    {
+      "epoch": 6.626601768153373,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00026987583192609517,
+      "loss": 0.5112,
+      "step": 133420
+    },
+    {
+      "epoch": 6.627098440448992,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00026983609814244564,
+      "loss": 0.5361,
+      "step": 133430
+    },
+    {
+      "epoch": 6.627595112744611,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002697963643587961,
+      "loss": 0.4426,
+      "step": 133440
+    },
+    {
+      "epoch": 6.6280917850402306,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002697566305751465,
+      "loss": 0.4809,
+      "step": 133450
+    },
+    {
+      "epoch": 6.62858845733585,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00026971689679149694,
+      "loss": 0.5134,
+      "step": 133460
+    },
+    {
+      "epoch": 6.629085129631469,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00026967716300784747,
+      "loss": 0.5242,
+      "step": 133470
+    },
+    {
+      "epoch": 6.629581801927088,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002696374292241979,
+      "loss": 0.4854,
+      "step": 133480
+    },
+    {
+      "epoch": 6.630078474222708,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00026959769544054836,
+      "loss": 0.4724,
+      "step": 133490
+    },
+    {
+      "epoch": 6.630575146518328,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002695579616568988,
+      "loss": 0.5376,
+      "step": 133500
+    },
+    {
+      "epoch": 6.631071818813947,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00026951822787324925,
+      "loss": 0.4984,
+      "step": 133510
+    },
+    {
+      "epoch": 6.631568491109566,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002694784940895997,
+      "loss": 0.5126,
+      "step": 133520
+    },
+    {
+      "epoch": 6.632065163405185,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00026943876030595013,
+      "loss": 0.511,
+      "step": 133530
+    },
+    {
+      "epoch": 6.6325618357008045,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002693990265223006,
+      "loss": 0.4826,
+      "step": 133540
+    },
+    {
+      "epoch": 6.633058507996424,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002693592927386511,
+      "loss": 0.4895,
+      "step": 133550
+    },
+    {
+      "epoch": 6.633555180292043,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002693195589550015,
+      "loss": 0.5184,
+      "step": 133560
+    },
+    {
+      "epoch": 6.634051852587662,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00026927982517135196,
+      "loss": 0.4698,
+      "step": 133570
+    },
+    {
+      "epoch": 6.634548524883282,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002692400913877024,
+      "loss": 0.5026,
+      "step": 133580
+    },
+    {
+      "epoch": 6.6350451971789015,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00026920035760405285,
+      "loss": 0.461,
+      "step": 133590
+    },
+    {
+      "epoch": 6.635541869474521,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002691606238204033,
+      "loss": 0.4852,
+      "step": 133600
+    },
+    {
+      "epoch": 6.63603854177014,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00026912089003675374,
+      "loss": 0.5103,
+      "step": 133610
+    },
+    {
+      "epoch": 6.636535214065759,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002690811562531042,
+      "loss": 0.4986,
+      "step": 133620
+    },
+    {
+      "epoch": 6.637031886361378,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002690414224694547,
+      "loss": 0.483,
+      "step": 133630
+    },
+    {
+      "epoch": 6.637528558656998,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002690016886858051,
+      "loss": 0.5236,
+      "step": 133640
+    },
+    {
+      "epoch": 6.638025230952618,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002689619549021556,
+      "loss": 0.4868,
+      "step": 133650
+    },
+    {
+      "epoch": 6.638521903248237,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00026892222111850604,
+      "loss": 0.4662,
+      "step": 133660
+    },
+    {
+      "epoch": 6.639018575543856,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00026888248733485646,
+      "loss": 0.4987,
+      "step": 133670
+    },
+    {
+      "epoch": 6.639515247839475,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00026884275355120693,
+      "loss": 0.4798,
+      "step": 133680
+    },
+    {
+      "epoch": 6.640011920135095,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00026880301976755735,
+      "loss": 0.5026,
+      "step": 133690
+    },
+    {
+      "epoch": 6.640508592430714,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0002687632859839078,
+      "loss": 0.5088,
+      "step": 133700
+    },
+    {
+      "epoch": 6.641005264726333,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002687235522002583,
+      "loss": 0.4896,
+      "step": 133710
+    },
+    {
+      "epoch": 6.641501937021953,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002686838184166087,
+      "loss": 0.5207,
+      "step": 133720
+    },
+    {
+      "epoch": 6.6419986093175725,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002686440846329592,
+      "loss": 0.5255,
+      "step": 133730
+    },
+    {
+      "epoch": 6.642495281613192,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00026860435084930965,
+      "loss": 0.4937,
+      "step": 133740
+    },
+    {
+      "epoch": 6.642991953908811,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00026856461706566007,
+      "loss": 0.4631,
+      "step": 133750
+    },
+    {
+      "epoch": 6.64348862620443,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00026852488328201054,
+      "loss": 0.5256,
+      "step": 133760
+    },
+    {
+      "epoch": 6.643985298500049,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.000268485149498361,
+      "loss": 0.505,
+      "step": 133770
+    },
+    {
+      "epoch": 6.644481970795669,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002684454157147115,
+      "loss": 0.4995,
+      "step": 133780
+    },
+    {
+      "epoch": 6.644978643091289,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002684056819310619,
+      "loss": 0.5255,
+      "step": 133790
+    },
+    {
+      "epoch": 6.645475315386908,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0002683659481474123,
+      "loss": 0.4723,
+      "step": 133800
+    },
+    {
+      "epoch": 6.645971987682527,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00026832621436376284,
+      "loss": 0.5012,
+      "step": 133810
+    },
+    {
+      "epoch": 6.646468659978146,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00026828648058011326,
+      "loss": 0.4947,
+      "step": 133820
+    },
+    {
+      "epoch": 6.646965332273766,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002682467467964637,
+      "loss": 0.5005,
+      "step": 133830
+    },
+    {
+      "epoch": 6.647462004569385,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00026820701301281415,
+      "loss": 0.5042,
+      "step": 133840
+    },
+    {
+      "epoch": 6.647958676865004,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0002681672792291646,
+      "loss": 0.4901,
+      "step": 133850
+    },
+    {
+      "epoch": 6.648455349160624,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002681275454455151,
+      "loss": 0.5007,
+      "step": 133860
+    },
+    {
+      "epoch": 6.648952021456243,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002680878116618655,
+      "loss": 0.4806,
+      "step": 133870
+    },
+    {
+      "epoch": 6.649448693751863,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002680480778782159,
+      "loss": 0.4836,
+      "step": 133880
+    },
+    {
+      "epoch": 6.649945366047482,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00026800834409456645,
+      "loss": 0.5106,
+      "step": 133890
+    },
+    {
+      "epoch": 6.650442038343101,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00026796861031091687,
+      "loss": 0.4862,
+      "step": 133900
+    },
+    {
+      "epoch": 6.65093871063872,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002679288765272673,
+      "loss": 0.4535,
+      "step": 133910
+    },
+    {
+      "epoch": 6.65143538293434,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00026788914274361776,
+      "loss": 0.5245,
+      "step": 133920
+    },
+    {
+      "epoch": 6.65193205522996,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00026784940895996823,
+      "loss": 0.5196,
+      "step": 133930
+    },
+    {
+      "epoch": 6.652428727525579,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002678096751763187,
+      "loss": 0.5149,
+      "step": 133940
+    },
+    {
+      "epoch": 6.652925399821198,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002677699413926691,
+      "loss": 0.4959,
+      "step": 133950
+    },
+    {
+      "epoch": 6.653422072116817,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002677302076090196,
+      "loss": 0.4859,
+      "step": 133960
+    },
+    {
+      "epoch": 6.653918744412437,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00026769047382537006,
+      "loss": 0.518,
+      "step": 133970
+    },
+    {
+      "epoch": 6.654415416708056,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002676507400417205,
+      "loss": 0.477,
+      "step": 133980
+    },
+    {
+      "epoch": 6.654912089003675,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0002676110062580709,
+      "loss": 0.483,
+      "step": 133990
+    },
+    {
+      "epoch": 6.655408761299295,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0002675712724744214,
+      "loss": 0.4983,
+      "step": 134000
+    },
+    {
+      "epoch": 6.655905433594914,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00026753153869077184,
+      "loss": 0.4666,
+      "step": 134010
+    },
+    {
+      "epoch": 6.656402105890534,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002674918049071223,
+      "loss": 0.5227,
+      "step": 134020
+    },
+    {
+      "epoch": 6.656898778186153,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002674520711234727,
+      "loss": 0.4894,
+      "step": 134030
+    },
+    {
+      "epoch": 6.657395450481772,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002674123373398232,
+      "loss": 0.4588,
+      "step": 134040
+    },
+    {
+      "epoch": 6.657892122777391,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00026737260355617367,
+      "loss": 0.4926,
+      "step": 134050
+    },
+    {
+      "epoch": 6.6583887950730105,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002673328697725241,
+      "loss": 0.4998,
+      "step": 134060
+    },
+    {
+      "epoch": 6.658885467368631,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00026729313598887456,
+      "loss": 0.5003,
+      "step": 134070
+    },
+    {
+      "epoch": 6.65938213966425,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00026725340220522503,
+      "loss": 0.5129,
+      "step": 134080
+    },
+    {
+      "epoch": 6.659878811959869,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00026721366842157545,
+      "loss": 0.4808,
+      "step": 134090
+    },
+    {
+      "epoch": 6.660375484255488,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002671739346379259,
+      "loss": 0.4807,
+      "step": 134100
+    },
+    {
+      "epoch": 6.6608721565511075,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002671342008542764,
+      "loss": 0.5135,
+      "step": 134110
+    },
+    {
+      "epoch": 6.661368828846727,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002670944670706268,
+      "loss": 0.4837,
+      "step": 134120
+    },
+    {
+      "epoch": 6.661865501142346,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002670547332869773,
+      "loss": 0.478,
+      "step": 134130
+    },
+    {
+      "epoch": 6.662362173437966,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002670149995033277,
+      "loss": 0.4888,
+      "step": 134140
+    },
+    {
+      "epoch": 6.662858845733585,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00026697526571967816,
+      "loss": 0.4739,
+      "step": 134150
+    },
+    {
+      "epoch": 6.663355518029205,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00026693553193602864,
+      "loss": 0.5055,
+      "step": 134160
+    },
+    {
+      "epoch": 6.663852190324824,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00026689579815237905,
+      "loss": 0.4849,
+      "step": 134170
+    },
+    {
+      "epoch": 6.664348862620443,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0002668560643687295,
+      "loss": 0.4776,
+      "step": 134180
+    },
+    {
+      "epoch": 6.664845534916062,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00026681633058508,
+      "loss": 0.5023,
+      "step": 134190
+    },
+    {
+      "epoch": 6.6653422072116815,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002667765968014304,
+      "loss": 0.4725,
+      "step": 134200
+    },
+    {
+      "epoch": 6.665838879507301,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002667368630177809,
+      "loss": 0.4813,
+      "step": 134210
+    },
+    {
+      "epoch": 6.666335551802921,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002666971292341313,
+      "loss": 0.4912,
+      "step": 134220
+    },
+    {
+      "epoch": 6.66683222409854,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00026665739545048183,
+      "loss": 0.5012,
+      "step": 134230
+    },
+    {
+      "epoch": 6.667328896394159,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00026661766166683224,
+      "loss": 0.4752,
+      "step": 134240
+    },
+    {
+      "epoch": 6.6678255686897785,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00026657792788318266,
+      "loss": 0.4683,
+      "step": 134250
+    },
+    {
+      "epoch": 6.668322240985398,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00026653819409953313,
+      "loss": 0.507,
+      "step": 134260
+    },
+    {
+      "epoch": 6.668818913281017,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002664984603158836,
+      "loss": 0.5005,
+      "step": 134270
+    },
+    {
+      "epoch": 6.669315585576636,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.000266458726532234,
+      "loss": 0.4964,
+      "step": 134280
+    },
+    {
+      "epoch": 6.669812257872255,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002664189927485845,
+      "loss": 0.5075,
+      "step": 134290
+    },
+    {
+      "epoch": 6.6703089301678755,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00026637925896493496,
+      "loss": 0.4932,
+      "step": 134300
+    },
+    {
+      "epoch": 6.670805602463495,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00026633952518128544,
+      "loss": 0.5096,
+      "step": 134310
+    },
+    {
+      "epoch": 6.671302274759114,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00026629979139763585,
+      "loss": 0.4881,
+      "step": 134320
+    },
+    {
+      "epoch": 6.671798947054733,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00026626005761398627,
+      "loss": 0.4874,
+      "step": 134330
+    },
+    {
+      "epoch": 6.672295619350352,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002662203238303368,
+      "loss": 0.486,
+      "step": 134340
+    },
+    {
+      "epoch": 6.672792291645972,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0002661805900466872,
+      "loss": 0.5361,
+      "step": 134350
+    },
+    {
+      "epoch": 6.673288963941591,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00026614085626303763,
+      "loss": 0.4694,
+      "step": 134360
+    },
+    {
+      "epoch": 6.673785636237211,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002661011224793881,
+      "loss": 0.4767,
+      "step": 134370
+    },
+    {
+      "epoch": 6.67428230853283,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00026606138869573857,
+      "loss": 0.4853,
+      "step": 134380
+    },
+    {
+      "epoch": 6.6747789808284494,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00026602165491208904,
+      "loss": 0.4636,
+      "step": 134390
+    },
+    {
+      "epoch": 6.675275653124069,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.00026598192112843946,
+      "loss": 0.4859,
+      "step": 134400
+    },
+    {
+      "epoch": 6.675772325419688,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00026594218734478993,
+      "loss": 0.5127,
+      "step": 134410
+    },
+    {
+      "epoch": 6.676268997715307,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002659024535611404,
+      "loss": 0.4887,
+      "step": 134420
+    },
+    {
+      "epoch": 6.676765670010926,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002658627197774908,
+      "loss": 0.5102,
+      "step": 134430
+    },
+    {
+      "epoch": 6.6772623423065465,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00026582298599384124,
+      "loss": 0.4927,
+      "step": 134440
+    },
+    {
+      "epoch": 6.677759014602166,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0002657832522101917,
+      "loss": 0.5023,
+      "step": 134450
+    },
+    {
+      "epoch": 6.678255686897785,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002657435184265422,
+      "loss": 0.4924,
+      "step": 134460
+    },
+    {
+      "epoch": 6.678752359193404,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00026570378464289265,
+      "loss": 0.4979,
+      "step": 134470
+    },
+    {
+      "epoch": 6.679249031489023,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00026566405085924307,
+      "loss": 0.4738,
+      "step": 134480
+    },
+    {
+      "epoch": 6.679745703784643,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00026562431707559354,
+      "loss": 0.4945,
+      "step": 134490
+    },
+    {
+      "epoch": 6.680242376080262,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.000265584583291944,
+      "loss": 0.4789,
+      "step": 134500
+    },
+    {
+      "epoch": 6.680739048375882,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00026554484950829443,
+      "loss": 0.4905,
+      "step": 134510
+    },
+    {
+      "epoch": 6.681235720671501,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00026550511572464485,
+      "loss": 0.4961,
+      "step": 134520
+    },
+    {
+      "epoch": 6.68173239296712,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00026546538194099537,
+      "loss": 0.5094,
+      "step": 134530
+    },
+    {
+      "epoch": 6.68222906526274,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002654256481573458,
+      "loss": 0.5259,
+      "step": 134540
+    },
+    {
+      "epoch": 6.682725737558359,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00026538591437369626,
+      "loss": 0.5188,
+      "step": 134550
+    },
+    {
+      "epoch": 6.683222409853978,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002653461805900467,
+      "loss": 0.4919,
+      "step": 134560
+    },
+    {
+      "epoch": 6.683719082149597,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00026530644680639715,
+      "loss": 0.4937,
+      "step": 134570
+    },
+    {
+      "epoch": 6.684215754445217,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002652667130227476,
+      "loss": 0.4981,
+      "step": 134580
+    },
+    {
+      "epoch": 6.684712426740837,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00026522697923909804,
+      "loss": 0.478,
+      "step": 134590
+    },
+    {
+      "epoch": 6.685209099036456,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002651872454554485,
+      "loss": 0.482,
+      "step": 134600
+    },
+    {
+      "epoch": 6.685705771332075,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.000265147511671799,
+      "loss": 0.4786,
+      "step": 134610
+    },
+    {
+      "epoch": 6.686202443627694,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002651077778881494,
+      "loss": 0.5206,
+      "step": 134620
+    },
+    {
+      "epoch": 6.686699115923314,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00026506804410449987,
+      "loss": 0.4924,
+      "step": 134630
+    },
+    {
+      "epoch": 6.687195788218933,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00026502831032085034,
+      "loss": 0.487,
+      "step": 134640
+    },
+    {
+      "epoch": 6.687692460514553,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00026498857653720076,
+      "loss": 0.4893,
+      "step": 134650
+    },
+    {
+      "epoch": 6.688189132810172,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00026494884275355123,
+      "loss": 0.5016,
+      "step": 134660
+    },
+    {
+      "epoch": 6.688685805105791,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00026490910896990165,
+      "loss": 0.48,
+      "step": 134670
+    },
+    {
+      "epoch": 6.689182477401411,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00026486937518625217,
+      "loss": 0.4858,
+      "step": 134680
+    },
+    {
+      "epoch": 6.68967914969703,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0002648296414026026,
+      "loss": 0.502,
+      "step": 134690
+    },
+    {
+      "epoch": 6.690175821992649,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.000264789907618953,
+      "loss": 0.4786,
+      "step": 134700
+    },
+    {
+      "epoch": 6.690672494288268,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002647501738353035,
+      "loss": 0.5083,
+      "step": 134710
+    },
+    {
+      "epoch": 6.691169166583888,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00026471044005165395,
+      "loss": 0.5071,
+      "step": 134720
+    },
+    {
+      "epoch": 6.691665838879508,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00026467070626800437,
+      "loss": 0.4903,
+      "step": 134730
+    },
+    {
+      "epoch": 6.692162511175127,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00026463097248435484,
+      "loss": 0.4844,
+      "step": 134740
+    },
+    {
+      "epoch": 6.692659183470746,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.00026459123870070525,
+      "loss": 0.5136,
+      "step": 134750
+    },
+    {
+      "epoch": 6.693155855766365,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002645515049170558,
+      "loss": 0.5069,
+      "step": 134760
+    },
+    {
+      "epoch": 6.6936525280619845,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002645117711334062,
+      "loss": 0.4855,
+      "step": 134770
+    },
+    {
+      "epoch": 6.694149200357604,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002644720373497566,
+      "loss": 0.4965,
+      "step": 134780
+    },
+    {
+      "epoch": 6.694645872653224,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002644323035661071,
+      "loss": 0.5055,
+      "step": 134790
+    },
+    {
+      "epoch": 6.695142544948843,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00026439256978245756,
+      "loss": 0.4824,
+      "step": 134800
+    },
+    {
+      "epoch": 6.695639217244462,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.000264352835998808,
+      "loss": 0.4634,
+      "step": 134810
+    },
+    {
+      "epoch": 6.6961358895400815,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00026431310221515844,
+      "loss": 0.4952,
+      "step": 134820
+    },
+    {
+      "epoch": 6.696632561835701,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002642733684315089,
+      "loss": 0.4851,
+      "step": 134830
+    },
+    {
+      "epoch": 6.69712923413132,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002642336346478594,
+      "loss": 0.4957,
+      "step": 134840
+    },
+    {
+      "epoch": 6.697625906426939,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0002641939008642098,
+      "loss": 0.5094,
+      "step": 134850
+    },
+    {
+      "epoch": 6.698122578722559,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002641541670805602,
+      "loss": 0.4804,
+      "step": 134860
+    },
+    {
+      "epoch": 6.698619251018179,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00026411443329691075,
+      "loss": 0.4908,
+      "step": 134870
+    },
+    {
+      "epoch": 6.699115923313798,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00026407469951326116,
+      "loss": 0.4931,
+      "step": 134880
+    },
+    {
+      "epoch": 6.699612595609417,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002640349657296116,
+      "loss": 0.4713,
+      "step": 134890
+    },
+    {
+      "epoch": 6.700109267905036,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00026399523194596205,
+      "loss": 0.4666,
+      "step": 134900
+    },
+    {
+      "epoch": 6.7006059402006555,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0002639554981623125,
+      "loss": 0.4869,
+      "step": 134910
+    },
+    {
+      "epoch": 6.701102612496275,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.000263915764378663,
+      "loss": 0.5096,
+      "step": 134920
+    },
+    {
+      "epoch": 6.701599284791894,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002638760305950134,
+      "loss": 0.4614,
+      "step": 134930
+    },
+    {
+      "epoch": 6.702095957087514,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002638362968113639,
+      "loss": 0.4738,
+      "step": 134940
+    },
+    {
+      "epoch": 6.702592629383133,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00026379656302771436,
+      "loss": 0.4953,
+      "step": 134950
+    },
+    {
+      "epoch": 6.7030893016787525,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00026375682924406477,
+      "loss": 0.5163,
+      "step": 134960
+    },
+    {
+      "epoch": 6.703585973974372,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002637170954604152,
+      "loss": 0.4771,
+      "step": 134970
+    },
+    {
+      "epoch": 6.704082646269991,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002636773616767657,
+      "loss": 0.5037,
+      "step": 134980
+    },
+    {
+      "epoch": 6.70457931856561,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00026363762789311613,
+      "loss": 0.4834,
+      "step": 134990
+    },
+    {
+      "epoch": 6.705075990861229,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002635978941094666,
+      "loss": 0.5574,
+      "step": 135000
+    },
+    {
+      "epoch": 6.705572663156849,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.000263558160325817,
+      "loss": 0.5143,
+      "step": 135010
+    },
+    {
+      "epoch": 6.706069335452469,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0002635184265421675,
+      "loss": 0.5004,
+      "step": 135020
+    },
+    {
+      "epoch": 6.706566007748088,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00026347869275851796,
+      "loss": 0.5572,
+      "step": 135030
+    },
+    {
+      "epoch": 6.707062680043707,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002634389589748684,
+      "loss": 0.4847,
+      "step": 135040
+    },
+    {
+      "epoch": 6.707559352339326,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00026339922519121885,
+      "loss": 0.4937,
+      "step": 135050
+    },
+    {
+      "epoch": 6.708056024634946,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002633594914075693,
+      "loss": 0.5069,
+      "step": 135060
+    },
+    {
+      "epoch": 6.708552696930565,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00026331975762391974,
+      "loss": 0.4741,
+      "step": 135070
+    },
+    {
+      "epoch": 6.709049369226184,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002632800238402702,
+      "loss": 0.4897,
+      "step": 135080
+    },
+    {
+      "epoch": 6.709546041521804,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00026324029005662063,
+      "loss": 0.4932,
+      "step": 135090
+    },
+    {
+      "epoch": 6.7100427138174235,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002632005562729711,
+      "loss": 0.4921,
+      "step": 135100
+    },
+    {
+      "epoch": 6.710539386113043,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00026316082248932157,
+      "loss": 0.4973,
+      "step": 135110
+    },
+    {
+      "epoch": 6.711036058408662,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.000263121088705672,
+      "loss": 0.5009,
+      "step": 135120
+    },
+    {
+      "epoch": 6.711532730704281,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00026308135492202246,
+      "loss": 0.501,
+      "step": 135130
+    },
+    {
+      "epoch": 6.7120294029999,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00026304162113837293,
+      "loss": 0.498,
+      "step": 135140
+    },
+    {
+      "epoch": 6.71252607529552,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00026300188735472335,
+      "loss": 0.5147,
+      "step": 135150
+    },
+    {
+      "epoch": 6.71302274759114,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002629621535710738,
+      "loss": 0.4895,
+      "step": 135160
+    },
+    {
+      "epoch": 6.713519419886759,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002629224197874243,
+      "loss": 0.4905,
+      "step": 135170
+    },
+    {
+      "epoch": 6.714016092182378,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002628826860037747,
+      "loss": 0.462,
+      "step": 135180
+    },
+    {
+      "epoch": 6.714512764477997,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0002628429522201252,
+      "loss": 0.5165,
+      "step": 135190
+    },
+    {
+      "epoch": 6.715009436773617,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002628032184364756,
+      "loss": 0.5055,
+      "step": 135200
+    },
+    {
+      "epoch": 6.715506109069236,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002627634846528261,
+      "loss": 0.4938,
+      "step": 135210
+    },
+    {
+      "epoch": 6.716002781364855,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00026272375086917654,
+      "loss": 0.4777,
+      "step": 135220
+    },
+    {
+      "epoch": 6.716499453660475,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00026268401708552696,
+      "loss": 0.4791,
+      "step": 135230
+    },
+    {
+      "epoch": 6.716996125956094,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00026264428330187743,
+      "loss": 0.4691,
+      "step": 135240
+    },
+    {
+      "epoch": 6.717492798251714,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002626045495182279,
+      "loss": 0.4933,
+      "step": 135250
+    },
+    {
+      "epoch": 6.717989470547333,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002625648157345783,
+      "loss": 0.5045,
+      "step": 135260
+    },
+    {
+      "epoch": 6.718486142842952,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002625250819509288,
+      "loss": 0.4828,
+      "step": 135270
+    },
+    {
+      "epoch": 6.718982815138571,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002624853481672792,
+      "loss": 0.4916,
+      "step": 135280
+    },
+    {
+      "epoch": 6.7194794874341905,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00026244561438362973,
+      "loss": 0.5101,
+      "step": 135290
+    },
+    {
+      "epoch": 6.719976159729811,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00026240588059998015,
+      "loss": 0.5071,
+      "step": 135300
+    },
+    {
+      "epoch": 6.72047283202543,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00026236614681633057,
+      "loss": 0.5014,
+      "step": 135310
+    },
+    {
+      "epoch": 6.720969504321049,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00026232641303268104,
+      "loss": 0.4723,
+      "step": 135320
+    },
+    {
+      "epoch": 6.721466176616668,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002622866792490315,
+      "loss": 0.4911,
+      "step": 135330
+    },
+    {
+      "epoch": 6.721962848912288,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0002622469454653819,
+      "loss": 0.5158,
+      "step": 135340
+    },
+    {
+      "epoch": 6.722459521207907,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0002622072116817324,
+      "loss": 0.4811,
+      "step": 135350
+    },
+    {
+      "epoch": 6.722956193503526,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00026216747789808287,
+      "loss": 0.5049,
+      "step": 135360
+    },
+    {
+      "epoch": 6.723452865799146,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00026212774411443334,
+      "loss": 0.5028,
+      "step": 135370
+    },
+    {
+      "epoch": 6.723949538094765,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00026208801033078376,
+      "loss": 0.5054,
+      "step": 135380
+    },
+    {
+      "epoch": 6.724446210390385,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002620482765471342,
+      "loss": 0.484,
+      "step": 135390
+    },
+    {
+      "epoch": 6.724942882686004,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002620085427634847,
+      "loss": 0.4932,
+      "step": 135400
+    },
+    {
+      "epoch": 6.725439554981623,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002619688089798351,
+      "loss": 0.525,
+      "step": 135410
+    },
+    {
+      "epoch": 6.725936227277242,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00026192907519618553,
+      "loss": 0.4812,
+      "step": 135420
+    },
+    {
+      "epoch": 6.7264328995728615,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.000261889341412536,
+      "loss": 0.4896,
+      "step": 135430
+    },
+    {
+      "epoch": 6.726929571868482,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002618496076288865,
+      "loss": 0.488,
+      "step": 135440
+    },
+    {
+      "epoch": 6.727426244164101,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00026180987384523695,
+      "loss": 0.5218,
+      "step": 135450
+    },
+    {
+      "epoch": 6.72792291645972,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00026177014006158736,
+      "loss": 0.4835,
+      "step": 135460
+    },
+    {
+      "epoch": 6.728419588755339,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00026173040627793784,
+      "loss": 0.5187,
+      "step": 135470
+    },
+    {
+      "epoch": 6.7289162610509585,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002616906724942883,
+      "loss": 0.4934,
+      "step": 135480
+    },
+    {
+      "epoch": 6.729412933346578,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0002616509387106387,
+      "loss": 0.4806,
+      "step": 135490
+    },
+    {
+      "epoch": 6.729909605642197,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002616112049269892,
+      "loss": 0.496,
+      "step": 135500
+    },
+    {
+      "epoch": 6.730406277937817,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00026157147114333967,
+      "loss": 0.5232,
+      "step": 135510
+    },
+    {
+      "epoch": 6.730902950233436,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002615317373596901,
+      "loss": 0.4889,
+      "step": 135520
+    },
+    {
+      "epoch": 6.7313996225290555,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00026149200357604056,
+      "loss": 0.5107,
+      "step": 135530
+    },
+    {
+      "epoch": 6.731896294824675,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00026145226979239097,
+      "loss": 0.5284,
+      "step": 135540
+    },
+    {
+      "epoch": 6.732392967120294,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00026141253600874144,
+      "loss": 0.4982,
+      "step": 135550
+    },
+    {
+      "epoch": 6.732889639415913,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002613728022250919,
+      "loss": 0.5103,
+      "step": 135560
+    },
+    {
+      "epoch": 6.7333863117115325,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00026133306844144233,
+      "loss": 0.4855,
+      "step": 135570
+    },
+    {
+      "epoch": 6.733882984007153,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002612933346577928,
+      "loss": 0.5072,
+      "step": 135580
+    },
+    {
+      "epoch": 6.734379656302772,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0002612536008741433,
+      "loss": 0.4879,
+      "step": 135590
+    },
+    {
+      "epoch": 6.734876328598391,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002612138670904937,
+      "loss": 0.5167,
+      "step": 135600
+    },
+    {
+      "epoch": 6.73537300089401,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00026117413330684416,
+      "loss": 0.4997,
+      "step": 135610
+    },
+    {
+      "epoch": 6.7358696731896295,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002611343995231946,
+      "loss": 0.5016,
+      "step": 135620
+    },
+    {
+      "epoch": 6.736366345485249,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00026109466573954505,
+      "loss": 0.5042,
+      "step": 135630
+    },
+    {
+      "epoch": 6.736863017780868,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002610549319558955,
+      "loss": 0.5122,
+      "step": 135640
+    },
+    {
+      "epoch": 6.737359690076487,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00026101519817224594,
+      "loss": 0.5026,
+      "step": 135650
+    },
+    {
+      "epoch": 6.737856362372106,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002609754643885964,
+      "loss": 0.4977,
+      "step": 135660
+    },
+    {
+      "epoch": 6.7383530346677265,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0002609357306049469,
+      "loss": 0.452,
+      "step": 135670
+    },
+    {
+      "epoch": 6.738849706963346,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002608959968212973,
+      "loss": 0.4967,
+      "step": 135680
+    },
+    {
+      "epoch": 6.739346379258965,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00026085626303764777,
+      "loss": 0.5096,
+      "step": 135690
+    },
+    {
+      "epoch": 6.739843051554584,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00026081652925399824,
+      "loss": 0.5105,
+      "step": 135700
+    },
+    {
+      "epoch": 6.740339723850203,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00026077679547034866,
+      "loss": 0.4877,
+      "step": 135710
+    },
+    {
+      "epoch": 6.740836396145823,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00026073706168669913,
+      "loss": 0.4725,
+      "step": 135720
+    },
+    {
+      "epoch": 6.741333068441442,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00026069732790304955,
+      "loss": 0.4899,
+      "step": 135730
+    },
+    {
+      "epoch": 6.741829740737062,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002606575941194001,
+      "loss": 0.5124,
+      "step": 135740
+    },
+    {
+      "epoch": 6.742326413032681,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002606178603357505,
+      "loss": 0.5151,
+      "step": 135750
+    },
+    {
+      "epoch": 6.7428230853283,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0002605781265521009,
+      "loss": 0.4804,
+      "step": 135760
+    },
+    {
+      "epoch": 6.74331975762392,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002605383927684514,
+      "loss": 0.4874,
+      "step": 135770
+    },
+    {
+      "epoch": 6.743816429919539,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00026049865898480185,
+      "loss": 0.5055,
+      "step": 135780
+    },
+    {
+      "epoch": 6.744313102215158,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00026045892520115227,
+      "loss": 0.4592,
+      "step": 135790
+    },
+    {
+      "epoch": 6.744809774510777,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00026041919141750274,
+      "loss": 0.4691,
+      "step": 135800
+    },
+    {
+      "epoch": 6.7453064468063975,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002603794576338532,
+      "loss": 0.5086,
+      "step": 135810
+    },
+    {
+      "epoch": 6.745803119102017,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0002603397238502037,
+      "loss": 0.5025,
+      "step": 135820
+    },
+    {
+      "epoch": 6.746299791397636,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002602999900665541,
+      "loss": 0.5012,
+      "step": 135830
+    },
+    {
+      "epoch": 6.746796463693255,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002602602562829045,
+      "loss": 0.4888,
+      "step": 135840
+    },
+    {
+      "epoch": 6.747293135988874,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.000260220522499255,
+      "loss": 0.4955,
+      "step": 135850
+    },
+    {
+      "epoch": 6.747789808284494,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00026018078871560546,
+      "loss": 0.5009,
+      "step": 135860
+    },
+    {
+      "epoch": 6.748286480580113,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00026014105493195593,
+      "loss": 0.5116,
+      "step": 135870
+    },
+    {
+      "epoch": 6.748783152875733,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00026010132114830635,
+      "loss": 0.5054,
+      "step": 135880
+    },
+    {
+      "epoch": 6.749279825171352,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002600615873646568,
+      "loss": 0.4901,
+      "step": 135890
+    },
+    {
+      "epoch": 6.749776497466971,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0002600218535810073,
+      "loss": 0.5006,
+      "step": 135900
+    },
+    {
+      "epoch": 6.750273169762591,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002599821197973577,
+      "loss": 0.4814,
+      "step": 135910
+    },
+    {
+      "epoch": 6.75076984205821,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002599423860137081,
+      "loss": 0.4964,
+      "step": 135920
+    },
+    {
+      "epoch": 6.751266514353829,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00025990265223005865,
+      "loss": 0.5052,
+      "step": 135930
+    },
+    {
+      "epoch": 6.751763186649448,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00025986291844640907,
+      "loss": 0.4758,
+      "step": 135940
+    },
+    {
+      "epoch": 6.752259858945068,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00025982318466275954,
+      "loss": 0.471,
+      "step": 135950
+    },
+    {
+      "epoch": 6.752756531240688,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00025978345087910996,
+      "loss": 0.4947,
+      "step": 135960
+    },
+    {
+      "epoch": 6.753253203536307,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00025974371709546043,
+      "loss": 0.5145,
+      "step": 135970
+    },
+    {
+      "epoch": 6.753749875831926,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002597039833118109,
+      "loss": 0.5098,
+      "step": 135980
+    },
+    {
+      "epoch": 6.754246548127545,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002596642495281613,
+      "loss": 0.4997,
+      "step": 135990
+    },
+    {
+      "epoch": 6.7547432204231646,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002596245157445118,
+      "loss": 0.4737,
+      "step": 136000
+    },
+    {
+      "epoch": 6.755239892718784,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00025958478196086226,
+      "loss": 0.5139,
+      "step": 136010
+    },
+    {
+      "epoch": 6.755736565014404,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002595450481772127,
+      "loss": 0.5038,
+      "step": 136020
+    },
+    {
+      "epoch": 6.756233237310023,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00025950531439356315,
+      "loss": 0.5041,
+      "step": 136030
+    },
+    {
+      "epoch": 6.756729909605642,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002594655806099136,
+      "loss": 0.5137,
+      "step": 136040
+    },
+    {
+      "epoch": 6.757226581901262,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00025942584682626404,
+      "loss": 0.4855,
+      "step": 136050
+    },
+    {
+      "epoch": 6.757723254196881,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002593861130426145,
+      "loss": 0.4702,
+      "step": 136060
+    },
+    {
+      "epoch": 6.7582199264925,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002593463792589649,
+      "loss": 0.5077,
+      "step": 136070
+    },
+    {
+      "epoch": 6.758716598788119,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002593066454753154,
+      "loss": 0.4921,
+      "step": 136080
+    },
+    {
+      "epoch": 6.759213271083739,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00025926691169166587,
+      "loss": 0.494,
+      "step": 136090
+    },
+    {
+      "epoch": 6.759709943379359,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0002592271779080163,
+      "loss": 0.4766,
+      "step": 136100
+    },
+    {
+      "epoch": 6.760206615674978,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00025918744412436676,
+      "loss": 0.4798,
+      "step": 136110
+    },
+    {
+      "epoch": 6.760703287970597,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00025914771034071723,
+      "loss": 0.4939,
+      "step": 136120
+    },
+    {
+      "epoch": 6.761199960266216,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00025910797655706764,
+      "loss": 0.5104,
+      "step": 136130
+    },
+    {
+      "epoch": 6.7616966325618355,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002590682427734181,
+      "loss": 0.5014,
+      "step": 136140
+    },
+    {
+      "epoch": 6.762193304857455,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00025902850898976853,
+      "loss": 0.4916,
+      "step": 136150
+    },
+    {
+      "epoch": 6.762689977153075,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.000258988775206119,
+      "loss": 0.5142,
+      "step": 136160
+    },
+    {
+      "epoch": 6.763186649448694,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002589490414224695,
+      "loss": 0.4729,
+      "step": 136170
+    },
+    {
+      "epoch": 6.763683321744313,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002589093076388199,
+      "loss": 0.4935,
+      "step": 136180
+    },
+    {
+      "epoch": 6.7641799940399325,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00025886957385517036,
+      "loss": 0.4747,
+      "step": 136190
+    },
+    {
+      "epoch": 6.764676666335552,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00025882984007152084,
+      "loss": 0.4835,
+      "step": 136200
+    },
+    {
+      "epoch": 6.765173338631171,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00025879010628787125,
+      "loss": 0.453,
+      "step": 136210
+    },
+    {
+      "epoch": 6.76567001092679,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002587503725042217,
+      "loss": 0.5033,
+      "step": 136220
+    },
+    {
+      "epoch": 6.76616668322241,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002587106387205722,
+      "loss": 0.5052,
+      "step": 136230
+    },
+    {
+      "epoch": 6.7666633555180296,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002586709049369226,
+      "loss": 0.4999,
+      "step": 136240
+    },
+    {
+      "epoch": 6.767160027813649,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002586311711532731,
+      "loss": 0.4879,
+      "step": 136250
+    },
+    {
+      "epoch": 6.767656700109268,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002585914373696235,
+      "loss": 0.5078,
+      "step": 136260
+    },
+    {
+      "epoch": 6.768153372404887,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.000258551703585974,
+      "loss": 0.5187,
+      "step": 136270
+    },
+    {
+      "epoch": 6.7686500447005065,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00025851196980232444,
+      "loss": 0.4798,
+      "step": 136280
+    },
+    {
+      "epoch": 6.769146716996126,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00025847223601867486,
+      "loss": 0.4725,
+      "step": 136290
+    },
+    {
+      "epoch": 6.769643389291746,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00025843250223502533,
+      "loss": 0.4921,
+      "step": 136300
+    },
+    {
+      "epoch": 6.770140061587365,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002583927684513758,
+      "loss": 0.5116,
+      "step": 136310
+    },
+    {
+      "epoch": 6.770636733882984,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002583530346677263,
+      "loss": 0.472,
+      "step": 136320
+    },
+    {
+      "epoch": 6.7711334061786035,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002583133008840767,
+      "loss": 0.4923,
+      "step": 136330
+    },
+    {
+      "epoch": 6.771630078474223,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00025827356710042716,
+      "loss": 0.4859,
+      "step": 136340
+    },
+    {
+      "epoch": 6.772126750769842,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00025823383331677763,
+      "loss": 0.4872,
+      "step": 136350
+    },
+    {
+      "epoch": 6.772623423065461,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00025819409953312805,
+      "loss": 0.5,
+      "step": 136360
+    },
+    {
+      "epoch": 6.77312009536108,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00025815436574947847,
+      "loss": 0.4895,
+      "step": 136370
+    },
+    {
+      "epoch": 6.7736167676567,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.000258114631965829,
+      "loss": 0.494,
+      "step": 136380
+    },
+    {
+      "epoch": 6.77411343995232,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002580748981821794,
+      "loss": 0.47,
+      "step": 136390
+    },
+    {
+      "epoch": 6.774610112247939,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002580351643985299,
+      "loss": 0.4972,
+      "step": 136400
+    },
+    {
+      "epoch": 6.775106784543558,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002579954306148803,
+      "loss": 0.495,
+      "step": 136410
+    },
+    {
+      "epoch": 6.775603456839177,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00025795569683123077,
+      "loss": 0.5207,
+      "step": 136420
+    },
+    {
+      "epoch": 6.776100129134797,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00025791596304758124,
+      "loss": 0.5088,
+      "step": 136430
+    },
+    {
+      "epoch": 6.776596801430416,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00025787622926393166,
+      "loss": 0.5184,
+      "step": 136440
+    },
+    {
+      "epoch": 6.777093473726035,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002578364954802821,
+      "loss": 0.4857,
+      "step": 136450
+    },
+    {
+      "epoch": 6.777590146021655,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002577967616966326,
+      "loss": 0.4996,
+      "step": 136460
+    },
+    {
+      "epoch": 6.778086818317274,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.000257757027912983,
+      "loss": 0.4963,
+      "step": 136470
+    },
+    {
+      "epoch": 6.778583490612894,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0002577172941293335,
+      "loss": 0.4915,
+      "step": 136480
+    },
+    {
+      "epoch": 6.779080162908513,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002576775603456839,
+      "loss": 0.5114,
+      "step": 136490
+    },
+    {
+      "epoch": 6.779576835204132,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002576378265620344,
+      "loss": 0.4829,
+      "step": 136500
+    },
+    {
+      "epoch": 6.780073507499751,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00025759809277838485,
+      "loss": 0.4805,
+      "step": 136510
+    },
+    {
+      "epoch": 6.780570179795371,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00025755835899473527,
+      "loss": 0.4742,
+      "step": 136520
+    },
+    {
+      "epoch": 6.781066852090991,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00025751862521108574,
+      "loss": 0.4905,
+      "step": 136530
+    },
+    {
+      "epoch": 6.78156352438661,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002574788914274362,
+      "loss": 0.5023,
+      "step": 136540
+    },
+    {
+      "epoch": 6.782060196682229,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00025743915764378663,
+      "loss": 0.4985,
+      "step": 136550
+    },
+    {
+      "epoch": 6.782556868977848,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002573994238601371,
+      "loss": 0.5227,
+      "step": 136560
+    },
+    {
+      "epoch": 6.783053541273468,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00025735969007648757,
+      "loss": 0.494,
+      "step": 136570
+    },
+    {
+      "epoch": 6.783550213569087,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.000257319956292838,
+      "loss": 0.5016,
+      "step": 136580
+    },
+    {
+      "epoch": 6.784046885864706,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00025728022250918846,
+      "loss": 0.4944,
+      "step": 136590
+    },
+    {
+      "epoch": 6.784543558160326,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002572404887255389,
+      "loss": 0.4921,
+      "step": 136600
+    },
+    {
+      "epoch": 6.785040230455945,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00025720075494188935,
+      "loss": 0.4741,
+      "step": 136610
+    },
+    {
+      "epoch": 6.785536902751565,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002571610211582398,
+      "loss": 0.4933,
+      "step": 136620
+    },
+    {
+      "epoch": 6.786033575047184,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00025712128737459024,
+      "loss": 0.4757,
+      "step": 136630
+    },
+    {
+      "epoch": 6.786530247342803,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002570815535909407,
+      "loss": 0.4882,
+      "step": 136640
+    },
+    {
+      "epoch": 6.787026919638422,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002570418198072912,
+      "loss": 0.4844,
+      "step": 136650
+    },
+    {
+      "epoch": 6.7875235919340415,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002570020860236416,
+      "loss": 0.5032,
+      "step": 136660
+    },
+    {
+      "epoch": 6.788020264229662,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00025696235223999207,
+      "loss": 0.4978,
+      "step": 136670
+    },
+    {
+      "epoch": 6.788516936525281,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00025692261845634254,
+      "loss": 0.5073,
+      "step": 136680
+    },
+    {
+      "epoch": 6.7890136088209,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.000256882884672693,
+      "loss": 0.4952,
+      "step": 136690
+    },
+    {
+      "epoch": 6.789510281116519,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00025684315088904343,
+      "loss": 0.5072,
+      "step": 136700
+    },
+    {
+      "epoch": 6.790006953412139,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00025680341710539384,
+      "loss": 0.4873,
+      "step": 136710
+    },
+    {
+      "epoch": 6.790503625707758,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002567636833217443,
+      "loss": 0.5001,
+      "step": 136720
+    },
+    {
+      "epoch": 6.791000298003377,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0002567239495380948,
+      "loss": 0.4856,
+      "step": 136730
+    },
+    {
+      "epoch": 6.791496970298997,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002566842157544452,
+      "loss": 0.4997,
+      "step": 136740
+    },
+    {
+      "epoch": 6.791993642594616,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002566444819707957,
+      "loss": 0.4973,
+      "step": 136750
+    },
+    {
+      "epoch": 6.792490314890236,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00025660474818714615,
+      "loss": 0.5042,
+      "step": 136760
+    },
+    {
+      "epoch": 6.792986987185855,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002565650144034966,
+      "loss": 0.5157,
+      "step": 136770
+    },
+    {
+      "epoch": 6.793483659481474,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00025652528061984704,
+      "loss": 0.5132,
+      "step": 136780
+    },
+    {
+      "epoch": 6.793980331777093,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00025648554683619745,
+      "loss": 0.4975,
+      "step": 136790
+    },
+    {
+      "epoch": 6.7944770040727125,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.000256445813052548,
+      "loss": 0.5163,
+      "step": 136800
+    },
+    {
+      "epoch": 6.794973676368333,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002564060792688984,
+      "loss": 0.4743,
+      "step": 136810
+    },
+    {
+      "epoch": 6.795470348663952,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002563663454852488,
+      "loss": 0.4745,
+      "step": 136820
+    },
+    {
+      "epoch": 6.795967020959571,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002563266117015993,
+      "loss": 0.5037,
+      "step": 136830
+    },
+    {
+      "epoch": 6.79646369325519,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00025628687791794976,
+      "loss": 0.453,
+      "step": 136840
+    },
+    {
+      "epoch": 6.7969603655508095,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002562471441343002,
+      "loss": 0.4808,
+      "step": 136850
+    },
+    {
+      "epoch": 6.797457037846429,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00025620741035065064,
+      "loss": 0.4868,
+      "step": 136860
+    },
+    {
+      "epoch": 6.797953710142048,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002561676765670011,
+      "loss": 0.4915,
+      "step": 136870
+    },
+    {
+      "epoch": 6.798450382437668,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002561279427833516,
+      "loss": 0.4938,
+      "step": 136880
+    },
+    {
+      "epoch": 6.798947054733287,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000256088208999702,
+      "loss": 0.4555,
+      "step": 136890
+    },
+    {
+      "epoch": 6.7994437270289065,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002560484752160524,
+      "loss": 0.5093,
+      "step": 136900
+    },
+    {
+      "epoch": 6.799940399324526,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00025600874143240295,
+      "loss": 0.4832,
+      "step": 136910
+    },
+    {
+      "epoch": 6.800437071620145,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00025596900764875336,
+      "loss": 0.4586,
+      "step": 136920
+    },
+    {
+      "epoch": 6.800933743915764,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00025592927386510384,
+      "loss": 0.4659,
+      "step": 136930
+    },
+    {
+      "epoch": 6.8014304162113834,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00025588954008145425,
+      "loss": 0.4829,
+      "step": 136940
+    },
+    {
+      "epoch": 6.801927088507004,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002558498062978047,
+      "loss": 0.4846,
+      "step": 136950
+    },
+    {
+      "epoch": 6.802423760802623,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002558100725141552,
+      "loss": 0.5186,
+      "step": 136960
+    },
+    {
+      "epoch": 6.802920433098242,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002557703387305056,
+      "loss": 0.4911,
+      "step": 136970
+    },
+    {
+      "epoch": 6.803417105393861,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00025573060494685603,
+      "loss": 0.4802,
+      "step": 136980
+    },
+    {
+      "epoch": 6.8039137776894805,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00025569087116320655,
+      "loss": 0.5097,
+      "step": 136990
+    },
+    {
+      "epoch": 6.8044104499851,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00025565113737955697,
+      "loss": 0.4759,
+      "step": 137000
+    },
+    {
+      "epoch": 6.804907122280719,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00025561140359590744,
+      "loss": 0.4875,
+      "step": 137010
+    },
+    {
+      "epoch": 6.805403794576338,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00025557166981225786,
+      "loss": 0.5124,
+      "step": 137020
+    },
+    {
+      "epoch": 6.805900466871958,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00025553193602860833,
+      "loss": 0.4782,
+      "step": 137030
+    },
+    {
+      "epoch": 6.8063971391675775,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002554922022449588,
+      "loss": 0.476,
+      "step": 137040
+    },
+    {
+      "epoch": 6.806893811463197,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002554524684613092,
+      "loss": 0.4942,
+      "step": 137050
+    },
+    {
+      "epoch": 6.807390483758816,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002554127346776597,
+      "loss": 0.5152,
+      "step": 137060
+    },
+    {
+      "epoch": 6.807887156054435,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00025537300089401016,
+      "loss": 0.5123,
+      "step": 137070
+    },
+    {
+      "epoch": 6.808383828350054,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0002553332671103606,
+      "loss": 0.4966,
+      "step": 137080
+    },
+    {
+      "epoch": 6.808880500645674,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00025529353332671105,
+      "loss": 0.4836,
+      "step": 137090
+    },
+    {
+      "epoch": 6.809377172941293,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002552537995430615,
+      "loss": 0.4629,
+      "step": 137100
+    },
+    {
+      "epoch": 6.809873845236913,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00025521406575941194,
+      "loss": 0.4719,
+      "step": 137110
+    },
+    {
+      "epoch": 6.810370517532532,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002551743319757624,
+      "loss": 0.4963,
+      "step": 137120
+    },
+    {
+      "epoch": 6.810867189828151,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00025513459819211283,
+      "loss": 0.4692,
+      "step": 137130
+    },
+    {
+      "epoch": 6.811363862123771,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00025509486440846335,
+      "loss": 0.5022,
+      "step": 137140
+    },
+    {
+      "epoch": 6.81186053441939,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00025505513062481377,
+      "loss": 0.463,
+      "step": 137150
+    },
+    {
+      "epoch": 6.812357206715009,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002550153968411642,
+      "loss": 0.5002,
+      "step": 137160
+    },
+    {
+      "epoch": 6.812853879010628,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00025497566305751466,
+      "loss": 0.4814,
+      "step": 137170
+    },
+    {
+      "epoch": 6.8133505513062484,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00025493592927386513,
+      "loss": 0.5067,
+      "step": 137180
+    },
+    {
+      "epoch": 6.813847223601868,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00025489619549021555,
+      "loss": 0.4942,
+      "step": 137190
+    },
+    {
+      "epoch": 6.814343895897487,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.000254856461706566,
+      "loss": 0.5138,
+      "step": 137200
+    },
+    {
+      "epoch": 6.814840568193106,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002548167279229165,
+      "loss": 0.4945,
+      "step": 137210
+    },
+    {
+      "epoch": 6.815337240488725,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00025477699413926696,
+      "loss": 0.4939,
+      "step": 137220
+    },
+    {
+      "epoch": 6.815833912784345,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002547372603556174,
+      "loss": 0.5035,
+      "step": 137230
+    },
+    {
+      "epoch": 6.816330585079964,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002546975265719678,
+      "loss": 0.4963,
+      "step": 137240
+    },
+    {
+      "epoch": 6.816827257375584,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00025465779278831827,
+      "loss": 0.4881,
+      "step": 137250
+    },
+    {
+      "epoch": 6.817323929671203,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.00025461805900466874,
+      "loss": 0.4813,
+      "step": 137260
+    },
+    {
+      "epoch": 6.817820601966822,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00025457832522101916,
+      "loss": 0.5031,
+      "step": 137270
+    },
+    {
+      "epoch": 6.818317274262442,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00025453859143736963,
+      "loss": 0.4995,
+      "step": 137280
+    },
+    {
+      "epoch": 6.818813946558061,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0002544988576537201,
+      "loss": 0.4715,
+      "step": 137290
+    },
+    {
+      "epoch": 6.81931061885368,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00025445912387007057,
+      "loss": 0.4713,
+      "step": 137300
+    },
+    {
+      "epoch": 6.819807291149299,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.000254419390086421,
+      "loss": 0.4607,
+      "step": 137310
+    },
+    {
+      "epoch": 6.820303963444919,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002543796563027714,
+      "loss": 0.4896,
+      "step": 137320
+    },
+    {
+      "epoch": 6.820800635740539,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00025433992251912193,
+      "loss": 0.4689,
+      "step": 137330
+    },
+    {
+      "epoch": 6.821297308036158,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00025430018873547235,
+      "loss": 0.51,
+      "step": 137340
+    },
+    {
+      "epoch": 6.821793980331777,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00025426045495182276,
+      "loss": 0.4849,
+      "step": 137350
+    },
+    {
+      "epoch": 6.822290652627396,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00025422072116817324,
+      "loss": 0.5056,
+      "step": 137360
+    },
+    {
+      "epoch": 6.8227873249230155,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0002541809873845237,
+      "loss": 0.4617,
+      "step": 137370
+    },
+    {
+      "epoch": 6.823283997218635,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002541412536008742,
+      "loss": 0.5141,
+      "step": 137380
+    },
+    {
+      "epoch": 6.823780669514255,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002541015198172246,
+      "loss": 0.487,
+      "step": 137390
+    },
+    {
+      "epoch": 6.824277341809874,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00025406178603357507,
+      "loss": 0.5098,
+      "step": 137400
+    },
+    {
+      "epoch": 6.824774014105493,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00025402205224992554,
+      "loss": 0.5122,
+      "step": 137410
+    },
+    {
+      "epoch": 6.825270686401113,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00025398231846627596,
+      "loss": 0.5249,
+      "step": 137420
+    },
+    {
+      "epoch": 6.825767358696732,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00025394258468262637,
+      "loss": 0.473,
+      "step": 137430
+    },
+    {
+      "epoch": 6.826264030992351,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002539028508989769,
+      "loss": 0.515,
+      "step": 137440
+    },
+    {
+      "epoch": 6.82676070328797,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002538631171153273,
+      "loss": 0.4566,
+      "step": 137450
+    },
+    {
+      "epoch": 6.82725737558359,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002538233833316778,
+      "loss": 0.47,
+      "step": 137460
+    },
+    {
+      "epoch": 6.82775404787921,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002537836495480282,
+      "loss": 0.5116,
+      "step": 137470
+    },
+    {
+      "epoch": 6.828250720174829,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002537439157643787,
+      "loss": 0.4773,
+      "step": 137480
+    },
+    {
+      "epoch": 6.828747392470448,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00025370418198072915,
+      "loss": 0.5096,
+      "step": 137490
+    },
+    {
+      "epoch": 6.829244064766067,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00025366444819707956,
+      "loss": 0.4903,
+      "step": 137500
+    },
+    {
+      "epoch": 6.8297407370616865,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00025362471441343004,
+      "loss": 0.5225,
+      "step": 137510
+    },
+    {
+      "epoch": 6.830237409357306,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002535849806297805,
+      "loss": 0.5026,
+      "step": 137520
+    },
+    {
+      "epoch": 6.830734081652926,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002535452468461309,
+      "loss": 0.5412,
+      "step": 137530
+    },
+    {
+      "epoch": 6.831230753948545,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002535055130624814,
+      "loss": 0.5015,
+      "step": 137540
+    },
+    {
+      "epoch": 6.831727426244164,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002534657792788318,
+      "loss": 0.5025,
+      "step": 137550
+    },
+    {
+      "epoch": 6.8322240985397835,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002534260454951823,
+      "loss": 0.4904,
+      "step": 137560
+    },
+    {
+      "epoch": 6.832720770835403,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00025338631171153275,
+      "loss": 0.4763,
+      "step": 137570
+    },
+    {
+      "epoch": 6.833217443131022,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00025334657792788317,
+      "loss": 0.4824,
+      "step": 137580
+    },
+    {
+      "epoch": 6.833714115426641,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00025330684414423364,
+      "loss": 0.5054,
+      "step": 137590
+    },
+    {
+      "epoch": 6.834210787722261,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002532671103605841,
+      "loss": 0.4908,
+      "step": 137600
+    },
+    {
+      "epoch": 6.8347074600178805,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00025322737657693453,
+      "loss": 0.4919,
+      "step": 137610
+    },
+    {
+      "epoch": 6.8352041323135,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.000253187642793285,
+      "loss": 0.5096,
+      "step": 137620
+    },
+    {
+      "epoch": 6.835700804609119,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002531479090096355,
+      "loss": 0.4745,
+      "step": 137630
+    },
+    {
+      "epoch": 6.836197476904738,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002531081752259859,
+      "loss": 0.4789,
+      "step": 137640
+    },
+    {
+      "epoch": 6.8366941492003575,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00025306844144233636,
+      "loss": 0.5087,
+      "step": 137650
+    },
+    {
+      "epoch": 6.837190821495977,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002530287076586868,
+      "loss": 0.5004,
+      "step": 137660
+    },
+    {
+      "epoch": 6.837687493791597,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002529889738750373,
+      "loss": 0.4793,
+      "step": 137670
+    },
+    {
+      "epoch": 6.838184166087216,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002529492400913877,
+      "loss": 0.5153,
+      "step": 137680
+    },
+    {
+      "epoch": 6.838680838382835,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.00025290950630773814,
+      "loss": 0.4701,
+      "step": 137690
+    },
+    {
+      "epoch": 6.8391775106784545,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002528697725240886,
+      "loss": 0.4855,
+      "step": 137700
+    },
+    {
+      "epoch": 6.839674182974074,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002528300387404391,
+      "loss": 0.5083,
+      "step": 137710
+    },
+    {
+      "epoch": 6.840170855269693,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002527903049567895,
+      "loss": 0.4645,
+      "step": 137720
+    },
+    {
+      "epoch": 6.840667527565312,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00025275057117313997,
+      "loss": 0.497,
+      "step": 137730
+    },
+    {
+      "epoch": 6.841164199860931,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00025271083738949044,
+      "loss": 0.4876,
+      "step": 137740
+    },
+    {
+      "epoch": 6.841660872156551,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002526711036058409,
+      "loss": 0.488,
+      "step": 137750
+    },
+    {
+      "epoch": 6.842157544452171,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00025263136982219133,
+      "loss": 0.5133,
+      "step": 137760
+    },
+    {
+      "epoch": 6.84265421674779,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00025259163603854175,
+      "loss": 0.4872,
+      "step": 137770
+    },
+    {
+      "epoch": 6.843150889043409,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002525519022548923,
+      "loss": 0.5093,
+      "step": 137780
+    },
+    {
+      "epoch": 6.843647561339028,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002525121684712427,
+      "loss": 0.5031,
+      "step": 137790
+    },
+    {
+      "epoch": 6.844144233634648,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002524724346875931,
+      "loss": 0.4943,
+      "step": 137800
+    },
+    {
+      "epoch": 6.844640905930267,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002524327009039436,
+      "loss": 0.4916,
+      "step": 137810
+    },
+    {
+      "epoch": 6.845137578225886,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00025239296712029405,
+      "loss": 0.5025,
+      "step": 137820
+    },
+    {
+      "epoch": 6.845634250521506,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002523532333366445,
+      "loss": 0.5072,
+      "step": 137830
+    },
+    {
+      "epoch": 6.846130922817125,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00025231349955299494,
+      "loss": 0.5001,
+      "step": 137840
+    },
+    {
+      "epoch": 6.846627595112745,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00025227376576934536,
+      "loss": 0.4927,
+      "step": 137850
+    },
+    {
+      "epoch": 6.847124267408364,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002522340319856959,
+      "loss": 0.5367,
+      "step": 137860
+    },
+    {
+      "epoch": 6.847620939703983,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002521942982020463,
+      "loss": 0.4874,
+      "step": 137870
+    },
+    {
+      "epoch": 6.848117611999602,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002521545644183967,
+      "loss": 0.4807,
+      "step": 137880
+    },
+    {
+      "epoch": 6.848614284295222,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002521148306347472,
+      "loss": 0.4903,
+      "step": 137890
+    },
+    {
+      "epoch": 6.849110956590842,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00025207509685109766,
+      "loss": 0.5168,
+      "step": 137900
+    },
+    {
+      "epoch": 6.849607628886461,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00025203536306744813,
+      "loss": 0.481,
+      "step": 137910
+    },
+    {
+      "epoch": 6.85010430118208,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00025199562928379855,
+      "loss": 0.4804,
+      "step": 137920
+    },
+    {
+      "epoch": 6.850600973477699,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.000251955895500149,
+      "loss": 0.5242,
+      "step": 137930
+    },
+    {
+      "epoch": 6.851097645773319,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002519161617164995,
+      "loss": 0.5137,
+      "step": 137940
+    },
+    {
+      "epoch": 6.851594318068938,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002518764279328499,
+      "loss": 0.516,
+      "step": 137950
+    },
+    {
+      "epoch": 6.852090990364557,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002518366941492004,
+      "loss": 0.5078,
+      "step": 137960
+    },
+    {
+      "epoch": 6.852587662660177,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00025179696036555085,
+      "loss": 0.4554,
+      "step": 137970
+    },
+    {
+      "epoch": 6.853084334955796,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00025175722658190127,
+      "loss": 0.5063,
+      "step": 137980
+    },
+    {
+      "epoch": 6.853581007251416,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00025171749279825174,
+      "loss": 0.4849,
+      "step": 137990
+    },
+    {
+      "epoch": 6.854077679547035,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00025167775901460216,
+      "loss": 0.4813,
+      "step": 138000
+    },
+    {
+      "epoch": 6.854574351842654,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00025163802523095263,
+      "loss": 0.4909,
+      "step": 138010
+    },
+    {
+      "epoch": 6.855071024138273,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002515982914473031,
+      "loss": 0.4865,
+      "step": 138020
+    },
+    {
+      "epoch": 6.8555676964338925,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0002515585576636535,
+      "loss": 0.4834,
+      "step": 138030
+    },
+    {
+      "epoch": 6.856064368729513,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.000251518823880004,
+      "loss": 0.4737,
+      "step": 138040
+    },
+    {
+      "epoch": 6.856561041025132,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00025147909009635446,
+      "loss": 0.4934,
+      "step": 138050
+    },
+    {
+      "epoch": 6.857057713320751,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002514393563127049,
+      "loss": 0.4939,
+      "step": 138060
+    },
+    {
+      "epoch": 6.85755438561637,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00025139962252905535,
+      "loss": 0.5089,
+      "step": 138070
+    },
+    {
+      "epoch": 6.8580510579119895,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002513598887454058,
+      "loss": 0.525,
+      "step": 138080
+    },
+    {
+      "epoch": 6.858547730207609,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00025132015496175624,
+      "loss": 0.5115,
+      "step": 138090
+    },
+    {
+      "epoch": 6.859044402503228,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002512804211781067,
+      "loss": 0.481,
+      "step": 138100
+    },
+    {
+      "epoch": 6.859541074798848,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002512406873944571,
+      "loss": 0.502,
+      "step": 138110
+    },
+    {
+      "epoch": 6.860037747094467,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002512009536108076,
+      "loss": 0.5023,
+      "step": 138120
+    },
+    {
+      "epoch": 6.860534419390087,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00025116121982715807,
+      "loss": 0.5104,
+      "step": 138130
+    },
+    {
+      "epoch": 6.861031091685706,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002511214860435085,
+      "loss": 0.4992,
+      "step": 138140
+    },
+    {
+      "epoch": 6.861527763981325,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00025108175225985896,
+      "loss": 0.4933,
+      "step": 138150
+    },
+    {
+      "epoch": 6.862024436276944,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002510420184762094,
+      "loss": 0.4964,
+      "step": 138160
+    },
+    {
+      "epoch": 6.8625211085725635,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00025100228469255984,
+      "loss": 0.4791,
+      "step": 138170
+    },
+    {
+      "epoch": 6.863017780868184,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002509625509089103,
+      "loss": 0.5395,
+      "step": 138180
+    },
+    {
+      "epoch": 6.863514453163803,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00025092281712526073,
+      "loss": 0.5023,
+      "step": 138190
+    },
+    {
+      "epoch": 6.864011125459422,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00025088308334161126,
+      "loss": 0.5093,
+      "step": 138200
+    },
+    {
+      "epoch": 6.864507797755041,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0002508433495579617,
+      "loss": 0.5053,
+      "step": 138210
+    },
+    {
+      "epoch": 6.8650044700506605,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002508036157743121,
+      "loss": 0.4806,
+      "step": 138220
+    },
+    {
+      "epoch": 6.86550114234628,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00025076388199066256,
+      "loss": 0.4829,
+      "step": 138230
+    },
+    {
+      "epoch": 6.865997814641899,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00025072414820701303,
+      "loss": 0.5352,
+      "step": 138240
+    },
+    {
+      "epoch": 6.866494486937519,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00025068441442336345,
+      "loss": 0.4913,
+      "step": 138250
+    },
+    {
+      "epoch": 6.866991159233138,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0002506446806397139,
+      "loss": 0.4969,
+      "step": 138260
+    },
+    {
+      "epoch": 6.8674878315287575,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002506049468560644,
+      "loss": 0.4607,
+      "step": 138270
+    },
+    {
+      "epoch": 6.867984503824377,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00025056521307241487,
+      "loss": 0.5346,
+      "step": 138280
+    },
+    {
+      "epoch": 6.868481176119996,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002505254792887653,
+      "loss": 0.4662,
+      "step": 138290
+    },
+    {
+      "epoch": 6.868977848415615,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002504857455051157,
+      "loss": 0.5014,
+      "step": 138300
+    },
+    {
+      "epoch": 6.869474520711234,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002504460117214662,
+      "loss": 0.4955,
+      "step": 138310
+    },
+    {
+      "epoch": 6.8699711930068545,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00025040627793781664,
+      "loss": 0.5091,
+      "step": 138320
+    },
+    {
+      "epoch": 6.870467865302474,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00025036654415416706,
+      "loss": 0.4925,
+      "step": 138330
+    },
+    {
+      "epoch": 6.870964537598093,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00025032681037051753,
+      "loss": 0.4968,
+      "step": 138340
+    },
+    {
+      "epoch": 6.871461209893712,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.000250287076586868,
+      "loss": 0.5242,
+      "step": 138350
+    },
+    {
+      "epoch": 6.8719578821893315,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0002502473428032185,
+      "loss": 0.5039,
+      "step": 138360
+    },
+    {
+      "epoch": 6.872454554484951,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002502076090195689,
+      "loss": 0.5184,
+      "step": 138370
+    },
+    {
+      "epoch": 6.87295122678057,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00025016787523591936,
+      "loss": 0.512,
+      "step": 138380
+    },
+    {
+      "epoch": 6.87344789907619,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00025012814145226983,
+      "loss": 0.476,
+      "step": 138390
+    },
+    {
+      "epoch": 6.873944571371809,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00025008840766862025,
+      "loss": 0.4987,
+      "step": 138400
+    },
+    {
+      "epoch": 6.8744412436674285,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002500486738849707,
+      "loss": 0.4966,
+      "step": 138410
+    },
+    {
+      "epoch": 6.874937915963048,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00025000894010132114,
+      "loss": 0.4859,
+      "step": 138420
+    },
+    {
+      "epoch": 6.875434588258667,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002499692063176716,
+      "loss": 0.5372,
+      "step": 138430
+    },
+    {
+      "epoch": 6.875931260554286,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0002499294725340221,
+      "loss": 0.4881,
+      "step": 138440
+    },
+    {
+      "epoch": 6.876427932849905,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002498897387503725,
+      "loss": 0.4923,
+      "step": 138450
+    },
+    {
+      "epoch": 6.876924605145525,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00024985000496672297,
+      "loss": 0.5507,
+      "step": 138460
+    },
+    {
+      "epoch": 6.877421277441144,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00024981027118307344,
+      "loss": 0.4744,
+      "step": 138470
+    },
+    {
+      "epoch": 6.877917949736764,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00024977053739942386,
+      "loss": 0.5066,
+      "step": 138480
+    },
+    {
+      "epoch": 6.878414622032383,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00024973080361577433,
+      "loss": 0.4819,
+      "step": 138490
+    },
+    {
+      "epoch": 6.878911294328002,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002496910698321248,
+      "loss": 0.4991,
+      "step": 138500
+    },
+    {
+      "epoch": 6.879407966623622,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002496513360484752,
+      "loss": 0.4835,
+      "step": 138510
+    },
+    {
+      "epoch": 6.879904638919241,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002496116022648257,
+      "loss": 0.4923,
+      "step": 138520
+    },
+    {
+      "epoch": 6.88040131121486,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002495718684811761,
+      "loss": 0.502,
+      "step": 138530
+    },
+    {
+      "epoch": 6.880897983510479,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002495321346975266,
+      "loss": 0.5097,
+      "step": 138540
+    },
+    {
+      "epoch": 6.881394655806099,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00024949240091387705,
+      "loss": 0.5251,
+      "step": 138550
+    },
+    {
+      "epoch": 6.881891328101719,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00024945266713022747,
+      "loss": 0.497,
+      "step": 138560
+    },
+    {
+      "epoch": 6.882388000397338,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00024941293334657794,
+      "loss": 0.5,
+      "step": 138570
+    },
+    {
+      "epoch": 6.882884672692957,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002493731995629284,
+      "loss": 0.5033,
+      "step": 138580
+    },
+    {
+      "epoch": 6.883381344988576,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00024933346577927883,
+      "loss": 0.4907,
+      "step": 138590
+    },
+    {
+      "epoch": 6.883878017284196,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002492937319956293,
+      "loss": 0.5065,
+      "step": 138600
+    },
+    {
+      "epoch": 6.884374689579815,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00024925399821197977,
+      "loss": 0.4797,
+      "step": 138610
+    },
+    {
+      "epoch": 6.884871361875435,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002492142644283302,
+      "loss": 0.4651,
+      "step": 138620
+    },
+    {
+      "epoch": 6.885368034171054,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00024917453064468066,
+      "loss": 0.5183,
+      "step": 138630
+    },
+    {
+      "epoch": 6.885864706466673,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002491347968610311,
+      "loss": 0.511,
+      "step": 138640
+    },
+    {
+      "epoch": 6.886361378762293,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00024909506307738155,
+      "loss": 0.4801,
+      "step": 138650
+    },
+    {
+      "epoch": 6.886858051057912,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.000249055329293732,
+      "loss": 0.4738,
+      "step": 138660
+    },
+    {
+      "epoch": 6.887354723353531,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00024901559551008244,
+      "loss": 0.4818,
+      "step": 138670
+    },
+    {
+      "epoch": 6.88785139564915,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0002489758617264329,
+      "loss": 0.4866,
+      "step": 138680
+    },
+    {
+      "epoch": 6.88834806794477,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0002489361279427834,
+      "loss": 0.4494,
+      "step": 138690
+    },
+    {
+      "epoch": 6.88884474024039,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002488963941591338,
+      "loss": 0.5095,
+      "step": 138700
+    },
+    {
+      "epoch": 6.889341412536009,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00024885666037548427,
+      "loss": 0.506,
+      "step": 138710
+    },
+    {
+      "epoch": 6.889838084831628,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0002488169265918347,
+      "loss": 0.4734,
+      "step": 138720
+    },
+    {
+      "epoch": 6.890334757127247,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002487771928081852,
+      "loss": 0.4911,
+      "step": 138730
+    },
+    {
+      "epoch": 6.8908314294228665,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002487374590245356,
+      "loss": 0.5237,
+      "step": 138740
+    },
+    {
+      "epoch": 6.891328101718486,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00024869772524088604,
+      "loss": 0.5002,
+      "step": 138750
+    },
+    {
+      "epoch": 6.891824774014106,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002486579914572365,
+      "loss": 0.4975,
+      "step": 138760
+    },
+    {
+      "epoch": 6.892321446309725,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.000248618257673587,
+      "loss": 0.5047,
+      "step": 138770
+    },
+    {
+      "epoch": 6.892818118605344,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002485785238899374,
+      "loss": 0.5074,
+      "step": 138780
+    },
+    {
+      "epoch": 6.8933147909009636,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002485387901062879,
+      "loss": 0.478,
+      "step": 138790
+    },
+    {
+      "epoch": 6.893811463196583,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00024849905632263835,
+      "loss": 0.4823,
+      "step": 138800
+    },
+    {
+      "epoch": 6.894308135492202,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002484593225389888,
+      "loss": 0.4991,
+      "step": 138810
+    },
+    {
+      "epoch": 6.894804807787821,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00024841958875533923,
+      "loss": 0.5135,
+      "step": 138820
+    },
+    {
+      "epoch": 6.895301480083441,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00024837985497168965,
+      "loss": 0.5066,
+      "step": 138830
+    },
+    {
+      "epoch": 6.895798152379061,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002483401211880402,
+      "loss": 0.4946,
+      "step": 138840
+    },
+    {
+      "epoch": 6.89629482467468,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002483003874043906,
+      "loss": 0.4861,
+      "step": 138850
+    },
+    {
+      "epoch": 6.896791496970299,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00024826065362074107,
+      "loss": 0.5071,
+      "step": 138860
+    },
+    {
+      "epoch": 6.897288169265918,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002482209198370915,
+      "loss": 0.4995,
+      "step": 138870
+    },
+    {
+      "epoch": 6.8977848415615375,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00024818118605344195,
+      "loss": 0.4846,
+      "step": 138880
+    },
+    {
+      "epoch": 6.898281513857157,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002481414522697924,
+      "loss": 0.4874,
+      "step": 138890
+    },
+    {
+      "epoch": 6.898778186152777,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00024810171848614284,
+      "loss": 0.5222,
+      "step": 138900
+    },
+    {
+      "epoch": 6.899274858448396,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002480619847024933,
+      "loss": 0.4916,
+      "step": 138910
+    },
+    {
+      "epoch": 6.899771530744015,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0002480222509188438,
+      "loss": 0.4765,
+      "step": 138920
+    },
+    {
+      "epoch": 6.9002682030396345,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002479825171351942,
+      "loss": 0.5114,
+      "step": 138930
+    },
+    {
+      "epoch": 6.900764875335254,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002479427833515447,
+      "loss": 0.4979,
+      "step": 138940
+    },
+    {
+      "epoch": 6.901261547630873,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002479030495678951,
+      "loss": 0.4637,
+      "step": 138950
+    },
+    {
+      "epoch": 6.901758219926492,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00024786331578424556,
+      "loss": 0.5241,
+      "step": 138960
+    },
+    {
+      "epoch": 6.902254892222112,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00024782358200059603,
+      "loss": 0.4974,
+      "step": 138970
+    },
+    {
+      "epoch": 6.9027515645177315,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00024778384821694645,
+      "loss": 0.5086,
+      "step": 138980
+    },
+    {
+      "epoch": 6.903248236813351,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0002477441144332969,
+      "loss": 0.4754,
+      "step": 138990
+    },
+    {
+      "epoch": 6.90374490910897,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002477043806496474,
+      "loss": 0.5033,
+      "step": 139000
+    },
+    {
+      "epoch": 6.904241581404589,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002476646468659978,
+      "loss": 0.5201,
+      "step": 139010
+    },
+    {
+      "epoch": 6.904738253700208,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002476249130823483,
+      "loss": 0.5056,
+      "step": 139020
+    },
+    {
+      "epoch": 6.905234925995828,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00024758517929869875,
+      "loss": 0.4932,
+      "step": 139030
+    },
+    {
+      "epoch": 6.905731598291448,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00024754544551504917,
+      "loss": 0.4779,
+      "step": 139040
+    },
+    {
+      "epoch": 6.906228270587067,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00024750571173139964,
+      "loss": 0.5336,
+      "step": 139050
+    },
+    {
+      "epoch": 6.906724942882686,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00024746597794775006,
+      "loss": 0.5058,
+      "step": 139060
+    },
+    {
+      "epoch": 6.9072216151783055,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00024742624416410053,
+      "loss": 0.5215,
+      "step": 139070
+    },
+    {
+      "epoch": 6.907718287473925,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.000247386510380451,
+      "loss": 0.4882,
+      "step": 139080
+    },
+    {
+      "epoch": 6.908214959769544,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002473467765968014,
+      "loss": 0.4795,
+      "step": 139090
+    },
+    {
+      "epoch": 6.908711632065163,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002473070428131519,
+      "loss": 0.486,
+      "step": 139100
+    },
+    {
+      "epoch": 6.909208304360782,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00024726730902950236,
+      "loss": 0.4911,
+      "step": 139110
+    },
+    {
+      "epoch": 6.9097049766564025,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002472275752458528,
+      "loss": 0.4978,
+      "step": 139120
+    },
+    {
+      "epoch": 6.910201648952022,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00024718784146220325,
+      "loss": 0.4819,
+      "step": 139130
+    },
+    {
+      "epoch": 6.910698321247641,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0002471481076785537,
+      "loss": 0.4984,
+      "step": 139140
+    },
+    {
+      "epoch": 6.91119499354326,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00024710837389490414,
+      "loss": 0.5102,
+      "step": 139150
+    },
+    {
+      "epoch": 6.911691665838879,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002470686401112546,
+      "loss": 0.4817,
+      "step": 139160
+    },
+    {
+      "epoch": 6.912188338134499,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00024702890632760503,
+      "loss": 0.5105,
+      "step": 139170
+    },
+    {
+      "epoch": 6.912685010430118,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00024698917254395555,
+      "loss": 0.4879,
+      "step": 139180
+    },
+    {
+      "epoch": 6.913181682725737,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00024694943876030597,
+      "loss": 0.507,
+      "step": 139190
+    },
+    {
+      "epoch": 6.913678355021357,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002469097049766564,
+      "loss": 0.5073,
+      "step": 139200
+    },
+    {
+      "epoch": 6.914175027316976,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00024686997119300686,
+      "loss": 0.4896,
+      "step": 139210
+    },
+    {
+      "epoch": 6.914671699612596,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00024683023740935733,
+      "loss": 0.4983,
+      "step": 139220
+    },
+    {
+      "epoch": 6.915168371908215,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002467905036257078,
+      "loss": 0.5106,
+      "step": 139230
+    },
+    {
+      "epoch": 6.915665044203834,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002467507698420582,
+      "loss": 0.4939,
+      "step": 139240
+    },
+    {
+      "epoch": 6.916161716499453,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00024671103605840864,
+      "loss": 0.541,
+      "step": 139250
+    },
+    {
+      "epoch": 6.9166583887950726,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00024667130227475916,
+      "loss": 0.5167,
+      "step": 139260
+    },
+    {
+      "epoch": 6.917155061090693,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002466315684911096,
+      "loss": 0.5126,
+      "step": 139270
+    },
+    {
+      "epoch": 6.917651733386312,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00024659183470746,
+      "loss": 0.5017,
+      "step": 139280
+    },
+    {
+      "epoch": 6.918148405681931,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00024655210092381047,
+      "loss": 0.5083,
+      "step": 139290
+    },
+    {
+      "epoch": 6.91864507797755,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00024651236714016094,
+      "loss": 0.5055,
+      "step": 139300
+    },
+    {
+      "epoch": 6.91914175027317,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002464726333565114,
+      "loss": 0.5173,
+      "step": 139310
+    },
+    {
+      "epoch": 6.919638422568789,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002464328995728618,
+      "loss": 0.4768,
+      "step": 139320
+    },
+    {
+      "epoch": 6.920135094864408,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002463931657892123,
+      "loss": 0.4742,
+      "step": 139330
+    },
+    {
+      "epoch": 6.920631767160028,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00024635343200556277,
+      "loss": 0.4894,
+      "step": 139340
+    },
+    {
+      "epoch": 6.921128439455647,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002463136982219132,
+      "loss": 0.4996,
+      "step": 139350
+    },
+    {
+      "epoch": 6.921625111751267,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002462739644382636,
+      "loss": 0.5211,
+      "step": 139360
+    },
+    {
+      "epoch": 6.922121784046886,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00024623423065461413,
+      "loss": 0.4981,
+      "step": 139370
+    },
+    {
+      "epoch": 6.922618456342505,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00024619449687096455,
+      "loss": 0.5044,
+      "step": 139380
+    },
+    {
+      "epoch": 6.923115128638124,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.000246154763087315,
+      "loss": 0.502,
+      "step": 139390
+    },
+    {
+      "epoch": 6.9236118009337435,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00024611502930366544,
+      "loss": 0.4946,
+      "step": 139400
+    },
+    {
+      "epoch": 6.924108473229364,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002460752955200159,
+      "loss": 0.5054,
+      "step": 139410
+    },
+    {
+      "epoch": 6.924605145524983,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0002460355617363664,
+      "loss": 0.4877,
+      "step": 139420
+    },
+    {
+      "epoch": 6.925101817820602,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002459958279527168,
+      "loss": 0.4919,
+      "step": 139430
+    },
+    {
+      "epoch": 6.925598490116221,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00024595609416906727,
+      "loss": 0.4642,
+      "step": 139440
+    },
+    {
+      "epoch": 6.9260951624118405,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00024591636038541774,
+      "loss": 0.5025,
+      "step": 139450
+    },
+    {
+      "epoch": 6.92659183470746,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00024587662660176815,
+      "loss": 0.4847,
+      "step": 139460
+    },
+    {
+      "epoch": 6.927088507003079,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002458368928181186,
+      "loss": 0.5113,
+      "step": 139470
+    },
+    {
+      "epoch": 6.927585179298699,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002457971590344691,
+      "loss": 0.4855,
+      "step": 139480
+    },
+    {
+      "epoch": 6.928081851594318,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002457574252508195,
+      "loss": 0.4971,
+      "step": 139490
+    },
+    {
+      "epoch": 6.928578523889938,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00024571769146717,
+      "loss": 0.5003,
+      "step": 139500
+    },
+    {
+      "epoch": 6.929075196185557,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002456779576835204,
+      "loss": 0.4825,
+      "step": 139510
+    },
+    {
+      "epoch": 6.929571868481176,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002456382238998709,
+      "loss": 0.501,
+      "step": 139520
+    },
+    {
+      "epoch": 6.930068540776795,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00024559849011622135,
+      "loss": 0.4937,
+      "step": 139530
+    },
+    {
+      "epoch": 6.9305652130724145,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00024555875633257176,
+      "loss": 0.4989,
+      "step": 139540
+    },
+    {
+      "epoch": 6.931061885368035,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00024551902254892223,
+      "loss": 0.4682,
+      "step": 139550
+    },
+    {
+      "epoch": 6.931558557663654,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002454792887652727,
+      "loss": 0.5281,
+      "step": 139560
+    },
+    {
+      "epoch": 6.932055229959273,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002454395549816231,
+      "loss": 0.5126,
+      "step": 139570
+    },
+    {
+      "epoch": 6.932551902254892,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002453998211979736,
+      "loss": 0.4697,
+      "step": 139580
+    },
+    {
+      "epoch": 6.9330485745505115,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.000245360087414324,
+      "loss": 0.4833,
+      "step": 139590
+    },
+    {
+      "epoch": 6.933545246846131,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002453203536306745,
+      "loss": 0.4721,
+      "step": 139600
+    },
+    {
+      "epoch": 6.93404191914175,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00024528061984702495,
+      "loss": 0.492,
+      "step": 139610
+    },
+    {
+      "epoch": 6.93453859143737,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00024524088606337537,
+      "loss": 0.5299,
+      "step": 139620
+    },
+    {
+      "epoch": 6.935035263732989,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00024520115227972584,
+      "loss": 0.4887,
+      "step": 139630
+    },
+    {
+      "epoch": 6.9355319360286085,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002451614184960763,
+      "loss": 0.4836,
+      "step": 139640
+    },
+    {
+      "epoch": 6.936028608324228,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00024512168471242673,
+      "loss": 0.4728,
+      "step": 139650
+    },
+    {
+      "epoch": 6.936525280619847,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002450819509287772,
+      "loss": 0.4863,
+      "step": 139660
+    },
+    {
+      "epoch": 6.937021952915466,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002450422171451277,
+      "loss": 0.4935,
+      "step": 139670
+    },
+    {
+      "epoch": 6.937518625211085,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.00024500248336147815,
+      "loss": 0.486,
+      "step": 139680
+    },
+    {
+      "epoch": 6.9380152975067055,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00024496274957782856,
+      "loss": 0.4928,
+      "step": 139690
+    },
+    {
+      "epoch": 6.938511969802325,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.000244923015794179,
+      "loss": 0.4964,
+      "step": 139700
+    },
+    {
+      "epoch": 6.939008642097944,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002448832820105295,
+      "loss": 0.4789,
+      "step": 139710
+    },
+    {
+      "epoch": 6.939505314393563,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002448435482268799,
+      "loss": 0.4749,
+      "step": 139720
+    },
+    {
+      "epoch": 6.9400019866891824,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00024480381444323034,
+      "loss": 0.4678,
+      "step": 139730
+    },
+    {
+      "epoch": 6.940498658984802,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002447640806595808,
+      "loss": 0.4679,
+      "step": 139740
+    },
+    {
+      "epoch": 6.940995331280421,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002447243468759313,
+      "loss": 0.4952,
+      "step": 139750
+    },
+    {
+      "epoch": 6.941492003576041,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00024468461309228175,
+      "loss": 0.4943,
+      "step": 139760
+    },
+    {
+      "epoch": 6.94198867587166,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00024464487930863217,
+      "loss": 0.4898,
+      "step": 139770
+    },
+    {
+      "epoch": 6.9424853481672795,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00024460514552498264,
+      "loss": 0.5022,
+      "step": 139780
+    },
+    {
+      "epoch": 6.942982020462899,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0002445654117413331,
+      "loss": 0.5112,
+      "step": 139790
+    },
+    {
+      "epoch": 6.943478692758518,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00024452567795768353,
+      "loss": 0.4766,
+      "step": 139800
+    },
+    {
+      "epoch": 6.943975365054137,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00024448594417403395,
+      "loss": 0.4633,
+      "step": 139810
+    },
+    {
+      "epoch": 6.944472037349756,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002444462103903844,
+      "loss": 0.4725,
+      "step": 139820
+    },
+    {
+      "epoch": 6.944968709645376,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002444064766067349,
+      "loss": 0.4767,
+      "step": 139830
+    },
+    {
+      "epoch": 6.945465381940996,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00024436674282308536,
+      "loss": 0.5269,
+      "step": 139840
+    },
+    {
+      "epoch": 6.945962054236615,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002443270090394358,
+      "loss": 0.4814,
+      "step": 139850
+    },
+    {
+      "epoch": 6.946458726532234,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00024428727525578625,
+      "loss": 0.4934,
+      "step": 139860
+    },
+    {
+      "epoch": 6.946955398827853,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002442475414721367,
+      "loss": 0.4957,
+      "step": 139870
+    },
+    {
+      "epoch": 6.947452071123473,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00024420780768848714,
+      "loss": 0.5135,
+      "step": 139880
+    },
+    {
+      "epoch": 6.947948743419092,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00024416807390483756,
+      "loss": 0.4758,
+      "step": 139890
+    },
+    {
+      "epoch": 6.948445415714711,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00024412834012118805,
+      "loss": 0.5051,
+      "step": 139900
+    },
+    {
+      "epoch": 6.94894208801033,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002440886063375385,
+      "loss": 0.4878,
+      "step": 139910
+    },
+    {
+      "epoch": 6.94943876030595,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00024404887255388897,
+      "loss": 0.4885,
+      "step": 139920
+    },
+    {
+      "epoch": 6.94993543260157,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00024400913877023941,
+      "loss": 0.4816,
+      "step": 139930
+    },
+    {
+      "epoch": 6.950432104897189,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00024396940498658983,
+      "loss": 0.505,
+      "step": 139940
+    },
+    {
+      "epoch": 6.950928777192808,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00024392967120294033,
+      "loss": 0.5088,
+      "step": 139950
+    },
+    {
+      "epoch": 6.951425449488427,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00024388993741929075,
+      "loss": 0.4806,
+      "step": 139960
+    },
+    {
+      "epoch": 6.951922121784047,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002438502036356412,
+      "loss": 0.4938,
+      "step": 139970
+    },
+    {
+      "epoch": 6.952418794079666,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00024381046985199166,
+      "loss": 0.4716,
+      "step": 139980
+    },
+    {
+      "epoch": 6.952915466375286,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002437707360683421,
+      "loss": 0.4989,
+      "step": 139990
+    },
+    {
+      "epoch": 6.953412138670905,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00024373100228469258,
+      "loss": 0.4904,
+      "step": 140000
+    },
+    {
+      "epoch": 6.953908810966524,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00024369126850104302,
+      "loss": 0.5107,
+      "step": 140010
+    },
+    {
+      "epoch": 6.954405483262144,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00024365153471739347,
+      "loss": 0.5227,
+      "step": 140020
+    },
+    {
+      "epoch": 6.954902155557763,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00024361180093374394,
+      "loss": 0.4979,
+      "step": 140030
+    },
+    {
+      "epoch": 6.955398827853382,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00024357206715009438,
+      "loss": 0.4795,
+      "step": 140040
+    },
+    {
+      "epoch": 6.955895500149001,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002435323333664448,
+      "loss": 0.485,
+      "step": 140050
+    },
+    {
+      "epoch": 6.956392172444621,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0002434925995827953,
+      "loss": 0.484,
+      "step": 140060
+    },
+    {
+      "epoch": 6.956888844740241,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00024345286579914571,
+      "loss": 0.5158,
+      "step": 140070
+    },
+    {
+      "epoch": 6.95738551703586,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002434131320154962,
+      "loss": 0.4753,
+      "step": 140080
+    },
+    {
+      "epoch": 6.957882189331479,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00024337339823184663,
+      "loss": 0.4996,
+      "step": 140090
+    },
+    {
+      "epoch": 6.958378861627098,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00024333366444819707,
+      "loss": 0.4916,
+      "step": 140100
+    },
+    {
+      "epoch": 6.9588755339227175,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00024329393066454755,
+      "loss": 0.4938,
+      "step": 140110
+    },
+    {
+      "epoch": 6.959372206218337,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.000243254196880898,
+      "loss": 0.5164,
+      "step": 140120
+    },
+    {
+      "epoch": 6.959868878513957,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00024321446309724846,
+      "loss": 0.4918,
+      "step": 140130
+    },
+    {
+      "epoch": 6.960365550809576,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002431747293135989,
+      "loss": 0.5013,
+      "step": 140140
+    },
+    {
+      "epoch": 6.960862223105195,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00024313499552994935,
+      "loss": 0.4931,
+      "step": 140150
+    },
+    {
+      "epoch": 6.9613588954008145,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00024309526174629982,
+      "loss": 0.4834,
+      "step": 140160
+    },
+    {
+      "epoch": 6.961855567696434,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00024305552796265027,
+      "loss": 0.4915,
+      "step": 140170
+    },
+    {
+      "epoch": 6.962352239992053,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00024301579417900068,
+      "loss": 0.4947,
+      "step": 140180
+    },
+    {
+      "epoch": 6.962848912287672,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00024297606039535118,
+      "loss": 0.5211,
+      "step": 140190
+    },
+    {
+      "epoch": 6.963345584583292,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002429363266117016,
+      "loss": 0.4868,
+      "step": 140200
+    },
+    {
+      "epoch": 6.963842256878912,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00024289659282805207,
+      "loss": 0.4687,
+      "step": 140210
+    },
+    {
+      "epoch": 6.964338929174531,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00024285685904440251,
+      "loss": 0.4978,
+      "step": 140220
+    },
+    {
+      "epoch": 6.96483560147015,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00024281712526075296,
+      "loss": 0.494,
+      "step": 140230
+    },
+    {
+      "epoch": 6.965332273765769,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00024277739147710343,
+      "loss": 0.4935,
+      "step": 140240
+    },
+    {
+      "epoch": 6.9658289460613885,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00024273765769345387,
+      "loss": 0.4853,
+      "step": 140250
+    },
+    {
+      "epoch": 6.966325618357008,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0002426979239098043,
+      "loss": 0.4798,
+      "step": 140260
+    },
+    {
+      "epoch": 6.966822290652628,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002426581901261548,
+      "loss": 0.4909,
+      "step": 140270
+    },
+    {
+      "epoch": 6.967318962948247,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002426184563425052,
+      "loss": 0.5131,
+      "step": 140280
+    },
+    {
+      "epoch": 6.967815635243866,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0002425787225588557,
+      "loss": 0.4682,
+      "step": 140290
+    },
+    {
+      "epoch": 6.9683123075394855,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00024253898877520612,
+      "loss": 0.4771,
+      "step": 140300
+    },
+    {
+      "epoch": 6.968808979835105,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00024249925499155657,
+      "loss": 0.4993,
+      "step": 140310
+    },
+    {
+      "epoch": 6.969305652130724,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00024245952120790704,
+      "loss": 0.4717,
+      "step": 140320
+    },
+    {
+      "epoch": 6.969802324426343,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00024241978742425748,
+      "loss": 0.4884,
+      "step": 140330
+    },
+    {
+      "epoch": 6.970298996721963,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00024238005364060793,
+      "loss": 0.4904,
+      "step": 140340
+    },
+    {
+      "epoch": 6.9707956690175825,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002423403198569584,
+      "loss": 0.4857,
+      "step": 140350
+    },
+    {
+      "epoch": 6.971292341313202,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00024230058607330884,
+      "loss": 0.4607,
+      "step": 140360
+    },
+    {
+      "epoch": 6.971789013608821,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002422608522896593,
+      "loss": 0.4999,
+      "step": 140370
+    },
+    {
+      "epoch": 6.97228568590444,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00024222111850600976,
+      "loss": 0.4969,
+      "step": 140380
+    },
+    {
+      "epoch": 6.972782358200059,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00024218138472236017,
+      "loss": 0.5017,
+      "step": 140390
+    },
+    {
+      "epoch": 6.973279030495679,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00024214165093871067,
+      "loss": 0.54,
+      "step": 140400
+    },
+    {
+      "epoch": 6.973775702791299,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002421019171550611,
+      "loss": 0.4736,
+      "step": 140410
+    },
+    {
+      "epoch": 6.974272375086918,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00024206218337141153,
+      "loss": 0.4898,
+      "step": 140420
+    },
+    {
+      "epoch": 6.974769047382537,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.000242022449587762,
+      "loss": 0.5352,
+      "step": 140430
+    },
+    {
+      "epoch": 6.9752657196781565,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00024198271580411245,
+      "loss": 0.4869,
+      "step": 140440
+    },
+    {
+      "epoch": 6.975762391973776,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00024194298202046292,
+      "loss": 0.4762,
+      "step": 140450
+    },
+    {
+      "epoch": 6.976259064269395,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00024190324823681337,
+      "loss": 0.4969,
+      "step": 140460
+    },
+    {
+      "epoch": 6.976755736565014,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002418635144531638,
+      "loss": 0.5136,
+      "step": 140470
+    },
+    {
+      "epoch": 6.977252408860634,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00024182378066951428,
+      "loss": 0.514,
+      "step": 140480
+    },
+    {
+      "epoch": 6.9777490811562535,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00024178404688586473,
+      "loss": 0.4833,
+      "step": 140490
+    },
+    {
+      "epoch": 6.978245753451873,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002417443131022152,
+      "loss": 0.4871,
+      "step": 140500
+    },
+    {
+      "epoch": 6.978742425747492,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00024170457931856561,
+      "loss": 0.5071,
+      "step": 140510
+    },
+    {
+      "epoch": 6.979239098043111,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00024166484553491606,
+      "loss": 0.5003,
+      "step": 140520
+    },
+    {
+      "epoch": 6.97973577033873,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00024162511175126653,
+      "loss": 0.4776,
+      "step": 140530
+    },
+    {
+      "epoch": 6.98023244263435,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00024158537796761697,
+      "loss": 0.5007,
+      "step": 140540
+    },
+    {
+      "epoch": 6.980729114929969,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00024154564418396742,
+      "loss": 0.4972,
+      "step": 140550
+    },
+    {
+      "epoch": 6.981225787225588,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002415059104003179,
+      "loss": 0.4996,
+      "step": 140560
+    },
+    {
+      "epoch": 6.981722459521208,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00024146617661666833,
+      "loss": 0.4959,
+      "step": 140570
+    },
+    {
+      "epoch": 6.982219131816827,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002414264428330188,
+      "loss": 0.5053,
+      "step": 140580
+    },
+    {
+      "epoch": 6.982715804112447,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00024138670904936925,
+      "loss": 0.5081,
+      "step": 140590
+    },
+    {
+      "epoch": 6.983212476408066,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00024134697526571967,
+      "loss": 0.4818,
+      "step": 140600
+    },
+    {
+      "epoch": 6.983709148703685,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00024130724148207017,
+      "loss": 0.5042,
+      "step": 140610
+    },
+    {
+      "epoch": 6.984205820999304,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00024126750769842058,
+      "loss": 0.5098,
+      "step": 140620
+    },
+    {
+      "epoch": 6.9847024932949235,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00024122777391477103,
+      "loss": 0.5089,
+      "step": 140630
+    },
+    {
+      "epoch": 6.985199165590544,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002411880401311215,
+      "loss": 0.4998,
+      "step": 140640
+    },
+    {
+      "epoch": 6.985695837886163,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00024114830634747194,
+      "loss": 0.5101,
+      "step": 140650
+    },
+    {
+      "epoch": 6.986192510181782,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00024110857256382241,
+      "loss": 0.4872,
+      "step": 140660
+    },
+    {
+      "epoch": 6.986689182477401,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00024106883878017286,
+      "loss": 0.4679,
+      "step": 140670
+    },
+    {
+      "epoch": 6.987185854773021,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002410291049965233,
+      "loss": 0.4831,
+      "step": 140680
+    },
+    {
+      "epoch": 6.98768252706864,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00024098937121287377,
+      "loss": 0.474,
+      "step": 140690
+    },
+    {
+      "epoch": 6.988179199364259,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00024094963742922422,
+      "loss": 0.492,
+      "step": 140700
+    },
+    {
+      "epoch": 6.988675871659879,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00024090990364557463,
+      "loss": 0.5069,
+      "step": 140710
+    },
+    {
+      "epoch": 6.989172543955498,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00024087016986192513,
+      "loss": 0.5026,
+      "step": 140720
+    },
+    {
+      "epoch": 6.989669216251118,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00024083043607827555,
+      "loss": 0.484,
+      "step": 140730
+    },
+    {
+      "epoch": 6.990165888546737,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00024079070229462605,
+      "loss": 0.498,
+      "step": 140740
+    },
+    {
+      "epoch": 6.990662560842356,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00024075096851097647,
+      "loss": 0.4986,
+      "step": 140750
+    },
+    {
+      "epoch": 6.991159233137975,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002407112347273269,
+      "loss": 0.5142,
+      "step": 140760
+    },
+    {
+      "epoch": 6.9916559054335945,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00024067150094367738,
+      "loss": 0.4693,
+      "step": 140770
+    },
+    {
+      "epoch": 6.992152577729215,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00024063176716002783,
+      "loss": 0.522,
+      "step": 140780
+    },
+    {
+      "epoch": 6.992649250024834,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00024059203337637824,
+      "loss": 0.4882,
+      "step": 140790
+    },
+    {
+      "epoch": 6.993145922320453,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00024055229959272874,
+      "loss": 0.502,
+      "step": 140800
+    },
+    {
+      "epoch": 6.993642594616072,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00024051256580907916,
+      "loss": 0.4914,
+      "step": 140810
+    },
+    {
+      "epoch": 6.9941392669116915,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00024047283202542966,
+      "loss": 0.4719,
+      "step": 140820
+    },
+    {
+      "epoch": 6.994635939207311,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00024043309824178007,
+      "loss": 0.472,
+      "step": 140830
+    },
+    {
+      "epoch": 6.99513261150293,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00024039336445813052,
+      "loss": 0.4798,
+      "step": 140840
+    },
+    {
+      "epoch": 6.99562928379855,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.000240353630674481,
+      "loss": 0.5351,
+      "step": 140850
+    },
+    {
+      "epoch": 6.996125956094169,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00024031389689083143,
+      "loss": 0.5033,
+      "step": 140860
+    },
+    {
+      "epoch": 6.9966226283897885,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00024027416310718188,
+      "loss": 0.4998,
+      "step": 140870
+    },
+    {
+      "epoch": 6.997119300685408,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00024023442932353235,
+      "loss": 0.4827,
+      "step": 140880
+    },
+    {
+      "epoch": 6.997615972981027,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002401946955398828,
+      "loss": 0.5031,
+      "step": 140890
+    },
+    {
+      "epoch": 6.998112645276646,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00024015496175623327,
+      "loss": 0.5173,
+      "step": 140900
+    },
+    {
+      "epoch": 6.9986093175722655,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002401152279725837,
+      "loss": 0.482,
+      "step": 140910
+    },
+    {
+      "epoch": 6.999105989867886,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00024007549418893413,
+      "loss": 0.4702,
+      "step": 140920
+    },
+    {
+      "epoch": 6.999602662163505,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00024003576040528463,
+      "loss": 0.4797,
+      "step": 140930
+    },
+    {
+      "epoch": 7.000099334459124,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00023999602662163504,
+      "loss": 0.5283,
+      "step": 140940
+    },
+    {
+      "epoch": 7.000596006754743,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00023995629283798554,
+      "loss": 0.4986,
+      "step": 140950
+    },
+    {
+      "epoch": 7.0010926790503625,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00023991655905433596,
+      "loss": 0.4611,
+      "step": 140960
+    },
+    {
+      "epoch": 7.001589351345982,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002398768252706864,
+      "loss": 0.4936,
+      "step": 140970
+    },
+    {
+      "epoch": 7.002086023641601,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00023983709148703687,
+      "loss": 0.4864,
+      "step": 140980
+    },
+    {
+      "epoch": 7.002582695937221,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00023979735770338732,
+      "loss": 0.4895,
+      "step": 140990
+    },
+    {
+      "epoch": 7.00307936823284,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00023975762391973776,
+      "loss": 0.4758,
+      "step": 141000
+    },
+    {
+      "epoch": 7.0035760405284595,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00023971789013608823,
+      "loss": 0.4665,
+      "step": 141010
+    },
+    {
+      "epoch": 7.004072712824079,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00023967815635243868,
+      "loss": 0.4966,
+      "step": 141020
+    },
+    {
+      "epoch": 7.004569385119698,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00023963842256878915,
+      "loss": 0.4663,
+      "step": 141030
+    },
+    {
+      "epoch": 7.005066057415317,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0002395986887851396,
+      "loss": 0.5,
+      "step": 141040
+    },
+    {
+      "epoch": 7.005562729710936,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00023955895500149,
+      "loss": 0.5082,
+      "step": 141050
+    },
+    {
+      "epoch": 7.006059402006556,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00023951922121784048,
+      "loss": 0.4666,
+      "step": 141060
+    },
+    {
+      "epoch": 7.006556074302176,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00023947948743419093,
+      "loss": 0.4773,
+      "step": 141070
+    },
+    {
+      "epoch": 7.007052746597795,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00023943975365054137,
+      "loss": 0.4744,
+      "step": 141080
+    },
+    {
+      "epoch": 7.007549418893414,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00023940001986689184,
+      "loss": 0.4914,
+      "step": 141090
+    },
+    {
+      "epoch": 7.008046091189033,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00023936028608324229,
+      "loss": 0.4801,
+      "step": 141100
+    },
+    {
+      "epoch": 7.008542763484653,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00023932055229959276,
+      "loss": 0.4721,
+      "step": 141110
+    },
+    {
+      "epoch": 7.009039435780272,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002392808185159432,
+      "loss": 0.4794,
+      "step": 141120
+    },
+    {
+      "epoch": 7.009536108075891,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00023924108473229362,
+      "loss": 0.5034,
+      "step": 141130
+    },
+    {
+      "epoch": 7.010032780371511,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00023920135094864412,
+      "loss": 0.4798,
+      "step": 141140
+    },
+    {
+      "epoch": 7.0105294526671305,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00023916161716499453,
+      "loss": 0.4586,
+      "step": 141150
+    },
+    {
+      "epoch": 7.01102612496275,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00023912188338134498,
+      "loss": 0.4931,
+      "step": 141160
+    },
+    {
+      "epoch": 7.011522797258369,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00023908214959769545,
+      "loss": 0.4674,
+      "step": 141170
+    },
+    {
+      "epoch": 7.012019469553988,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002390424158140459,
+      "loss": 0.496,
+      "step": 141180
+    },
+    {
+      "epoch": 7.012516141849607,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00023900268203039637,
+      "loss": 0.4728,
+      "step": 141190
+    },
+    {
+      "epoch": 7.013012814145227,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002389629482467468,
+      "loss": 0.4576,
+      "step": 141200
+    },
+    {
+      "epoch": 7.013509486440847,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00023892321446309725,
+      "loss": 0.4785,
+      "step": 141210
+    },
+    {
+      "epoch": 7.014006158736466,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00023888348067944773,
+      "loss": 0.4936,
+      "step": 141220
+    },
+    {
+      "epoch": 7.014502831032085,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00023884374689579817,
+      "loss": 0.4896,
+      "step": 141230
+    },
+    {
+      "epoch": 7.014999503327704,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002388040131121486,
+      "loss": 0.4937,
+      "step": 141240
+    },
+    {
+      "epoch": 7.015496175623324,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00023876427932849909,
+      "loss": 0.4706,
+      "step": 141250
+    },
+    {
+      "epoch": 7.015992847918943,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002387245455448495,
+      "loss": 0.5109,
+      "step": 141260
+    },
+    {
+      "epoch": 7.016489520214562,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002386848117612,
+      "loss": 0.4743,
+      "step": 141270
+    },
+    {
+      "epoch": 7.016986192510182,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00023864507797755042,
+      "loss": 0.4832,
+      "step": 141280
+    },
+    {
+      "epoch": 7.017482864805801,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00023860534419390086,
+      "loss": 0.5202,
+      "step": 141290
+    },
+    {
+      "epoch": 7.017979537101421,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00023856561041025133,
+      "loss": 0.519,
+      "step": 141300
+    },
+    {
+      "epoch": 7.01847620939704,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00023852587662660178,
+      "loss": 0.4728,
+      "step": 141310
+    },
+    {
+      "epoch": 7.018972881692659,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00023848614284295225,
+      "loss": 0.4885,
+      "step": 141320
+    },
+    {
+      "epoch": 7.019469553988278,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002384464090593027,
+      "loss": 0.4653,
+      "step": 141330
+    },
+    {
+      "epoch": 7.0199662262838975,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002384066752756531,
+      "loss": 0.4746,
+      "step": 141340
+    },
+    {
+      "epoch": 7.020462898579518,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002383669414920036,
+      "loss": 0.4718,
+      "step": 141350
+    },
+    {
+      "epoch": 7.020959570875137,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00023832720770835403,
+      "loss": 0.4733,
+      "step": 141360
+    },
+    {
+      "epoch": 7.021456243170756,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00023828747392470447,
+      "loss": 0.4683,
+      "step": 141370
+    },
+    {
+      "epoch": 7.021952915466375,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00023824774014105494,
+      "loss": 0.4735,
+      "step": 141380
+    },
+    {
+      "epoch": 7.022449587761995,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00023820800635740539,
+      "loss": 0.4853,
+      "step": 141390
+    },
+    {
+      "epoch": 7.022946260057614,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00023816827257375586,
+      "loss": 0.4634,
+      "step": 141400
+    },
+    {
+      "epoch": 7.023442932353233,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002381285387901063,
+      "loss": 0.4686,
+      "step": 141410
+    },
+    {
+      "epoch": 7.023939604648852,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00023808880500645675,
+      "loss": 0.5208,
+      "step": 141420
+    },
+    {
+      "epoch": 7.024436276944472,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00023804907122280722,
+      "loss": 0.4788,
+      "step": 141430
+    },
+    {
+      "epoch": 7.024932949240092,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00023800933743915766,
+      "loss": 0.5135,
+      "step": 141440
+    },
+    {
+      "epoch": 7.025429621535711,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00023796960365550808,
+      "loss": 0.4973,
+      "step": 141450
+    },
+    {
+      "epoch": 7.02592629383133,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00023792986987185858,
+      "loss": 0.4715,
+      "step": 141460
+    },
+    {
+      "epoch": 7.026422966126949,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.000237890136088209,
+      "loss": 0.5018,
+      "step": 141470
+    },
+    {
+      "epoch": 7.0269196384225685,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0002378504023045595,
+      "loss": 0.5192,
+      "step": 141480
+    },
+    {
+      "epoch": 7.027416310718188,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002378106685209099,
+      "loss": 0.5151,
+      "step": 141490
+    },
+    {
+      "epoch": 7.027912983013808,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00023777093473726035,
+      "loss": 0.4586,
+      "step": 141500
+    },
+    {
+      "epoch": 7.028409655309427,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00023773120095361083,
+      "loss": 0.4731,
+      "step": 141510
+    },
+    {
+      "epoch": 7.028906327605046,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00023769146716996127,
+      "loss": 0.4952,
+      "step": 141520
+    },
+    {
+      "epoch": 7.0294029999006655,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00023765173338631171,
+      "loss": 0.4799,
+      "step": 141530
+    },
+    {
+      "epoch": 7.029899672196285,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00023761199960266219,
+      "loss": 0.5019,
+      "step": 141540
+    },
+    {
+      "epoch": 7.030396344491904,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00023757226581901263,
+      "loss": 0.4547,
+      "step": 141550
+    },
+    {
+      "epoch": 7.030893016787523,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0002375325320353631,
+      "loss": 0.502,
+      "step": 141560
+    },
+    {
+      "epoch": 7.031389689083143,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00023749279825171355,
+      "loss": 0.5338,
+      "step": 141570
+    },
+    {
+      "epoch": 7.0318863613787626,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00023745306446806396,
+      "loss": 0.4831,
+      "step": 141580
+    },
+    {
+      "epoch": 7.032383033674382,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00023741333068441446,
+      "loss": 0.4868,
+      "step": 141590
+    },
+    {
+      "epoch": 7.032879705970001,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00023737359690076488,
+      "loss": 0.482,
+      "step": 141600
+    },
+    {
+      "epoch": 7.03337637826562,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00023733386311711532,
+      "loss": 0.4651,
+      "step": 141610
+    },
+    {
+      "epoch": 7.0338730505612395,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0002372941293334658,
+      "loss": 0.47,
+      "step": 141620
+    },
+    {
+      "epoch": 7.034369722856859,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00023725439554981624,
+      "loss": 0.4866,
+      "step": 141630
+    },
+    {
+      "epoch": 7.034866395152479,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002372146617661667,
+      "loss": 0.4972,
+      "step": 141640
+    },
+    {
+      "epoch": 7.035363067448098,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00023717492798251715,
+      "loss": 0.4595,
+      "step": 141650
+    },
+    {
+      "epoch": 7.035859739743717,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00023713519419886757,
+      "loss": 0.459,
+      "step": 141660
+    },
+    {
+      "epoch": 7.0363564120393365,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00023709546041521807,
+      "loss": 0.491,
+      "step": 141670
+    },
+    {
+      "epoch": 7.036853084334956,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00023705572663156849,
+      "loss": 0.4738,
+      "step": 141680
+    },
+    {
+      "epoch": 7.037349756630575,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00023701599284791893,
+      "loss": 0.4718,
+      "step": 141690
+    },
+    {
+      "epoch": 7.037846428926194,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002369762590642694,
+      "loss": 0.4733,
+      "step": 141700
+    },
+    {
+      "epoch": 7.038343101221814,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00023693652528061985,
+      "loss": 0.4742,
+      "step": 141710
+    },
+    {
+      "epoch": 7.0388397735174335,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00023689679149697032,
+      "loss": 0.4897,
+      "step": 141720
+    },
+    {
+      "epoch": 7.039336445813053,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00023685705771332076,
+      "loss": 0.5069,
+      "step": 141730
+    },
+    {
+      "epoch": 7.039833118108672,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002368173239296712,
+      "loss": 0.4781,
+      "step": 141740
+    },
+    {
+      "epoch": 7.040329790404291,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00023677759014602168,
+      "loss": 0.4808,
+      "step": 141750
+    },
+    {
+      "epoch": 7.04082646269991,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00023673785636237212,
+      "loss": 0.5024,
+      "step": 141760
+    },
+    {
+      "epoch": 7.04132313499553,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002366981225787226,
+      "loss": 0.4859,
+      "step": 141770
+    },
+    {
+      "epoch": 7.041819807291149,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00023665838879507304,
+      "loss": 0.4987,
+      "step": 141780
+    },
+    {
+      "epoch": 7.042316479586769,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00023661865501142345,
+      "loss": 0.458,
+      "step": 141790
+    },
+    {
+      "epoch": 7.042813151882388,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00023657892122777395,
+      "loss": 0.5042,
+      "step": 141800
+    },
+    {
+      "epoch": 7.043309824178007,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00023653918744412437,
+      "loss": 0.4712,
+      "step": 141810
+    },
+    {
+      "epoch": 7.043806496473627,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00023649945366047481,
+      "loss": 0.4868,
+      "step": 141820
+    },
+    {
+      "epoch": 7.044303168769246,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00023645971987682529,
+      "loss": 0.4499,
+      "step": 141830
+    },
+    {
+      "epoch": 7.044799841064865,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00023641998609317573,
+      "loss": 0.4485,
+      "step": 141840
+    },
+    {
+      "epoch": 7.045296513360484,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002363802523095262,
+      "loss": 0.4767,
+      "step": 141850
+    },
+    {
+      "epoch": 7.0457931856561045,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00023634051852587665,
+      "loss": 0.485,
+      "step": 141860
+    },
+    {
+      "epoch": 7.046289857951724,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002363007847422271,
+      "loss": 0.4986,
+      "step": 141870
+    },
+    {
+      "epoch": 7.046786530247343,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00023626105095857756,
+      "loss": 0.4705,
+      "step": 141880
+    },
+    {
+      "epoch": 7.047283202542962,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.000236221317174928,
+      "loss": 0.4768,
+      "step": 141890
+    },
+    {
+      "epoch": 7.047779874838581,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00023618158339127842,
+      "loss": 0.4889,
+      "step": 141900
+    },
+    {
+      "epoch": 7.048276547134201,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002361418496076289,
+      "loss": 0.4702,
+      "step": 141910
+    },
+    {
+      "epoch": 7.04877321942982,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00023610211582397934,
+      "loss": 0.4579,
+      "step": 141920
+    },
+    {
+      "epoch": 7.04926989172544,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002360623820403298,
+      "loss": 0.4773,
+      "step": 141930
+    },
+    {
+      "epoch": 7.049766564021059,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00023602264825668025,
+      "loss": 0.4548,
+      "step": 141940
+    },
+    {
+      "epoch": 7.050263236316678,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002359829144730307,
+      "loss": 0.4606,
+      "step": 141950
+    },
+    {
+      "epoch": 7.050759908612298,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00023594318068938117,
+      "loss": 0.4856,
+      "step": 141960
+    },
+    {
+      "epoch": 7.051256580907917,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002359034469057316,
+      "loss": 0.5125,
+      "step": 141970
+    },
+    {
+      "epoch": 7.051753253203536,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00023586371312208203,
+      "loss": 0.4888,
+      "step": 141980
+    },
+    {
+      "epoch": 7.052249925499155,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00023582397933843253,
+      "loss": 0.4924,
+      "step": 141990
+    },
+    {
+      "epoch": 7.052746597794775,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00023578424555478295,
+      "loss": 0.4722,
+      "step": 142000
+    },
+    {
+      "epoch": 7.053243270090395,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00023574451177113344,
+      "loss": 0.4774,
+      "step": 142010
+    },
+    {
+      "epoch": 7.053739942386014,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00023570477798748386,
+      "loss": 0.4848,
+      "step": 142020
+    },
+    {
+      "epoch": 7.054236614681633,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0002356650442038343,
+      "loss": 0.4656,
+      "step": 142030
+    },
+    {
+      "epoch": 7.054733286977252,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00023562531042018478,
+      "loss": 0.5006,
+      "step": 142040
+    },
+    {
+      "epoch": 7.0552299592728716,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00023558557663653522,
+      "loss": 0.4561,
+      "step": 142050
+    },
+    {
+      "epoch": 7.055726631568491,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00023554584285288567,
+      "loss": 0.5041,
+      "step": 142060
+    },
+    {
+      "epoch": 7.056223303864111,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00023550610906923614,
+      "loss": 0.4918,
+      "step": 142070
+    },
+    {
+      "epoch": 7.05671997615973,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00023546637528558658,
+      "loss": 0.4901,
+      "step": 142080
+    },
+    {
+      "epoch": 7.057216648455349,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00023542664150193705,
+      "loss": 0.4613,
+      "step": 142090
+    },
+    {
+      "epoch": 7.057713320750969,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002353869077182875,
+      "loss": 0.4986,
+      "step": 142100
+    },
+    {
+      "epoch": 7.058209993046588,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00023534717393463791,
+      "loss": 0.4867,
+      "step": 142110
+    },
+    {
+      "epoch": 7.058706665342207,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0002353074401509884,
+      "loss": 0.4918,
+      "step": 142120
+    },
+    {
+      "epoch": 7.059203337637826,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00023526770636733883,
+      "loss": 0.4797,
+      "step": 142130
+    },
+    {
+      "epoch": 7.0597000099334455,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00023522797258368927,
+      "loss": 0.4604,
+      "step": 142140
+    },
+    {
+      "epoch": 7.060196682229066,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00023518823880003975,
+      "loss": 0.4562,
+      "step": 142150
+    },
+    {
+      "epoch": 7.060693354524685,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0002351485050163902,
+      "loss": 0.4999,
+      "step": 142160
+    },
+    {
+      "epoch": 7.061190026820304,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00023510877123274066,
+      "loss": 0.5229,
+      "step": 142170
+    },
+    {
+      "epoch": 7.061686699115923,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002350690374490911,
+      "loss": 0.4936,
+      "step": 142180
+    },
+    {
+      "epoch": 7.0621833714115425,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00023502930366544152,
+      "loss": 0.4659,
+      "step": 142190
+    },
+    {
+      "epoch": 7.062680043707162,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00023498956988179202,
+      "loss": 0.5114,
+      "step": 142200
+    },
+    {
+      "epoch": 7.063176716002781,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00023494983609814244,
+      "loss": 0.4818,
+      "step": 142210
+    },
+    {
+      "epoch": 7.063673388298401,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00023491010231449294,
+      "loss": 0.4942,
+      "step": 142220
+    },
+    {
+      "epoch": 7.06417006059402,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00023487036853084335,
+      "loss": 0.4806,
+      "step": 142230
+    },
+    {
+      "epoch": 7.0646667328896395,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002348306347471938,
+      "loss": 0.5028,
+      "step": 142240
+    },
+    {
+      "epoch": 7.065163405185259,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00023479090096354427,
+      "loss": 0.4712,
+      "step": 142250
+    },
+    {
+      "epoch": 7.065660077480878,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002347511671798947,
+      "loss": 0.4996,
+      "step": 142260
+    },
+    {
+      "epoch": 7.066156749776497,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00023471143339624516,
+      "loss": 0.4829,
+      "step": 142270
+    },
+    {
+      "epoch": 7.066653422072116,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00023467169961259563,
+      "loss": 0.5225,
+      "step": 142280
+    },
+    {
+      "epoch": 7.067150094367737,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00023463196582894607,
+      "loss": 0.4617,
+      "step": 142290
+    },
+    {
+      "epoch": 7.067646766663356,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00023459223204529654,
+      "loss": 0.4767,
+      "step": 142300
+    },
+    {
+      "epoch": 7.068143438958975,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.000234552498261647,
+      "loss": 0.5274,
+      "step": 142310
+    },
+    {
+      "epoch": 7.068640111254594,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002345127644779974,
+      "loss": 0.4848,
+      "step": 142320
+    },
+    {
+      "epoch": 7.0691367835502135,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002344730306943479,
+      "loss": 0.4745,
+      "step": 142330
+    },
+    {
+      "epoch": 7.069633455845833,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00023443329691069832,
+      "loss": 0.4779,
+      "step": 142340
+    },
+    {
+      "epoch": 7.070130128141452,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00023439356312704877,
+      "loss": 0.4938,
+      "step": 142350
+    },
+    {
+      "epoch": 7.070626800437072,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00023435382934339924,
+      "loss": 0.5046,
+      "step": 142360
+    },
+    {
+      "epoch": 7.071123472732691,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00023431409555974968,
+      "loss": 0.4688,
+      "step": 142370
+    },
+    {
+      "epoch": 7.0716201450283105,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00023427436177610015,
+      "loss": 0.4591,
+      "step": 142380
+    },
+    {
+      "epoch": 7.07211681732393,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0002342346279924506,
+      "loss": 0.4907,
+      "step": 142390
+    },
+    {
+      "epoch": 7.072613489619549,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00023419489420880104,
+      "loss": 0.4604,
+      "step": 142400
+    },
+    {
+      "epoch": 7.073110161915168,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002341551604251515,
+      "loss": 0.503,
+      "step": 142410
+    },
+    {
+      "epoch": 7.073606834210787,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.00023411542664150196,
+      "loss": 0.4991,
+      "step": 142420
+    },
+    {
+      "epoch": 7.0741035065064075,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00023407569285785237,
+      "loss": 0.472,
+      "step": 142430
+    },
+    {
+      "epoch": 7.074600178802027,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00023403595907420287,
+      "loss": 0.4888,
+      "step": 142440
+    },
+    {
+      "epoch": 7.075096851097646,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002339962252905533,
+      "loss": 0.4842,
+      "step": 142450
+    },
+    {
+      "epoch": 7.075593523393265,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00023395649150690376,
+      "loss": 0.4966,
+      "step": 142460
+    },
+    {
+      "epoch": 7.076090195688884,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002339167577232542,
+      "loss": 0.4971,
+      "step": 142470
+    },
+    {
+      "epoch": 7.076586867984504,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00023387702393960465,
+      "loss": 0.4729,
+      "step": 142480
+    },
+    {
+      "epoch": 7.077083540280123,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00023383729015595512,
+      "loss": 0.5048,
+      "step": 142490
+    },
+    {
+      "epoch": 7.077580212575742,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00023379755637230557,
+      "loss": 0.4961,
+      "step": 142500
+    },
+    {
+      "epoch": 7.078076884871362,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00023375782258865598,
+      "loss": 0.497,
+      "step": 142510
+    },
+    {
+      "epoch": 7.0785735571669814,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00023371808880500648,
+      "loss": 0.4709,
+      "step": 142520
+    },
+    {
+      "epoch": 7.079070229462601,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002336783550213569,
+      "loss": 0.514,
+      "step": 142530
+    },
+    {
+      "epoch": 7.07956690175822,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002336386212377074,
+      "loss": 0.4985,
+      "step": 142540
+    },
+    {
+      "epoch": 7.080063574053839,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002335988874540578,
+      "loss": 0.4723,
+      "step": 142550
+    },
+    {
+      "epoch": 7.080560246349458,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00023355915367040826,
+      "loss": 0.5067,
+      "step": 142560
+    },
+    {
+      "epoch": 7.081056918645078,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00023351941988675873,
+      "loss": 0.4608,
+      "step": 142570
+    },
+    {
+      "epoch": 7.081553590940698,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00023347968610310917,
+      "loss": 0.4775,
+      "step": 142580
+    },
+    {
+      "epoch": 7.082050263236317,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00023343995231945964,
+      "loss": 0.4837,
+      "step": 142590
+    },
+    {
+      "epoch": 7.082546935531936,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002334002185358101,
+      "loss": 0.4759,
+      "step": 142600
+    },
+    {
+      "epoch": 7.083043607827555,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00023336048475216053,
+      "loss": 0.4956,
+      "step": 142610
+    },
+    {
+      "epoch": 7.083540280123175,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.000233320750968511,
+      "loss": 0.4865,
+      "step": 142620
+    },
+    {
+      "epoch": 7.084036952418794,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00023328101718486145,
+      "loss": 0.4843,
+      "step": 142630
+    },
+    {
+      "epoch": 7.084533624714413,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00023324128340121187,
+      "loss": 0.4738,
+      "step": 142640
+    },
+    {
+      "epoch": 7.085030297010033,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00023320154961756236,
+      "loss": 0.4977,
+      "step": 142650
+    },
+    {
+      "epoch": 7.085526969305652,
+      "grad_norm": 0.125,
+      "learning_rate": 0.00023316181583391278,
+      "loss": 0.4792,
+      "step": 142660
+    },
+    {
+      "epoch": 7.086023641601272,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00023312208205026328,
+      "loss": 0.4879,
+      "step": 142670
+    },
+    {
+      "epoch": 7.086520313896891,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002330823482666137,
+      "loss": 0.5005,
+      "step": 142680
+    },
+    {
+      "epoch": 7.08701698619251,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00023304261448296414,
+      "loss": 0.4676,
+      "step": 142690
+    },
+    {
+      "epoch": 7.087513658488129,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002330028806993146,
+      "loss": 0.4853,
+      "step": 142700
+    },
+    {
+      "epoch": 7.0880103307837485,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00023296314691566506,
+      "loss": 0.5005,
+      "step": 142710
+    },
+    {
+      "epoch": 7.088507003079369,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002329234131320155,
+      "loss": 0.4629,
+      "step": 142720
+    },
+    {
+      "epoch": 7.089003675374988,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00023288367934836597,
+      "loss": 0.4831,
+      "step": 142730
+    },
+    {
+      "epoch": 7.089500347670607,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00023284394556471642,
+      "loss": 0.5082,
+      "step": 142740
+    },
+    {
+      "epoch": 7.089997019966226,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0002328042117810669,
+      "loss": 0.4694,
+      "step": 142750
+    },
+    {
+      "epoch": 7.090493692261846,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002327644779974173,
+      "loss": 0.4828,
+      "step": 142760
+    },
+    {
+      "epoch": 7.090990364557465,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00023272474421376775,
+      "loss": 0.5281,
+      "step": 142770
+    },
+    {
+      "epoch": 7.091487036853084,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00023268501043011822,
+      "loss": 0.4916,
+      "step": 142780
+    },
+    {
+      "epoch": 7.091983709148703,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00023264527664646867,
+      "loss": 0.49,
+      "step": 142790
+    },
+    {
+      "epoch": 7.092480381444323,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002326055428628191,
+      "loss": 0.4367,
+      "step": 142800
+    },
+    {
+      "epoch": 7.092977053739943,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00023256580907916958,
+      "loss": 0.4794,
+      "step": 142810
+    },
+    {
+      "epoch": 7.093473726035562,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00023252607529552003,
+      "loss": 0.468,
+      "step": 142820
+    },
+    {
+      "epoch": 7.093970398331181,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002324863415118705,
+      "loss": 0.506,
+      "step": 142830
+    },
+    {
+      "epoch": 7.0944670706268,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00023244660772822094,
+      "loss": 0.4884,
+      "step": 142840
+    },
+    {
+      "epoch": 7.0949637429224195,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00023240687394457136,
+      "loss": 0.5043,
+      "step": 142850
+    },
+    {
+      "epoch": 7.095460415218039,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00023236714016092186,
+      "loss": 0.4999,
+      "step": 142860
+    },
+    {
+      "epoch": 7.095957087513659,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00023232740637727227,
+      "loss": 0.4508,
+      "step": 142870
+    },
+    {
+      "epoch": 7.096453759809278,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00023228767259362272,
+      "loss": 0.4809,
+      "step": 142880
+    },
+    {
+      "epoch": 7.096950432104897,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002322479388099732,
+      "loss": 0.4774,
+      "step": 142890
+    },
+    {
+      "epoch": 7.0974471044005165,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00023220820502632363,
+      "loss": 0.4958,
+      "step": 142900
+    },
+    {
+      "epoch": 7.097943776696136,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002321684712426741,
+      "loss": 0.5162,
+      "step": 142910
+    },
+    {
+      "epoch": 7.098440448991755,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00023212873745902455,
+      "loss": 0.4788,
+      "step": 142920
+    },
+    {
+      "epoch": 7.098937121287374,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.000232089003675375,
+      "loss": 0.4658,
+      "step": 142930
+    },
+    {
+      "epoch": 7.099433793582994,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00023204926989172546,
+      "loss": 0.5278,
+      "step": 142940
+    },
+    {
+      "epoch": 7.0999304658786135,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002320095361080759,
+      "loss": 0.5104,
+      "step": 142950
+    },
+    {
+      "epoch": 7.100427138174233,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00023196980232442633,
+      "loss": 0.4983,
+      "step": 142960
+    },
+    {
+      "epoch": 7.100923810469852,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00023193006854077682,
+      "loss": 0.4831,
+      "step": 142970
+    },
+    {
+      "epoch": 7.101420482765471,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00023189033475712724,
+      "loss": 0.4977,
+      "step": 142980
+    },
+    {
+      "epoch": 7.1019171550610904,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00023185060097347774,
+      "loss": 0.481,
+      "step": 142990
+    },
+    {
+      "epoch": 7.10241382735671,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00023181086718982816,
+      "loss": 0.5166,
+      "step": 143000
+    },
+    {
+      "epoch": 7.10291049965233,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002317711334061786,
+      "loss": 0.5247,
+      "step": 143010
+    },
+    {
+      "epoch": 7.103407171947949,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00023173139962252907,
+      "loss": 0.4746,
+      "step": 143020
+    },
+    {
+      "epoch": 7.103903844243568,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00023169166583887952,
+      "loss": 0.5252,
+      "step": 143030
+    },
+    {
+      "epoch": 7.1044005165391875,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00023165193205523,
+      "loss": 0.4897,
+      "step": 143040
+    },
+    {
+      "epoch": 7.104897188834807,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00023161219827158043,
+      "loss": 0.4883,
+      "step": 143050
+    },
+    {
+      "epoch": 7.105393861130426,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00023157246448793085,
+      "loss": 0.5,
+      "step": 143060
+    },
+    {
+      "epoch": 7.105890533426045,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00023153273070428135,
+      "loss": 0.4721,
+      "step": 143070
+    },
+    {
+      "epoch": 7.106387205721665,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00023149299692063177,
+      "loss": 0.4643,
+      "step": 143080
+    },
+    {
+      "epoch": 7.1068838780172845,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002314532631369822,
+      "loss": 0.461,
+      "step": 143090
+    },
+    {
+      "epoch": 7.107380550312904,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00023141352935333268,
+      "loss": 0.5495,
+      "step": 143100
+    },
+    {
+      "epoch": 7.107877222608523,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00023137379556968313,
+      "loss": 0.4675,
+      "step": 143110
+    },
+    {
+      "epoch": 7.108373894904142,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002313340617860336,
+      "loss": 0.4878,
+      "step": 143120
+    },
+    {
+      "epoch": 7.108870567199761,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00023129432800238404,
+      "loss": 0.4986,
+      "step": 143130
+    },
+    {
+      "epoch": 7.109367239495381,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00023125459421873449,
+      "loss": 0.4905,
+      "step": 143140
+    },
+    {
+      "epoch": 7.109863911791001,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00023121486043508496,
+      "loss": 0.4642,
+      "step": 143150
+    },
+    {
+      "epoch": 7.11036058408662,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002311751266514354,
+      "loss": 0.4544,
+      "step": 143160
+    },
+    {
+      "epoch": 7.110857256382239,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00023113539286778582,
+      "loss": 0.4911,
+      "step": 143170
+    },
+    {
+      "epoch": 7.111353928677858,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00023109565908413632,
+      "loss": 0.4653,
+      "step": 143180
+    },
+    {
+      "epoch": 7.111850600973478,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00023105592530048673,
+      "loss": 0.4804,
+      "step": 143190
+    },
+    {
+      "epoch": 7.112347273269097,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00023101619151683723,
+      "loss": 0.4988,
+      "step": 143200
+    },
+    {
+      "epoch": 7.112843945564716,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00023097645773318765,
+      "loss": 0.4689,
+      "step": 143210
+    },
+    {
+      "epoch": 7.113340617860335,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002309367239495381,
+      "loss": 0.506,
+      "step": 143220
+    },
+    {
+      "epoch": 7.1138372901559555,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00023089699016588856,
+      "loss": 0.5052,
+      "step": 143230
+    },
+    {
+      "epoch": 7.114333962451575,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.000230857256382239,
+      "loss": 0.4921,
+      "step": 143240
+    },
+    {
+      "epoch": 7.114830634747194,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00023081752259858945,
+      "loss": 0.5034,
+      "step": 143250
+    },
+    {
+      "epoch": 7.115327307042813,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00023077778881493992,
+      "loss": 0.4747,
+      "step": 143260
+    },
+    {
+      "epoch": 7.115823979338432,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00023073805503129037,
+      "loss": 0.4833,
+      "step": 143270
+    },
+    {
+      "epoch": 7.116320651634052,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00023069832124764084,
+      "loss": 0.4776,
+      "step": 143280
+    },
+    {
+      "epoch": 7.116817323929671,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00023065858746399128,
+      "loss": 0.452,
+      "step": 143290
+    },
+    {
+      "epoch": 7.117313996225291,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002306188536803417,
+      "loss": 0.499,
+      "step": 143300
+    },
+    {
+      "epoch": 7.11781066852091,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00023057911989669217,
+      "loss": 0.5051,
+      "step": 143310
+    },
+    {
+      "epoch": 7.118307340816529,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00023053938611304262,
+      "loss": 0.487,
+      "step": 143320
+    },
+    {
+      "epoch": 7.118804013112149,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.00023049965232939306,
+      "loss": 0.505,
+      "step": 143330
+    },
+    {
+      "epoch": 7.119300685407768,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00023045991854574353,
+      "loss": 0.5393,
+      "step": 143340
+    },
+    {
+      "epoch": 7.119797357703387,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00023042018476209398,
+      "loss": 0.4801,
+      "step": 143350
+    },
+    {
+      "epoch": 7.120294029999006,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00023038045097844445,
+      "loss": 0.5147,
+      "step": 143360
+    },
+    {
+      "epoch": 7.120790702294626,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002303407171947949,
+      "loss": 0.4547,
+      "step": 143370
+    },
+    {
+      "epoch": 7.121287374590246,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0002303009834111453,
+      "loss": 0.4809,
+      "step": 143380
+    },
+    {
+      "epoch": 7.121784046885865,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002302612496274958,
+      "loss": 0.5019,
+      "step": 143390
+    },
+    {
+      "epoch": 7.122280719181484,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00023022151584384623,
+      "loss": 0.5201,
+      "step": 143400
+    },
+    {
+      "epoch": 7.122777391477103,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00023018178206019667,
+      "loss": 0.4755,
+      "step": 143410
+    },
+    {
+      "epoch": 7.1232740637727225,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00023014204827654714,
+      "loss": 0.4496,
+      "step": 143420
+    },
+    {
+      "epoch": 7.123770736068342,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00023010231449289759,
+      "loss": 0.5056,
+      "step": 143430
+    },
+    {
+      "epoch": 7.124267408363962,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00023006258070924806,
+      "loss": 0.4708,
+      "step": 143440
+    },
+    {
+      "epoch": 7.124764080659581,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002300228469255985,
+      "loss": 0.5275,
+      "step": 143450
+    },
+    {
+      "epoch": 7.1252607529552,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00022998311314194895,
+      "loss": 0.4782,
+      "step": 143460
+    },
+    {
+      "epoch": 7.12575742525082,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00022994337935829942,
+      "loss": 0.4861,
+      "step": 143470
+    },
+    {
+      "epoch": 7.126254097546439,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00022990364557464986,
+      "loss": 0.5009,
+      "step": 143480
+    },
+    {
+      "epoch": 7.126750769842058,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00022986391179100033,
+      "loss": 0.4984,
+      "step": 143490
+    },
+    {
+      "epoch": 7.127247442137677,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00022982417800735078,
+      "loss": 0.4795,
+      "step": 143500
+    },
+    {
+      "epoch": 7.1277441144332965,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002297844442237012,
+      "loss": 0.498,
+      "step": 143510
+    },
+    {
+      "epoch": 7.128240786728917,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002297447104400517,
+      "loss": 0.4772,
+      "step": 143520
+    },
+    {
+      "epoch": 7.128737459024536,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002297049766564021,
+      "loss": 0.4932,
+      "step": 143530
+    },
+    {
+      "epoch": 7.129234131320155,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00022966524287275255,
+      "loss": 0.5033,
+      "step": 143540
+    },
+    {
+      "epoch": 7.129730803615774,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00022962550908910302,
+      "loss": 0.4864,
+      "step": 143550
+    },
+    {
+      "epoch": 7.1302274759113935,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00022958577530545347,
+      "loss": 0.4603,
+      "step": 143560
+    },
+    {
+      "epoch": 7.130724148207013,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00022954604152180394,
+      "loss": 0.4982,
+      "step": 143570
+    },
+    {
+      "epoch": 7.131220820502632,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00022950630773815438,
+      "loss": 0.4506,
+      "step": 143580
+    },
+    {
+      "epoch": 7.131717492798252,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00022946657395450483,
+      "loss": 0.5013,
+      "step": 143590
+    },
+    {
+      "epoch": 7.132214165093871,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002294268401708553,
+      "loss": 0.4878,
+      "step": 143600
+    },
+    {
+      "epoch": 7.1327108373894905,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00022938710638720572,
+      "loss": 0.5103,
+      "step": 143610
+    },
+    {
+      "epoch": 7.13320750968511,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00022934737260355616,
+      "loss": 0.4955,
+      "step": 143620
+    },
+    {
+      "epoch": 7.133704181980729,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00022930763881990663,
+      "loss": 0.4691,
+      "step": 143630
+    },
+    {
+      "epoch": 7.134200854276348,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00022926790503625708,
+      "loss": 0.4869,
+      "step": 143640
+    },
+    {
+      "epoch": 7.134697526571967,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00022922817125260755,
+      "loss": 0.5061,
+      "step": 143650
+    },
+    {
+      "epoch": 7.1351941988675875,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.000229188437468958,
+      "loss": 0.4892,
+      "step": 143660
+    },
+    {
+      "epoch": 7.135690871163207,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00022914870368530844,
+      "loss": 0.4848,
+      "step": 143670
+    },
+    {
+      "epoch": 7.136187543458826,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002291089699016589,
+      "loss": 0.4725,
+      "step": 143680
+    },
+    {
+      "epoch": 7.136684215754445,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00022906923611800935,
+      "loss": 0.4675,
+      "step": 143690
+    },
+    {
+      "epoch": 7.1371808880500645,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00022902950233435977,
+      "loss": 0.4967,
+      "step": 143700
+    },
+    {
+      "epoch": 7.137677560345684,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00022898976855071027,
+      "loss": 0.486,
+      "step": 143710
+    },
+    {
+      "epoch": 7.138174232641303,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00022895003476706069,
+      "loss": 0.5012,
+      "step": 143720
+    },
+    {
+      "epoch": 7.138670904936923,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00022891030098341118,
+      "loss": 0.492,
+      "step": 143730
+    },
+    {
+      "epoch": 7.139167577232542,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002288705671997616,
+      "loss": 0.462,
+      "step": 143740
+    },
+    {
+      "epoch": 7.1396642495281615,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00022883083341611205,
+      "loss": 0.4899,
+      "step": 143750
+    },
+    {
+      "epoch": 7.140160921823781,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00022879109963246252,
+      "loss": 0.4951,
+      "step": 143760
+    },
+    {
+      "epoch": 7.1406575941194,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00022875136584881296,
+      "loss": 0.4571,
+      "step": 143770
+    },
+    {
+      "epoch": 7.141154266415019,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002287116320651634,
+      "loss": 0.4952,
+      "step": 143780
+    },
+    {
+      "epoch": 7.141650938710638,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.00022867189828151388,
+      "loss": 0.4873,
+      "step": 143790
+    },
+    {
+      "epoch": 7.1421476110062585,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00022863216449786432,
+      "loss": 0.4599,
+      "step": 143800
+    },
+    {
+      "epoch": 7.142644283301878,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002285924307142148,
+      "loss": 0.4905,
+      "step": 143810
+    },
+    {
+      "epoch": 7.143140955597497,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00022855269693056524,
+      "loss": 0.4881,
+      "step": 143820
+    },
+    {
+      "epoch": 7.143637627893116,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00022851296314691565,
+      "loss": 0.4771,
+      "step": 143830
+    },
+    {
+      "epoch": 7.144134300188735,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00022847322936326615,
+      "loss": 0.4965,
+      "step": 143840
+    },
+    {
+      "epoch": 7.144630972484355,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00022843349557961657,
+      "loss": 0.5024,
+      "step": 143850
+    },
+    {
+      "epoch": 7.145127644779974,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00022839376179596704,
+      "loss": 0.4866,
+      "step": 143860
+    },
+    {
+      "epoch": 7.145624317075594,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00022835402801231748,
+      "loss": 0.5024,
+      "step": 143870
+    },
+    {
+      "epoch": 7.146120989371213,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00022831429422866793,
+      "loss": 0.4544,
+      "step": 143880
+    },
+    {
+      "epoch": 7.146617661666832,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002282745604450184,
+      "loss": 0.4847,
+      "step": 143890
+    },
+    {
+      "epoch": 7.147114333962452,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.00022823482666136884,
+      "loss": 0.4629,
+      "step": 143900
+    },
+    {
+      "epoch": 7.147611006258071,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00022819509287771926,
+      "loss": 0.5042,
+      "step": 143910
+    },
+    {
+      "epoch": 7.14810767855369,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00022815535909406976,
+      "loss": 0.4641,
+      "step": 143920
+    },
+    {
+      "epoch": 7.148604350849309,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00022811562531042018,
+      "loss": 0.4813,
+      "step": 143930
+    },
+    {
+      "epoch": 7.149101023144929,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00022807589152677068,
+      "loss": 0.4937,
+      "step": 143940
+    },
+    {
+      "epoch": 7.149597695440549,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002280361577431211,
+      "loss": 0.4746,
+      "step": 143950
+    },
+    {
+      "epoch": 7.150094367736168,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00022799642395947154,
+      "loss": 0.4798,
+      "step": 143960
+    },
+    {
+      "epoch": 7.150591040031787,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.000227956690175822,
+      "loss": 0.5005,
+      "step": 143970
+    },
+    {
+      "epoch": 7.151087712327406,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00022791695639217245,
+      "loss": 0.4803,
+      "step": 143980
+    },
+    {
+      "epoch": 7.151584384623026,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002278772226085229,
+      "loss": 0.4651,
+      "step": 143990
+    },
+    {
+      "epoch": 7.152081056918645,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00022783748882487337,
+      "loss": 0.4723,
+      "step": 144000
+    },
+    {
+      "epoch": 7.152577729214264,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0002277977550412238,
+      "loss": 0.4977,
+      "step": 144010
+    },
+    {
+      "epoch": 7.153074401509884,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00022775802125757428,
+      "loss": 0.5051,
+      "step": 144020
+    },
+    {
+      "epoch": 7.153571073805503,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00022771828747392473,
+      "loss": 0.5129,
+      "step": 144030
+    },
+    {
+      "epoch": 7.154067746101123,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00022767855369027515,
+      "loss": 0.4887,
+      "step": 144040
+    },
+    {
+      "epoch": 7.154564418396742,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00022763881990662564,
+      "loss": 0.5253,
+      "step": 144050
+    },
+    {
+      "epoch": 7.155061090692361,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00022759908612297606,
+      "loss": 0.4954,
+      "step": 144060
+    },
+    {
+      "epoch": 7.15555776298798,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002275593523393265,
+      "loss": 0.4925,
+      "step": 144070
+    },
+    {
+      "epoch": 7.1560544352835995,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00022751961855567698,
+      "loss": 0.4706,
+      "step": 144080
+    },
+    {
+      "epoch": 7.15655110757922,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00022747988477202742,
+      "loss": 0.4558,
+      "step": 144090
+    },
+    {
+      "epoch": 7.157047779874839,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002274401509883779,
+      "loss": 0.4933,
+      "step": 144100
+    },
+    {
+      "epoch": 7.157544452170458,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00022740041720472834,
+      "loss": 0.484,
+      "step": 144110
+    },
+    {
+      "epoch": 7.158041124466077,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00022736068342107878,
+      "loss": 0.5118,
+      "step": 144120
+    },
+    {
+      "epoch": 7.1585377967616965,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00022732094963742925,
+      "loss": 0.4969,
+      "step": 144130
+    },
+    {
+      "epoch": 7.159034469057316,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002272812158537797,
+      "loss": 0.4972,
+      "step": 144140
+    },
+    {
+      "epoch": 7.159531141352935,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002272414820701301,
+      "loss": 0.5192,
+      "step": 144150
+    },
+    {
+      "epoch": 7.160027813648554,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00022720174828648058,
+      "loss": 0.4866,
+      "step": 144160
+    },
+    {
+      "epoch": 7.160524485944174,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00022716201450283103,
+      "loss": 0.4867,
+      "step": 144170
+    },
+    {
+      "epoch": 7.161021158239794,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002271222807191815,
+      "loss": 0.5067,
+      "step": 144180
+    },
+    {
+      "epoch": 7.161517830535413,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00022708254693553194,
+      "loss": 0.4791,
+      "step": 144190
+    },
+    {
+      "epoch": 7.162014502831032,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0002270428131518824,
+      "loss": 0.4805,
+      "step": 144200
+    },
+    {
+      "epoch": 7.162511175126651,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00022700307936823286,
+      "loss": 0.4814,
+      "step": 144210
+    },
+    {
+      "epoch": 7.1630078474222705,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002269633455845833,
+      "loss": 0.4961,
+      "step": 144220
+    },
+    {
+      "epoch": 7.16350451971789,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00022692361180093372,
+      "loss": 0.4874,
+      "step": 144230
+    },
+    {
+      "epoch": 7.16400119201351,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00022688387801728422,
+      "loss": 0.5099,
+      "step": 144240
+    },
+    {
+      "epoch": 7.164497864309129,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00022684414423363464,
+      "loss": 0.5031,
+      "step": 144250
+    },
+    {
+      "epoch": 7.164994536604748,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00022680441044998514,
+      "loss": 0.5086,
+      "step": 144260
+    },
+    {
+      "epoch": 7.1654912089003675,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00022676467666633555,
+      "loss": 0.4491,
+      "step": 144270
+    },
+    {
+      "epoch": 7.165987881195987,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.000226724942882686,
+      "loss": 0.4925,
+      "step": 144280
+    },
+    {
+      "epoch": 7.166484553491606,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00022668520909903647,
+      "loss": 0.4784,
+      "step": 144290
+    },
+    {
+      "epoch": 7.166981225787225,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0002266454753153869,
+      "loss": 0.5344,
+      "step": 144300
+    },
+    {
+      "epoch": 7.167477898082845,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00022660574153173738,
+      "loss": 0.4703,
+      "step": 144310
+    },
+    {
+      "epoch": 7.1679745703784645,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00022656600774808783,
+      "loss": 0.502,
+      "step": 144320
+    },
+    {
+      "epoch": 7.168471242674084,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00022652627396443827,
+      "loss": 0.4975,
+      "step": 144330
+    },
+    {
+      "epoch": 7.168967914969703,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00022648654018078874,
+      "loss": 0.4867,
+      "step": 144340
+    },
+    {
+      "epoch": 7.169464587265322,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002264468063971392,
+      "loss": 0.4843,
+      "step": 144350
+    },
+    {
+      "epoch": 7.169961259560941,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0002264070726134896,
+      "loss": 0.4771,
+      "step": 144360
+    },
+    {
+      "epoch": 7.170457931856561,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002263673388298401,
+      "loss": 0.4978,
+      "step": 144370
+    },
+    {
+      "epoch": 7.170954604152181,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00022632760504619052,
+      "loss": 0.4906,
+      "step": 144380
+    },
+    {
+      "epoch": 7.1714512764478,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00022628787126254102,
+      "loss": 0.4696,
+      "step": 144390
+    },
+    {
+      "epoch": 7.171947948743419,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00022624813747889144,
+      "loss": 0.4589,
+      "step": 144400
+    },
+    {
+      "epoch": 7.1724446210390385,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00022620840369524188,
+      "loss": 0.4837,
+      "step": 144410
+    },
+    {
+      "epoch": 7.172941293334658,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00022616866991159235,
+      "loss": 0.4701,
+      "step": 144420
+    },
+    {
+      "epoch": 7.173437965630277,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002261289361279428,
+      "loss": 0.4935,
+      "step": 144430
+    },
+    {
+      "epoch": 7.173934637925896,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00022608920234429324,
+      "loss": 0.5182,
+      "step": 144440
+    },
+    {
+      "epoch": 7.174431310221516,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002260494685606437,
+      "loss": 0.4686,
+      "step": 144450
+    },
+    {
+      "epoch": 7.1749279825171355,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00022600973477699413,
+      "loss": 0.4775,
+      "step": 144460
+    },
+    {
+      "epoch": 7.175424654812755,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00022597000099334463,
+      "loss": 0.5098,
+      "step": 144470
+    },
+    {
+      "epoch": 7.175921327108374,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00022593026720969504,
+      "loss": 0.4768,
+      "step": 144480
+    },
+    {
+      "epoch": 7.176417999403993,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002258905334260455,
+      "loss": 0.4488,
+      "step": 144490
+    },
+    {
+      "epoch": 7.176914671699612,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00022585079964239596,
+      "loss": 0.4852,
+      "step": 144500
+    },
+    {
+      "epoch": 7.177411343995232,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002258110658587464,
+      "loss": 0.4838,
+      "step": 144510
+    },
+    {
+      "epoch": 7.177908016290852,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00022577133207509685,
+      "loss": 0.4947,
+      "step": 144520
+    },
+    {
+      "epoch": 7.178404688586471,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00022573159829144732,
+      "loss": 0.4788,
+      "step": 144530
+    },
+    {
+      "epoch": 7.17890136088209,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00022569186450779776,
+      "loss": 0.4921,
+      "step": 144540
+    },
+    {
+      "epoch": 7.179398033177709,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00022565213072414824,
+      "loss": 0.473,
+      "step": 144550
+    },
+    {
+      "epoch": 7.179894705473329,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00022561239694049868,
+      "loss": 0.4832,
+      "step": 144560
+    },
+    {
+      "epoch": 7.180391377768948,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002255726631568491,
+      "loss": 0.4857,
+      "step": 144570
+    },
+    {
+      "epoch": 7.180888050064567,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002255329293731996,
+      "loss": 0.4534,
+      "step": 144580
+    },
+    {
+      "epoch": 7.181384722360186,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00022549319558955,
+      "loss": 0.5032,
+      "step": 144590
+    },
+    {
+      "epoch": 7.181881394655806,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00022545346180590046,
+      "loss": 0.4603,
+      "step": 144600
+    },
+    {
+      "epoch": 7.182378066951426,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00022541372802225093,
+      "loss": 0.4819,
+      "step": 144610
+    },
+    {
+      "epoch": 7.182874739247045,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00022537399423860137,
+      "loss": 0.5002,
+      "step": 144620
+    },
+    {
+      "epoch": 7.183371411542664,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00022533426045495184,
+      "loss": 0.4795,
+      "step": 144630
+    },
+    {
+      "epoch": 7.183868083838283,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0002252945266713023,
+      "loss": 0.4916,
+      "step": 144640
+    },
+    {
+      "epoch": 7.184364756133903,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00022525479288765273,
+      "loss": 0.4951,
+      "step": 144650
+    },
+    {
+      "epoch": 7.184861428429522,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002252150591040032,
+      "loss": 0.5165,
+      "step": 144660
+    },
+    {
+      "epoch": 7.185358100725142,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00022517532532035365,
+      "loss": 0.4848,
+      "step": 144670
+    },
+    {
+      "epoch": 7.185854773020761,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00022513559153670407,
+      "loss": 0.4596,
+      "step": 144680
+    },
+    {
+      "epoch": 7.18635144531638,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00022509585775305456,
+      "loss": 0.4846,
+      "step": 144690
+    },
+    {
+      "epoch": 7.186848117612,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00022505612396940498,
+      "loss": 0.4743,
+      "step": 144700
+    },
+    {
+      "epoch": 7.187344789907619,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00022501639018575545,
+      "loss": 0.5068,
+      "step": 144710
+    },
+    {
+      "epoch": 7.187841462203238,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002249766564021059,
+      "loss": 0.4759,
+      "step": 144720
+    },
+    {
+      "epoch": 7.188338134498857,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00022493692261845634,
+      "loss": 0.4783,
+      "step": 144730
+    },
+    {
+      "epoch": 7.188834806794477,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0002248971888348068,
+      "loss": 0.4935,
+      "step": 144740
+    },
+    {
+      "epoch": 7.189331479090097,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00022485745505115726,
+      "loss": 0.4897,
+      "step": 144750
+    },
+    {
+      "epoch": 7.189828151385716,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00022481772126750773,
+      "loss": 0.5137,
+      "step": 144760
+    },
+    {
+      "epoch": 7.190324823681335,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00022477798748385817,
+      "loss": 0.4782,
+      "step": 144770
+    },
+    {
+      "epoch": 7.190821495976954,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002247382537002086,
+      "loss": 0.4985,
+      "step": 144780
+    },
+    {
+      "epoch": 7.1913181682725735,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002246985199165591,
+      "loss": 0.4739,
+      "step": 144790
+    },
+    {
+      "epoch": 7.191814840568193,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002246587861329095,
+      "loss": 0.4992,
+      "step": 144800
+    },
+    {
+      "epoch": 7.192311512863813,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00022461905234925995,
+      "loss": 0.4707,
+      "step": 144810
+    },
+    {
+      "epoch": 7.192808185159432,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00022457931856561042,
+      "loss": 0.4752,
+      "step": 144820
+    },
+    {
+      "epoch": 7.193304857455051,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00022453958478196086,
+      "loss": 0.4861,
+      "step": 144830
+    },
+    {
+      "epoch": 7.1938015297506706,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00022449985099831134,
+      "loss": 0.4754,
+      "step": 144840
+    },
+    {
+      "epoch": 7.19429820204629,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00022446011721466178,
+      "loss": 0.483,
+      "step": 144850
+    },
+    {
+      "epoch": 7.194794874341909,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00022442038343101222,
+      "loss": 0.4937,
+      "step": 144860
+    },
+    {
+      "epoch": 7.195291546637528,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0002243806496473627,
+      "loss": 0.4744,
+      "step": 144870
+    },
+    {
+      "epoch": 7.1957882189331475,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00022434091586371314,
+      "loss": 0.502,
+      "step": 144880
+    },
+    {
+      "epoch": 7.196284891228768,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00022430118208006356,
+      "loss": 0.4647,
+      "step": 144890
+    },
+    {
+      "epoch": 7.196781563524387,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00022426144829641406,
+      "loss": 0.5108,
+      "step": 144900
+    },
+    {
+      "epoch": 7.197278235820006,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00022422171451276447,
+      "loss": 0.5119,
+      "step": 144910
+    },
+    {
+      "epoch": 7.197774908115625,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00022418198072911497,
+      "loss": 0.4668,
+      "step": 144920
+    },
+    {
+      "epoch": 7.1982715804112445,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002241422469454654,
+      "loss": 0.513,
+      "step": 144930
+    },
+    {
+      "epoch": 7.198768252706864,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00022410251316181583,
+      "loss": 0.4853,
+      "step": 144940
+    },
+    {
+      "epoch": 7.199264925002483,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002240627793781663,
+      "loss": 0.4552,
+      "step": 144950
+    },
+    {
+      "epoch": 7.199761597298103,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00022402304559451675,
+      "loss": 0.4684,
+      "step": 144960
+    },
+    {
+      "epoch": 7.200258269593722,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002239833118108672,
+      "loss": 0.4928,
+      "step": 144970
+    },
+    {
+      "epoch": 7.2007549418893415,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00022394357802721766,
+      "loss": 0.5074,
+      "step": 144980
+    },
+    {
+      "epoch": 7.201251614184961,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002239038442435681,
+      "loss": 0.4706,
+      "step": 144990
+    },
+    {
+      "epoch": 7.20174828648058,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00022386411045991858,
+      "loss": 0.5032,
+      "step": 145000
+    },
+    {
+      "epoch": 7.202244958776199,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.000223824376676269,
+      "loss": 0.4878,
+      "step": 145010
+    },
+    {
+      "epoch": 7.202741631071818,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00022378464289261944,
+      "loss": 0.4989,
+      "step": 145020
+    },
+    {
+      "epoch": 7.2032383033674385,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0002237449091089699,
+      "loss": 0.4722,
+      "step": 145030
+    },
+    {
+      "epoch": 7.203734975663058,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00022370517532532036,
+      "loss": 0.4766,
+      "step": 145040
+    },
+    {
+      "epoch": 7.204231647958677,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002236654415416708,
+      "loss": 0.5251,
+      "step": 145050
+    },
+    {
+      "epoch": 7.204728320254296,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00022362570775802127,
+      "loss": 0.4952,
+      "step": 145060
+    },
+    {
+      "epoch": 7.205224992549915,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00022358597397437172,
+      "loss": 0.4873,
+      "step": 145070
+    },
+    {
+      "epoch": 7.205721664845535,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002235462401907222,
+      "loss": 0.4997,
+      "step": 145080
+    },
+    {
+      "epoch": 7.206218337141154,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00022350650640707263,
+      "loss": 0.4896,
+      "step": 145090
+    },
+    {
+      "epoch": 7.206715009436774,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00022346677262342305,
+      "loss": 0.4994,
+      "step": 145100
+    },
+    {
+      "epoch": 7.207211681732393,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00022342703883977355,
+      "loss": 0.5009,
+      "step": 145110
+    },
+    {
+      "epoch": 7.2077083540280125,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00022338730505612396,
+      "loss": 0.4934,
+      "step": 145120
+    },
+    {
+      "epoch": 7.208205026323632,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00022334757127247446,
+      "loss": 0.4681,
+      "step": 145130
+    },
+    {
+      "epoch": 7.208701698619251,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00022330783748882488,
+      "loss": 0.4923,
+      "step": 145140
+    },
+    {
+      "epoch": 7.20919837091487,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00022326810370517532,
+      "loss": 0.4874,
+      "step": 145150
+    },
+    {
+      "epoch": 7.209695043210489,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002232283699215258,
+      "loss": 0.501,
+      "step": 145160
+    },
+    {
+      "epoch": 7.2101917155061095,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00022318863613787624,
+      "loss": 0.4875,
+      "step": 145170
+    },
+    {
+      "epoch": 7.210688387801729,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00022314890235422668,
+      "loss": 0.4814,
+      "step": 145180
+    },
+    {
+      "epoch": 7.211185060097348,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00022310916857057716,
+      "loss": 0.4958,
+      "step": 145190
+    },
+    {
+      "epoch": 7.211681732392967,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002230694347869276,
+      "loss": 0.4566,
+      "step": 145200
+    },
+    {
+      "epoch": 7.212178404688586,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00022302970100327807,
+      "loss": 0.5105,
+      "step": 145210
+    },
+    {
+      "epoch": 7.212675076984206,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00022298996721962852,
+      "loss": 0.4881,
+      "step": 145220
+    },
+    {
+      "epoch": 7.213171749279825,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00022295023343597893,
+      "loss": 0.4852,
+      "step": 145230
+    },
+    {
+      "epoch": 7.213668421575445,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00022291049965232943,
+      "loss": 0.4856,
+      "step": 145240
+    },
+    {
+      "epoch": 7.214165093871064,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00022287076586867985,
+      "loss": 0.5042,
+      "step": 145250
+    },
+    {
+      "epoch": 7.214661766166683,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002228310320850303,
+      "loss": 0.4929,
+      "step": 145260
+    },
+    {
+      "epoch": 7.215158438462303,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00022279129830138076,
+      "loss": 0.4727,
+      "step": 145270
+    },
+    {
+      "epoch": 7.215655110757922,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0002227515645177312,
+      "loss": 0.5022,
+      "step": 145280
+    },
+    {
+      "epoch": 7.216151783053541,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00022271183073408168,
+      "loss": 0.4897,
+      "step": 145290
+    },
+    {
+      "epoch": 7.21664845534916,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00022267209695043212,
+      "loss": 0.5057,
+      "step": 145300
+    },
+    {
+      "epoch": 7.21714512764478,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00022263236316678254,
+      "loss": 0.464,
+      "step": 145310
+    },
+    {
+      "epoch": 7.2176417999404,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00022259262938313304,
+      "loss": 0.4611,
+      "step": 145320
+    },
+    {
+      "epoch": 7.218138472236019,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00022255289559948346,
+      "loss": 0.4799,
+      "step": 145330
+    },
+    {
+      "epoch": 7.218635144531638,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002225131618158339,
+      "loss": 0.4788,
+      "step": 145340
+    },
+    {
+      "epoch": 7.219131816827257,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00022247342803218437,
+      "loss": 0.4758,
+      "step": 145350
+    },
+    {
+      "epoch": 7.219628489122877,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00022243369424853482,
+      "loss": 0.4607,
+      "step": 145360
+    },
+    {
+      "epoch": 7.220125161418496,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002223939604648853,
+      "loss": 0.512,
+      "step": 145370
+    },
+    {
+      "epoch": 7.220621833714115,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00022235422668123573,
+      "loss": 0.4601,
+      "step": 145380
+    },
+    {
+      "epoch": 7.221118506009735,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00022231449289758618,
+      "loss": 0.4845,
+      "step": 145390
+    },
+    {
+      "epoch": 7.221615178305354,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00022227475911393665,
+      "loss": 0.4872,
+      "step": 145400
+    },
+    {
+      "epoch": 7.222111850600974,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002222350253302871,
+      "loss": 0.4676,
+      "step": 145410
+    },
+    {
+      "epoch": 7.222608522896593,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0002221952915466375,
+      "loss": 0.4826,
+      "step": 145420
+    },
+    {
+      "epoch": 7.223105195192212,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.000222155557762988,
+      "loss": 0.5007,
+      "step": 145430
+    },
+    {
+      "epoch": 7.223601867487831,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00022211582397933842,
+      "loss": 0.4959,
+      "step": 145440
+    },
+    {
+      "epoch": 7.2240985397834505,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00022207609019568892,
+      "loss": 0.5069,
+      "step": 145450
+    },
+    {
+      "epoch": 7.224595212079071,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00022203635641203934,
+      "loss": 0.5187,
+      "step": 145460
+    },
+    {
+      "epoch": 7.22509188437469,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00022199662262838978,
+      "loss": 0.4887,
+      "step": 145470
+    },
+    {
+      "epoch": 7.225588556670309,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00022195688884474026,
+      "loss": 0.4891,
+      "step": 145480
+    },
+    {
+      "epoch": 7.226085228965928,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002219171550610907,
+      "loss": 0.5209,
+      "step": 145490
+    },
+    {
+      "epoch": 7.2265819012615475,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00022187742127744114,
+      "loss": 0.4798,
+      "step": 145500
+    },
+    {
+      "epoch": 7.227078573557167,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00022183768749379162,
+      "loss": 0.529,
+      "step": 145510
+    },
+    {
+      "epoch": 7.227575245852786,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00022179795371014206,
+      "loss": 0.4911,
+      "step": 145520
+    },
+    {
+      "epoch": 7.228071918148406,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00022175821992649253,
+      "loss": 0.499,
+      "step": 145530
+    },
+    {
+      "epoch": 7.228568590444025,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00022171848614284298,
+      "loss": 0.49,
+      "step": 145540
+    },
+    {
+      "epoch": 7.229065262739645,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002216787523591934,
+      "loss": 0.5261,
+      "step": 145550
+    },
+    {
+      "epoch": 7.229561935035264,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00022163901857554386,
+      "loss": 0.4873,
+      "step": 145560
+    },
+    {
+      "epoch": 7.230058607330883,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0002215992847918943,
+      "loss": 0.494,
+      "step": 145570
+    },
+    {
+      "epoch": 7.230555279626502,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00022155955100824478,
+      "loss": 0.4886,
+      "step": 145580
+    },
+    {
+      "epoch": 7.2310519519221215,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00022151981722459522,
+      "loss": 0.4963,
+      "step": 145590
+    },
+    {
+      "epoch": 7.231548624217741,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00022148008344094567,
+      "loss": 0.4728,
+      "step": 145600
+    },
+    {
+      "epoch": 7.232045296513361,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00022144034965729614,
+      "loss": 0.4815,
+      "step": 145610
+    },
+    {
+      "epoch": 7.23254196880898,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00022140061587364658,
+      "loss": 0.5064,
+      "step": 145620
+    },
+    {
+      "epoch": 7.233038641104599,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.000221360882089997,
+      "loss": 0.4569,
+      "step": 145630
+    },
+    {
+      "epoch": 7.2335353134002185,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002213211483063475,
+      "loss": 0.4598,
+      "step": 145640
+    },
+    {
+      "epoch": 7.234031985695838,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00022128141452269792,
+      "loss": 0.4865,
+      "step": 145650
+    },
+    {
+      "epoch": 7.234528657991457,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00022124168073904841,
+      "loss": 0.5023,
+      "step": 145660
+    },
+    {
+      "epoch": 7.235025330287076,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00022120194695539883,
+      "loss": 0.4864,
+      "step": 145670
+    },
+    {
+      "epoch": 7.235522002582696,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00022116221317174928,
+      "loss": 0.5131,
+      "step": 145680
+    },
+    {
+      "epoch": 7.2360186748783155,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00022112247938809975,
+      "loss": 0.4828,
+      "step": 145690
+    },
+    {
+      "epoch": 7.236515347173935,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002210827456044502,
+      "loss": 0.4809,
+      "step": 145700
+    },
+    {
+      "epoch": 7.237012019469554,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00022104301182080064,
+      "loss": 0.4749,
+      "step": 145710
+    },
+    {
+      "epoch": 7.237508691765173,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002210032780371511,
+      "loss": 0.5027,
+      "step": 145720
+    },
+    {
+      "epoch": 7.238005364060792,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00022096354425350155,
+      "loss": 0.521,
+      "step": 145730
+    },
+    {
+      "epoch": 7.238502036356412,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00022092381046985202,
+      "loss": 0.4808,
+      "step": 145740
+    },
+    {
+      "epoch": 7.238998708652032,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00022088407668620247,
+      "loss": 0.5053,
+      "step": 145750
+    },
+    {
+      "epoch": 7.239495380947651,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00022084434290255288,
+      "loss": 0.4967,
+      "step": 145760
+    },
+    {
+      "epoch": 7.23999205324327,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00022080460911890338,
+      "loss": 0.4675,
+      "step": 145770
+    },
+    {
+      "epoch": 7.2404887255388894,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002207648753352538,
+      "loss": 0.5069,
+      "step": 145780
+    },
+    {
+      "epoch": 7.240985397834509,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00022072514155160424,
+      "loss": 0.4736,
+      "step": 145790
+    },
+    {
+      "epoch": 7.241482070130128,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00022068540776795472,
+      "loss": 0.4937,
+      "step": 145800
+    },
+    {
+      "epoch": 7.241978742425747,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00022064567398430516,
+      "loss": 0.4887,
+      "step": 145810
+    },
+    {
+      "epoch": 7.242475414721367,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00022060594020065563,
+      "loss": 0.5057,
+      "step": 145820
+    },
+    {
+      "epoch": 7.2429720870169865,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00022056620641700608,
+      "loss": 0.4826,
+      "step": 145830
+    },
+    {
+      "epoch": 7.243468759312606,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00022052647263335652,
+      "loss": 0.4793,
+      "step": 145840
+    },
+    {
+      "epoch": 7.243965431608225,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.000220486738849707,
+      "loss": 0.4725,
+      "step": 145850
+    },
+    {
+      "epoch": 7.244462103903844,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002204470050660574,
+      "loss": 0.5033,
+      "step": 145860
+    },
+    {
+      "epoch": 7.244958776199463,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00022040727128240785,
+      "loss": 0.4814,
+      "step": 145870
+    },
+    {
+      "epoch": 7.245455448495083,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00022036753749875832,
+      "loss": 0.4836,
+      "step": 145880
+    },
+    {
+      "epoch": 7.245952120790703,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00022032780371510877,
+      "loss": 0.4845,
+      "step": 145890
+    },
+    {
+      "epoch": 7.246448793086322,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00022028806993145924,
+      "loss": 0.4973,
+      "step": 145900
+    },
+    {
+      "epoch": 7.246945465381941,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00022024833614780968,
+      "loss": 0.4808,
+      "step": 145910
+    },
+    {
+      "epoch": 7.24744213767756,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00022020860236416013,
+      "loss": 0.494,
+      "step": 145920
+    },
+    {
+      "epoch": 7.24793880997318,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002201688685805106,
+      "loss": 0.4976,
+      "step": 145930
+    },
+    {
+      "epoch": 7.248435482268799,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00022012913479686104,
+      "loss": 0.4904,
+      "step": 145940
+    },
+    {
+      "epoch": 7.248932154564418,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00022008940101321146,
+      "loss": 0.4855,
+      "step": 145950
+    },
+    {
+      "epoch": 7.249428826860038,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00022004966722956196,
+      "loss": 0.4814,
+      "step": 145960
+    },
+    {
+      "epoch": 7.249925499155657,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00022000993344591238,
+      "loss": 0.4844,
+      "step": 145970
+    },
+    {
+      "epoch": 7.250422171451277,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00021997019966226287,
+      "loss": 0.4592,
+      "step": 145980
+    },
+    {
+      "epoch": 7.250918843746896,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002199304658786133,
+      "loss": 0.4822,
+      "step": 145990
+    },
+    {
+      "epoch": 7.251415516042515,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00021989073209496374,
+      "loss": 0.5044,
+      "step": 146000
+    },
+    {
+      "epoch": 7.251912188338134,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002198509983113142,
+      "loss": 0.4562,
+      "step": 146010
+    },
+    {
+      "epoch": 7.252408860633754,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00021981126452766465,
+      "loss": 0.4825,
+      "step": 146020
+    },
+    {
+      "epoch": 7.252905532929373,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00021977153074401512,
+      "loss": 0.4722,
+      "step": 146030
+    },
+    {
+      "epoch": 7.253402205224993,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00021973179696036557,
+      "loss": 0.5035,
+      "step": 146040
+    },
+    {
+      "epoch": 7.253898877520612,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.000219692063176716,
+      "loss": 0.4719,
+      "step": 146050
+    },
+    {
+      "epoch": 7.254395549816231,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00021965232939306648,
+      "loss": 0.471,
+      "step": 146060
+    },
+    {
+      "epoch": 7.254892222111851,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00021961259560941693,
+      "loss": 0.4632,
+      "step": 146070
+    },
+    {
+      "epoch": 7.25538889440747,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00021957286182576734,
+      "loss": 0.4776,
+      "step": 146080
+    },
+    {
+      "epoch": 7.255885566703089,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00021953312804211784,
+      "loss": 0.4698,
+      "step": 146090
+    },
+    {
+      "epoch": 7.256382238998708,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.00021949339425846826,
+      "loss": 0.4762,
+      "step": 146100
+    },
+    {
+      "epoch": 7.256878911294328,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00021945366047481876,
+      "loss": 0.4938,
+      "step": 146110
+    },
+    {
+      "epoch": 7.257375583589948,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00021941392669116918,
+      "loss": 0.4769,
+      "step": 146120
+    },
+    {
+      "epoch": 7.257872255885567,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00021937419290751962,
+      "loss": 0.5028,
+      "step": 146130
+    },
+    {
+      "epoch": 7.258368928181186,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002193344591238701,
+      "loss": 0.5018,
+      "step": 146140
+    },
+    {
+      "epoch": 7.258865600476805,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00021929472534022054,
+      "loss": 0.4779,
+      "step": 146150
+    },
+    {
+      "epoch": 7.2593622727724245,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00021925499155657095,
+      "loss": 0.4888,
+      "step": 146160
+    },
+    {
+      "epoch": 7.259858945068044,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00021921525777292145,
+      "loss": 0.5051,
+      "step": 146170
+    },
+    {
+      "epoch": 7.260355617363664,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00021917552398927187,
+      "loss": 0.5003,
+      "step": 146180
+    },
+    {
+      "epoch": 7.260852289659283,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00021913579020562237,
+      "loss": 0.48,
+      "step": 146190
+    },
+    {
+      "epoch": 7.261348961954902,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00021909605642197278,
+      "loss": 0.493,
+      "step": 146200
+    },
+    {
+      "epoch": 7.2618456342505215,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00021905632263832323,
+      "loss": 0.4842,
+      "step": 146210
+    },
+    {
+      "epoch": 7.262342306546141,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0002190165888546737,
+      "loss": 0.493,
+      "step": 146220
+    },
+    {
+      "epoch": 7.26283897884176,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.00021897685507102414,
+      "loss": 0.4798,
+      "step": 146230
+    },
+    {
+      "epoch": 7.263335651137379,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002189371212873746,
+      "loss": 0.4905,
+      "step": 146240
+    },
+    {
+      "epoch": 7.2638323234329985,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00021889738750372506,
+      "loss": 0.4747,
+      "step": 146250
+    },
+    {
+      "epoch": 7.264328995728619,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002188576537200755,
+      "loss": 0.4507,
+      "step": 146260
+    },
+    {
+      "epoch": 7.264825668024238,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00021881791993642598,
+      "loss": 0.4637,
+      "step": 146270
+    },
+    {
+      "epoch": 7.265322340319857,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00021877818615277642,
+      "loss": 0.4874,
+      "step": 146280
+    },
+    {
+      "epoch": 7.265819012615476,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00021873845236912684,
+      "loss": 0.5362,
+      "step": 146290
+    },
+    {
+      "epoch": 7.2663156849110955,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00021869871858547733,
+      "loss": 0.477,
+      "step": 146300
+    },
+    {
+      "epoch": 7.266812357206715,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00021865898480182775,
+      "loss": 0.4905,
+      "step": 146310
+    },
+    {
+      "epoch": 7.267309029502334,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0002186192510181782,
+      "loss": 0.4728,
+      "step": 146320
+    },
+    {
+      "epoch": 7.267805701797954,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00021857951723452867,
+      "loss": 0.4751,
+      "step": 146330
+    },
+    {
+      "epoch": 7.268302374093573,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0002185397834508791,
+      "loss": 0.5212,
+      "step": 146340
+    },
+    {
+      "epoch": 7.2687990463891925,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00021850004966722958,
+      "loss": 0.4893,
+      "step": 146350
+    },
+    {
+      "epoch": 7.269295718684812,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00021846031588358003,
+      "loss": 0.4892,
+      "step": 146360
+    },
+    {
+      "epoch": 7.269792390980431,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00021842058209993047,
+      "loss": 0.4533,
+      "step": 146370
+    },
+    {
+      "epoch": 7.27028906327605,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00021838084831628094,
+      "loss": 0.4917,
+      "step": 146380
+    },
+    {
+      "epoch": 7.270785735571669,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002183411145326314,
+      "loss": 0.4851,
+      "step": 146390
+    },
+    {
+      "epoch": 7.2712824078672895,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00021830138074898186,
+      "loss": 0.5024,
+      "step": 146400
+    },
+    {
+      "epoch": 7.271779080162909,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00021826164696533228,
+      "loss": 0.5082,
+      "step": 146410
+    },
+    {
+      "epoch": 7.272275752458528,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00021822191318168272,
+      "loss": 0.465,
+      "step": 146420
+    },
+    {
+      "epoch": 7.272772424754147,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002181821793980332,
+      "loss": 0.496,
+      "step": 146430
+    },
+    {
+      "epoch": 7.273269097049766,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00021814244561438364,
+      "loss": 0.4859,
+      "step": 146440
+    },
+    {
+      "epoch": 7.273765769345386,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00021810271183073408,
+      "loss": 0.502,
+      "step": 146450
+    },
+    {
+      "epoch": 7.274262441641005,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00021806297804708455,
+      "loss": 0.5207,
+      "step": 146460
+    },
+    {
+      "epoch": 7.274759113936625,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.000218023244263435,
+      "loss": 0.4643,
+      "step": 146470
+    },
+    {
+      "epoch": 7.275255786232244,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00021798351047978547,
+      "loss": 0.4876,
+      "step": 146480
+    },
+    {
+      "epoch": 7.2757524585278635,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002179437766961359,
+      "loss": 0.4683,
+      "step": 146490
+    },
+    {
+      "epoch": 7.276249130823483,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00021790404291248633,
+      "loss": 0.4881,
+      "step": 146500
+    },
+    {
+      "epoch": 7.276745803119102,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00021786430912883683,
+      "loss": 0.4642,
+      "step": 146510
+    },
+    {
+      "epoch": 7.277242475414721,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00021782457534518724,
+      "loss": 0.4739,
+      "step": 146520
+    },
+    {
+      "epoch": 7.27773914771034,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002177848415615377,
+      "loss": 0.467,
+      "step": 146530
+    },
+    {
+      "epoch": 7.2782358200059605,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00021774510777788816,
+      "loss": 0.5216,
+      "step": 146540
+    },
+    {
+      "epoch": 7.27873249230158,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002177053739942386,
+      "loss": 0.5084,
+      "step": 146550
+    },
+    {
+      "epoch": 7.279229164597199,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00021766564021058908,
+      "loss": 0.5143,
+      "step": 146560
+    },
+    {
+      "epoch": 7.279725836892818,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00021762590642693952,
+      "loss": 0.519,
+      "step": 146570
+    },
+    {
+      "epoch": 7.280222509188437,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00021758617264328996,
+      "loss": 0.4847,
+      "step": 146580
+    },
+    {
+      "epoch": 7.280719181484057,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00021754643885964043,
+      "loss": 0.4794,
+      "step": 146590
+    },
+    {
+      "epoch": 7.281215853779676,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00021750670507599088,
+      "loss": 0.4951,
+      "step": 146600
+    },
+    {
+      "epoch": 7.281712526075296,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002174669712923413,
+      "loss": 0.5,
+      "step": 146610
+    },
+    {
+      "epoch": 7.282209198370915,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002174272375086918,
+      "loss": 0.4903,
+      "step": 146620
+    },
+    {
+      "epoch": 7.282705870666534,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002173875037250422,
+      "loss": 0.4716,
+      "step": 146630
+    },
+    {
+      "epoch": 7.283202542962154,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002173477699413927,
+      "loss": 0.4708,
+      "step": 146640
+    },
+    {
+      "epoch": 7.283699215257773,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00021730803615774313,
+      "loss": 0.5213,
+      "step": 146650
+    },
+    {
+      "epoch": 7.284195887553392,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00021726830237409357,
+      "loss": 0.5167,
+      "step": 146660
+    },
+    {
+      "epoch": 7.284692559849011,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00021722856859044404,
+      "loss": 0.4967,
+      "step": 146670
+    },
+    {
+      "epoch": 7.285189232144631,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0002171888348067945,
+      "loss": 0.5061,
+      "step": 146680
+    },
+    {
+      "epoch": 7.285685904440251,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00021714910102314493,
+      "loss": 0.4926,
+      "step": 146690
+    },
+    {
+      "epoch": 7.28618257673587,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002171093672394954,
+      "loss": 0.4887,
+      "step": 146700
+    },
+    {
+      "epoch": 7.286679249031489,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00021706963345584582,
+      "loss": 0.5073,
+      "step": 146710
+    },
+    {
+      "epoch": 7.287175921327108,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00021702989967219632,
+      "loss": 0.4765,
+      "step": 146720
+    },
+    {
+      "epoch": 7.287672593622728,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00021699016588854674,
+      "loss": 0.4991,
+      "step": 146730
+    },
+    {
+      "epoch": 7.288169265918347,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00021695043210489718,
+      "loss": 0.4757,
+      "step": 146740
+    },
+    {
+      "epoch": 7.288665938213966,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00021691069832124765,
+      "loss": 0.4766,
+      "step": 146750
+    },
+    {
+      "epoch": 7.289162610509586,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002168709645375981,
+      "loss": 0.5002,
+      "step": 146760
+    },
+    {
+      "epoch": 7.289659282805205,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00021683123075394854,
+      "loss": 0.4682,
+      "step": 146770
+    },
+    {
+      "epoch": 7.290155955100825,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.000216791496970299,
+      "loss": 0.4651,
+      "step": 146780
+    },
+    {
+      "epoch": 7.290652627396444,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00021675176318664946,
+      "loss": 0.4819,
+      "step": 146790
+    },
+    {
+      "epoch": 7.291149299692063,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00021671202940299993,
+      "loss": 0.4793,
+      "step": 146800
+    },
+    {
+      "epoch": 7.291645971987682,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00021667229561935037,
+      "loss": 0.4844,
+      "step": 146810
+    },
+    {
+      "epoch": 7.2921426442833015,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002166325618357008,
+      "loss": 0.4949,
+      "step": 146820
+    },
+    {
+      "epoch": 7.292639316578922,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002165928280520513,
+      "loss": 0.4655,
+      "step": 146830
+    },
+    {
+      "epoch": 7.293135988874541,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002165530942684017,
+      "loss": 0.4366,
+      "step": 146840
+    },
+    {
+      "epoch": 7.29363266117016,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002165133604847522,
+      "loss": 0.4994,
+      "step": 146850
+    },
+    {
+      "epoch": 7.294129333465779,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00021647362670110262,
+      "loss": 0.4936,
+      "step": 146860
+    },
+    {
+      "epoch": 7.2946260057613985,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00021643389291745306,
+      "loss": 0.4645,
+      "step": 146870
+    },
+    {
+      "epoch": 7.295122678057018,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00021639415913380354,
+      "loss": 0.4965,
+      "step": 146880
+    },
+    {
+      "epoch": 7.295619350352637,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00021635442535015398,
+      "loss": 0.5206,
+      "step": 146890
+    },
+    {
+      "epoch": 7.296116022648257,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00021631469156650442,
+      "loss": 0.4899,
+      "step": 146900
+    },
+    {
+      "epoch": 7.296612694943876,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002162749577828549,
+      "loss": 0.4816,
+      "step": 146910
+    },
+    {
+      "epoch": 7.2971093672394955,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00021623522399920534,
+      "loss": 0.4872,
+      "step": 146920
+    },
+    {
+      "epoch": 7.297606039535115,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002161954902155558,
+      "loss": 0.5007,
+      "step": 146930
+    },
+    {
+      "epoch": 7.298102711830734,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00021615575643190625,
+      "loss": 0.4545,
+      "step": 146940
+    },
+    {
+      "epoch": 7.298599384126353,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00021611602264825667,
+      "loss": 0.484,
+      "step": 146950
+    },
+    {
+      "epoch": 7.2990960564219725,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00021607628886460717,
+      "loss": 0.4755,
+      "step": 146960
+    },
+    {
+      "epoch": 7.299592728717592,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0002160365550809576,
+      "loss": 0.4786,
+      "step": 146970
+    },
+    {
+      "epoch": 7.300089401013212,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00021599682129730803,
+      "loss": 0.4916,
+      "step": 146980
+    },
+    {
+      "epoch": 7.300586073308831,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002159570875136585,
+      "loss": 0.4628,
+      "step": 146990
+    },
+    {
+      "epoch": 7.30108274560445,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00021591735373000895,
+      "loss": 0.4911,
+      "step": 147000
+    },
+    {
+      "epoch": 7.3015794179000695,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00021587761994635942,
+      "loss": 0.4863,
+      "step": 147010
+    },
+    {
+      "epoch": 7.302076090195689,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00021583788616270986,
+      "loss": 0.4797,
+      "step": 147020
+    },
+    {
+      "epoch": 7.302572762491308,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00021579815237906028,
+      "loss": 0.4951,
+      "step": 147030
+    },
+    {
+      "epoch": 7.303069434786927,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00021575841859541078,
+      "loss": 0.5233,
+      "step": 147040
+    },
+    {
+      "epoch": 7.303566107082547,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0002157186848117612,
+      "loss": 0.4678,
+      "step": 147050
+    },
+    {
+      "epoch": 7.3040627793781665,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00021567895102811164,
+      "loss": 0.4642,
+      "step": 147060
+    },
+    {
+      "epoch": 7.304559451673786,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0002156392172444621,
+      "loss": 0.5134,
+      "step": 147070
+    },
+    {
+      "epoch": 7.305056123969405,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00021559948346081256,
+      "loss": 0.5011,
+      "step": 147080
+    },
+    {
+      "epoch": 7.305552796265024,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00021555974967716303,
+      "loss": 0.5014,
+      "step": 147090
+    },
+    {
+      "epoch": 7.306049468560643,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00021552001589351347,
+      "loss": 0.4747,
+      "step": 147100
+    },
+    {
+      "epoch": 7.306546140856263,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00021548028210986392,
+      "loss": 0.4914,
+      "step": 147110
+    },
+    {
+      "epoch": 7.307042813151883,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002154405483262144,
+      "loss": 0.4917,
+      "step": 147120
+    },
+    {
+      "epoch": 7.307539485447502,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00021540081454256483,
+      "loss": 0.4962,
+      "step": 147130
+    },
+    {
+      "epoch": 7.308036157743121,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00021536108075891525,
+      "loss": 0.5183,
+      "step": 147140
+    },
+    {
+      "epoch": 7.30853283003874,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00021532134697526575,
+      "loss": 0.4937,
+      "step": 147150
+    },
+    {
+      "epoch": 7.30902950233436,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00021528161319161616,
+      "loss": 0.5082,
+      "step": 147160
+    },
+    {
+      "epoch": 7.309526174629979,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00021524187940796666,
+      "loss": 0.4975,
+      "step": 147170
+    },
+    {
+      "epoch": 7.310022846925598,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00021520214562431708,
+      "loss": 0.4993,
+      "step": 147180
+    },
+    {
+      "epoch": 7.310519519221218,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00021516241184066752,
+      "loss": 0.4736,
+      "step": 147190
+    },
+    {
+      "epoch": 7.3110161915168375,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.000215122678057018,
+      "loss": 0.4645,
+      "step": 147200
+    },
+    {
+      "epoch": 7.311512863812457,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00021508294427336844,
+      "loss": 0.4641,
+      "step": 147210
+    },
+    {
+      "epoch": 7.312009536108076,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002150432104897189,
+      "loss": 0.4738,
+      "step": 147220
+    },
+    {
+      "epoch": 7.312506208403695,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00021500347670606935,
+      "loss": 0.5394,
+      "step": 147230
+    },
+    {
+      "epoch": 7.313002880699314,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002149637429224198,
+      "loss": 0.4848,
+      "step": 147240
+    },
+    {
+      "epoch": 7.313499552994934,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00021492400913877027,
+      "loss": 0.4947,
+      "step": 147250
+    },
+    {
+      "epoch": 7.313996225290554,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002148842753551207,
+      "loss": 0.4941,
+      "step": 147260
+    },
+    {
+      "epoch": 7.314492897586173,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00021484454157147113,
+      "loss": 0.484,
+      "step": 147270
+    },
+    {
+      "epoch": 7.314989569881792,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002148048077878216,
+      "loss": 0.496,
+      "step": 147280
+    },
+    {
+      "epoch": 7.315486242177411,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00021476507400417205,
+      "loss": 0.4739,
+      "step": 147290
+    },
+    {
+      "epoch": 7.315982914473031,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00021472534022052252,
+      "loss": 0.5054,
+      "step": 147300
+    },
+    {
+      "epoch": 7.31647958676865,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00021468560643687296,
+      "loss": 0.4693,
+      "step": 147310
+    },
+    {
+      "epoch": 7.316976259064269,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002146458726532234,
+      "loss": 0.4808,
+      "step": 147320
+    },
+    {
+      "epoch": 7.317472931359889,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00021460613886957388,
+      "loss": 0.488,
+      "step": 147330
+    },
+    {
+      "epoch": 7.317969603655508,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00021456640508592432,
+      "loss": 0.5135,
+      "step": 147340
+    },
+    {
+      "epoch": 7.318466275951128,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00021452667130227474,
+      "loss": 0.4665,
+      "step": 147350
+    },
+    {
+      "epoch": 7.318962948246747,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00021448693751862524,
+      "loss": 0.4793,
+      "step": 147360
+    },
+    {
+      "epoch": 7.319459620542366,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00021444720373497566,
+      "loss": 0.5066,
+      "step": 147370
+    },
+    {
+      "epoch": 7.319956292837985,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00021440746995132615,
+      "loss": 0.4914,
+      "step": 147380
+    },
+    {
+      "epoch": 7.3204529651336046,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00021436773616767657,
+      "loss": 0.487,
+      "step": 147390
+    },
+    {
+      "epoch": 7.320949637429225,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00021432800238402702,
+      "loss": 0.501,
+      "step": 147400
+    },
+    {
+      "epoch": 7.321446309724844,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0002142882686003775,
+      "loss": 0.4784,
+      "step": 147410
+    },
+    {
+      "epoch": 7.321942982020463,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00021424853481672793,
+      "loss": 0.5204,
+      "step": 147420
+    },
+    {
+      "epoch": 7.322439654316082,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00021420880103307838,
+      "loss": 0.4816,
+      "step": 147430
+    },
+    {
+      "epoch": 7.322936326611702,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00021416906724942885,
+      "loss": 0.4802,
+      "step": 147440
+    },
+    {
+      "epoch": 7.323432998907321,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002141293334657793,
+      "loss": 0.4924,
+      "step": 147450
+    },
+    {
+      "epoch": 7.32392967120294,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00021408959968212976,
+      "loss": 0.4798,
+      "step": 147460
+    },
+    {
+      "epoch": 7.324426343498559,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002140498658984802,
+      "loss": 0.4849,
+      "step": 147470
+    },
+    {
+      "epoch": 7.324923015794179,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00021401013211483062,
+      "loss": 0.4835,
+      "step": 147480
+    },
+    {
+      "epoch": 7.325419688089799,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00021397039833118112,
+      "loss": 0.4794,
+      "step": 147490
+    },
+    {
+      "epoch": 7.325916360385418,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00021393066454753154,
+      "loss": 0.4891,
+      "step": 147500
+    },
+    {
+      "epoch": 7.326413032681037,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00021389093076388198,
+      "loss": 0.4946,
+      "step": 147510
+    },
+    {
+      "epoch": 7.326909704976656,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00021385119698023246,
+      "loss": 0.509,
+      "step": 147520
+    },
+    {
+      "epoch": 7.3274063772722755,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002138114631965829,
+      "loss": 0.4729,
+      "step": 147530
+    },
+    {
+      "epoch": 7.327903049567895,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00021377172941293337,
+      "loss": 0.4801,
+      "step": 147540
+    },
+    {
+      "epoch": 7.328399721863515,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00021373199562928381,
+      "loss": 0.5032,
+      "step": 147550
+    },
+    {
+      "epoch": 7.328896394159134,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00021369226184563423,
+      "loss": 0.4887,
+      "step": 147560
+    },
+    {
+      "epoch": 7.329393066454753,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00021365252806198473,
+      "loss": 0.4697,
+      "step": 147570
+    },
+    {
+      "epoch": 7.3298897387503725,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00021361279427833515,
+      "loss": 0.4832,
+      "step": 147580
+    },
+    {
+      "epoch": 7.330386411045992,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0002135730604946856,
+      "loss": 0.5004,
+      "step": 147590
+    },
+    {
+      "epoch": 7.330883083341611,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00021353332671103606,
+      "loss": 0.5154,
+      "step": 147600
+    },
+    {
+      "epoch": 7.33137975563723,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002134935929273865,
+      "loss": 0.4936,
+      "step": 147610
+    },
+    {
+      "epoch": 7.331876427932849,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00021345385914373698,
+      "loss": 0.4782,
+      "step": 147620
+    },
+    {
+      "epoch": 7.3323731002284696,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00021341412536008742,
+      "loss": 0.4783,
+      "step": 147630
+    },
+    {
+      "epoch": 7.332869772524089,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00021337439157643787,
+      "loss": 0.4588,
+      "step": 147640
+    },
+    {
+      "epoch": 7.333366444819708,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00021333465779278834,
+      "loss": 0.4956,
+      "step": 147650
+    },
+    {
+      "epoch": 7.333863117115327,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00021329492400913878,
+      "loss": 0.4903,
+      "step": 147660
+    },
+    {
+      "epoch": 7.3343597894109465,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00021325519022548925,
+      "loss": 0.4925,
+      "step": 147670
+    },
+    {
+      "epoch": 7.334856461706566,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002132154564418397,
+      "loss": 0.4826,
+      "step": 147680
+    },
+    {
+      "epoch": 7.335353134002185,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00021317572265819012,
+      "loss": 0.5,
+      "step": 147690
+    },
+    {
+      "epoch": 7.335849806297805,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00021313598887454061,
+      "loss": 0.4536,
+      "step": 147700
+    },
+    {
+      "epoch": 7.336346478593424,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00021309625509089103,
+      "loss": 0.4697,
+      "step": 147710
+    },
+    {
+      "epoch": 7.3368431508890435,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00021305652130724148,
+      "loss": 0.4588,
+      "step": 147720
+    },
+    {
+      "epoch": 7.337339823184663,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00021301678752359195,
+      "loss": 0.4895,
+      "step": 147730
+    },
+    {
+      "epoch": 7.337836495480282,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002129770537399424,
+      "loss": 0.4911,
+      "step": 147740
+    },
+    {
+      "epoch": 7.338333167775901,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00021293731995629286,
+      "loss": 0.4846,
+      "step": 147750
+    },
+    {
+      "epoch": 7.33882984007152,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0002128975861726433,
+      "loss": 0.4762,
+      "step": 147760
+    },
+    {
+      "epoch": 7.3393265123671405,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00021285785238899375,
+      "loss": 0.4769,
+      "step": 147770
+    },
+    {
+      "epoch": 7.33982318466276,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00021281811860534422,
+      "loss": 0.496,
+      "step": 147780
+    },
+    {
+      "epoch": 7.340319856958379,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00021277838482169467,
+      "loss": 0.4716,
+      "step": 147790
+    },
+    {
+      "epoch": 7.340816529253998,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00021273865103804508,
+      "loss": 0.4979,
+      "step": 147800
+    },
+    {
+      "epoch": 7.341313201549617,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00021269891725439558,
+      "loss": 0.4991,
+      "step": 147810
+    },
+    {
+      "epoch": 7.341809873845237,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.000212659183470746,
+      "loss": 0.5062,
+      "step": 147820
+    },
+    {
+      "epoch": 7.342306546140856,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00021261944968709647,
+      "loss": 0.5022,
+      "step": 147830
+    },
+    {
+      "epoch": 7.342803218436476,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00021257971590344691,
+      "loss": 0.5222,
+      "step": 147840
+    },
+    {
+      "epoch": 7.343299890732095,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00021253998211979736,
+      "loss": 0.4817,
+      "step": 147850
+    },
+    {
+      "epoch": 7.343796563027714,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00021250024833614783,
+      "loss": 0.4652,
+      "step": 147860
+    },
+    {
+      "epoch": 7.344293235323334,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00021246051455249827,
+      "loss": 0.4908,
+      "step": 147870
+    },
+    {
+      "epoch": 7.344789907618953,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0002124207807688487,
+      "loss": 0.4937,
+      "step": 147880
+    },
+    {
+      "epoch": 7.345286579914572,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0002123810469851992,
+      "loss": 0.4813,
+      "step": 147890
+    },
+    {
+      "epoch": 7.345783252210191,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002123413132015496,
+      "loss": 0.4756,
+      "step": 147900
+    },
+    {
+      "epoch": 7.3462799245058115,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002123015794179001,
+      "loss": 0.4978,
+      "step": 147910
+    },
+    {
+      "epoch": 7.346776596801431,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00021226184563425052,
+      "loss": 0.4986,
+      "step": 147920
+    },
+    {
+      "epoch": 7.34727326909705,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00021222211185060097,
+      "loss": 0.4819,
+      "step": 147930
+    },
+    {
+      "epoch": 7.347769941392669,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00021218237806695144,
+      "loss": 0.4643,
+      "step": 147940
+    },
+    {
+      "epoch": 7.348266613688288,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00021214264428330188,
+      "loss": 0.4582,
+      "step": 147950
+    },
+    {
+      "epoch": 7.348763285983908,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00021210291049965233,
+      "loss": 0.4971,
+      "step": 147960
+    },
+    {
+      "epoch": 7.349259958279527,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0002120631767160028,
+      "loss": 0.4929,
+      "step": 147970
+    },
+    {
+      "epoch": 7.349756630575147,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00021202344293235324,
+      "loss": 0.4662,
+      "step": 147980
+    },
+    {
+      "epoch": 7.350253302870766,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00021198370914870371,
+      "loss": 0.4468,
+      "step": 147990
+    },
+    {
+      "epoch": 7.350749975166385,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00021194397536505416,
+      "loss": 0.5046,
+      "step": 148000
+    },
+    {
+      "epoch": 7.351246647462005,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00021190424158140458,
+      "loss": 0.4765,
+      "step": 148010
+    },
+    {
+      "epoch": 7.351743319757624,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00021186450779775507,
+      "loss": 0.4758,
+      "step": 148020
+    },
+    {
+      "epoch": 7.352239992053243,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002118247740141055,
+      "loss": 0.4849,
+      "step": 148030
+    },
+    {
+      "epoch": 7.352736664348862,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00021178504023045594,
+      "loss": 0.4877,
+      "step": 148040
+    },
+    {
+      "epoch": 7.353233336644482,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002117453064468064,
+      "loss": 0.4915,
+      "step": 148050
+    },
+    {
+      "epoch": 7.353730008940102,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00021170557266315685,
+      "loss": 0.5046,
+      "step": 148060
+    },
+    {
+      "epoch": 7.354226681235721,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00021166583887950732,
+      "loss": 0.4907,
+      "step": 148070
+    },
+    {
+      "epoch": 7.35472335353134,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00021162610509585777,
+      "loss": 0.4744,
+      "step": 148080
+    },
+    {
+      "epoch": 7.355220025826959,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002115863713122082,
+      "loss": 0.4648,
+      "step": 148090
+    },
+    {
+      "epoch": 7.355716698122579,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00021154663752855868,
+      "loss": 0.5027,
+      "step": 148100
+    },
+    {
+      "epoch": 7.356213370418198,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0002115069037449091,
+      "loss": 0.4799,
+      "step": 148110
+    },
+    {
+      "epoch": 7.356710042713817,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002114671699612596,
+      "loss": 0.4903,
+      "step": 148120
+    },
+    {
+      "epoch": 7.357206715009437,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00021142743617761002,
+      "loss": 0.4935,
+      "step": 148130
+    },
+    {
+      "epoch": 7.357703387305056,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00021138770239396046,
+      "loss": 0.4986,
+      "step": 148140
+    },
+    {
+      "epoch": 7.358200059600676,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00021134796861031093,
+      "loss": 0.4964,
+      "step": 148150
+    },
+    {
+      "epoch": 7.358696731896295,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00021130823482666137,
+      "loss": 0.4879,
+      "step": 148160
+    },
+    {
+      "epoch": 7.359193404191914,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00021126850104301182,
+      "loss": 0.5062,
+      "step": 148170
+    },
+    {
+      "epoch": 7.359690076487533,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0002112287672593623,
+      "loss": 0.5151,
+      "step": 148180
+    },
+    {
+      "epoch": 7.3601867487831525,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00021118903347571273,
+      "loss": 0.5111,
+      "step": 148190
+    },
+    {
+      "epoch": 7.360683421078773,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002111492996920632,
+      "loss": 0.5055,
+      "step": 148200
+    },
+    {
+      "epoch": 7.361180093374392,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00021110956590841365,
+      "loss": 0.4919,
+      "step": 148210
+    },
+    {
+      "epoch": 7.361676765670011,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.00021106983212476407,
+      "loss": 0.4405,
+      "step": 148220
+    },
+    {
+      "epoch": 7.36217343796563,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00021103009834111457,
+      "loss": 0.4251,
+      "step": 148230
+    },
+    {
+      "epoch": 7.3626701102612495,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00021099036455746498,
+      "loss": 0.5104,
+      "step": 148240
+    },
+    {
+      "epoch": 7.363166782556869,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00021095063077381543,
+      "loss": 0.4527,
+      "step": 148250
+    },
+    {
+      "epoch": 7.363663454852488,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002109108969901659,
+      "loss": 0.4996,
+      "step": 148260
+    },
+    {
+      "epoch": 7.364160127148108,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00021087116320651634,
+      "loss": 0.4782,
+      "step": 148270
+    },
+    {
+      "epoch": 7.364656799443727,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00021083142942286681,
+      "loss": 0.5066,
+      "step": 148280
+    },
+    {
+      "epoch": 7.3651534717393465,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00021079169563921726,
+      "loss": 0.4703,
+      "step": 148290
+    },
+    {
+      "epoch": 7.365650144034966,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002107519618555677,
+      "loss": 0.4993,
+      "step": 148300
+    },
+    {
+      "epoch": 7.366146816330585,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00021071222807191817,
+      "loss": 0.4699,
+      "step": 148310
+    },
+    {
+      "epoch": 7.366643488626204,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00021067249428826862,
+      "loss": 0.4557,
+      "step": 148320
+    },
+    {
+      "epoch": 7.367140160921823,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00021063276050461904,
+      "loss": 0.4764,
+      "step": 148330
+    },
+    {
+      "epoch": 7.367636833217443,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00021059302672096953,
+      "loss": 0.471,
+      "step": 148340
+    },
+    {
+      "epoch": 7.368133505513063,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00021055329293731995,
+      "loss": 0.5214,
+      "step": 148350
+    },
+    {
+      "epoch": 7.368630177808682,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00021051355915367045,
+      "loss": 0.4973,
+      "step": 148360
+    },
+    {
+      "epoch": 7.369126850104301,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00021047382537002087,
+      "loss": 0.4635,
+      "step": 148370
+    },
+    {
+      "epoch": 7.3696235223999205,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0002104340915863713,
+      "loss": 0.4943,
+      "step": 148380
+    },
+    {
+      "epoch": 7.37012019469554,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00021039435780272178,
+      "loss": 0.5044,
+      "step": 148390
+    },
+    {
+      "epoch": 7.370616866991159,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00021035462401907223,
+      "loss": 0.492,
+      "step": 148400
+    },
+    {
+      "epoch": 7.371113539286778,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00021031489023542264,
+      "loss": 0.4722,
+      "step": 148410
+    },
+    {
+      "epoch": 7.371610211582398,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00021027515645177314,
+      "loss": 0.4814,
+      "step": 148420
+    },
+    {
+      "epoch": 7.3721068838780175,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00021023542266812356,
+      "loss": 0.4697,
+      "step": 148430
+    },
+    {
+      "epoch": 7.372603556173637,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00021019568888447406,
+      "loss": 0.4714,
+      "step": 148440
+    },
+    {
+      "epoch": 7.373100228469256,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00021015595510082448,
+      "loss": 0.4722,
+      "step": 148450
+    },
+    {
+      "epoch": 7.373596900764875,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00021011622131717492,
+      "loss": 0.4903,
+      "step": 148460
+    },
+    {
+      "epoch": 7.374093573060494,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002100764875335254,
+      "loss": 0.4621,
+      "step": 148470
+    },
+    {
+      "epoch": 7.374590245356114,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00021003675374987583,
+      "loss": 0.4817,
+      "step": 148480
+    },
+    {
+      "epoch": 7.375086917651734,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0002099970199662263,
+      "loss": 0.5023,
+      "step": 148490
+    },
+    {
+      "epoch": 7.375583589947353,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00020995728618257675,
+      "loss": 0.5172,
+      "step": 148500
+    },
+    {
+      "epoch": 7.376080262242972,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0002099175523989272,
+      "loss": 0.4956,
+      "step": 148510
+    },
+    {
+      "epoch": 7.376576934538591,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00020987781861527767,
+      "loss": 0.5087,
+      "step": 148520
+    },
+    {
+      "epoch": 7.377073606834211,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002098380848316281,
+      "loss": 0.5029,
+      "step": 148530
+    },
+    {
+      "epoch": 7.37757027912983,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00020979835104797853,
+      "loss": 0.4689,
+      "step": 148540
+    },
+    {
+      "epoch": 7.378066951425449,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00020975861726432903,
+      "loss": 0.495,
+      "step": 148550
+    },
+    {
+      "epoch": 7.378563623721069,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00020971888348067944,
+      "loss": 0.4624,
+      "step": 148560
+    },
+    {
+      "epoch": 7.3790602960166884,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00020967914969702994,
+      "loss": 0.4616,
+      "step": 148570
+    },
+    {
+      "epoch": 7.379556968312308,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00020963941591338036,
+      "loss": 0.4881,
+      "step": 148580
+    },
+    {
+      "epoch": 7.380053640607927,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0002095996821297308,
+      "loss": 0.4747,
+      "step": 148590
+    },
+    {
+      "epoch": 7.380550312903546,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00020955994834608127,
+      "loss": 0.5043,
+      "step": 148600
+    },
+    {
+      "epoch": 7.381046985199165,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00020952021456243172,
+      "loss": 0.4839,
+      "step": 148610
+    },
+    {
+      "epoch": 7.381543657494785,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00020948048077878216,
+      "loss": 0.4851,
+      "step": 148620
+    },
+    {
+      "epoch": 7.382040329790405,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00020944074699513263,
+      "loss": 0.478,
+      "step": 148630
+    },
+    {
+      "epoch": 7.382537002086024,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00020940101321148308,
+      "loss": 0.4842,
+      "step": 148640
+    },
+    {
+      "epoch": 7.383033674381643,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00020936127942783355,
+      "loss": 0.4903,
+      "step": 148650
+    },
+    {
+      "epoch": 7.383530346677262,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.000209321545644184,
+      "loss": 0.5152,
+      "step": 148660
+    },
+    {
+      "epoch": 7.384027018972882,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002092818118605344,
+      "loss": 0.4867,
+      "step": 148670
+    },
+    {
+      "epoch": 7.384523691268501,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00020924207807688488,
+      "loss": 0.4878,
+      "step": 148680
+    },
+    {
+      "epoch": 7.38502036356412,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00020920234429323533,
+      "loss": 0.5052,
+      "step": 148690
+    },
+    {
+      "epoch": 7.38551703585974,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00020916261050958577,
+      "loss": 0.4869,
+      "step": 148700
+    },
+    {
+      "epoch": 7.386013708155359,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00020912287672593624,
+      "loss": 0.4599,
+      "step": 148710
+    },
+    {
+      "epoch": 7.386510380450979,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002090831429422867,
+      "loss": 0.4942,
+      "step": 148720
+    },
+    {
+      "epoch": 7.387007052746598,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00020904340915863716,
+      "loss": 0.5085,
+      "step": 148730
+    },
+    {
+      "epoch": 7.387503725042217,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002090036753749876,
+      "loss": 0.5101,
+      "step": 148740
+    },
+    {
+      "epoch": 7.388000397337836,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00020896394159133802,
+      "loss": 0.4933,
+      "step": 148750
+    },
+    {
+      "epoch": 7.3884970696334555,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00020892420780768852,
+      "loss": 0.4913,
+      "step": 148760
+    },
+    {
+      "epoch": 7.388993741929076,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.00020888447402403894,
+      "loss": 0.4553,
+      "step": 148770
+    },
+    {
+      "epoch": 7.389490414224695,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00020884474024038938,
+      "loss": 0.4928,
+      "step": 148780
+    },
+    {
+      "epoch": 7.389987086520314,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00020880500645673985,
+      "loss": 0.4955,
+      "step": 148790
+    },
+    {
+      "epoch": 7.390483758815933,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002087652726730903,
+      "loss": 0.4831,
+      "step": 148800
+    },
+    {
+      "epoch": 7.390980431111553,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00020872553888944077,
+      "loss": 0.4747,
+      "step": 148810
+    },
+    {
+      "epoch": 7.391477103407172,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0002086858051057912,
+      "loss": 0.4693,
+      "step": 148820
+    },
+    {
+      "epoch": 7.391973775702791,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00020864607132214165,
+      "loss": 0.4614,
+      "step": 148830
+    },
+    {
+      "epoch": 7.39247044799841,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00020860633753849213,
+      "loss": 0.4809,
+      "step": 148840
+    },
+    {
+      "epoch": 7.39296712029403,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00020856660375484257,
+      "loss": 0.4973,
+      "step": 148850
+    },
+    {
+      "epoch": 7.39346379258965,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.000208526869971193,
+      "loss": 0.4956,
+      "step": 148860
+    },
+    {
+      "epoch": 7.393960464885269,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00020848713618754349,
+      "loss": 0.4837,
+      "step": 148870
+    },
+    {
+      "epoch": 7.394457137180888,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0002084474024038939,
+      "loss": 0.4962,
+      "step": 148880
+    },
+    {
+      "epoch": 7.394953809476507,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0002084076686202444,
+      "loss": 0.4933,
+      "step": 148890
+    },
+    {
+      "epoch": 7.3954504817721265,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00020836793483659482,
+      "loss": 0.4801,
+      "step": 148900
+    },
+    {
+      "epoch": 7.395947154067746,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00020832820105294526,
+      "loss": 0.5075,
+      "step": 148910
+    },
+    {
+      "epoch": 7.396443826363366,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00020828846726929573,
+      "loss": 0.5015,
+      "step": 148920
+    },
+    {
+      "epoch": 7.396940498658985,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00020824873348564618,
+      "loss": 0.4751,
+      "step": 148930
+    },
+    {
+      "epoch": 7.397437170954604,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00020820899970199665,
+      "loss": 0.4797,
+      "step": 148940
+    },
+    {
+      "epoch": 7.3979338432502235,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002081692659183471,
+      "loss": 0.4828,
+      "step": 148950
+    },
+    {
+      "epoch": 7.398430515545843,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002081295321346975,
+      "loss": 0.5123,
+      "step": 148960
+    },
+    {
+      "epoch": 7.398927187841462,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.000208089798351048,
+      "loss": 0.4848,
+      "step": 148970
+    },
+    {
+      "epoch": 7.399423860137081,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00020805006456739843,
+      "loss": 0.4856,
+      "step": 148980
+    },
+    {
+      "epoch": 7.399920532432701,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00020801033078374887,
+      "loss": 0.4928,
+      "step": 148990
+    },
+    {
+      "epoch": 7.4004172047283205,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00020797059700009934,
+      "loss": 0.5128,
+      "step": 149000
+    },
+    {
+      "epoch": 7.40091387702394,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002079308632164498,
+      "loss": 0.5278,
+      "step": 149010
+    },
+    {
+      "epoch": 7.401410549319559,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00020789112943280026,
+      "loss": 0.4889,
+      "step": 149020
+    },
+    {
+      "epoch": 7.401907221615178,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002078513956491507,
+      "loss": 0.4796,
+      "step": 149030
+    },
+    {
+      "epoch": 7.4024038939107974,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00020781166186550115,
+      "loss": 0.4768,
+      "step": 149040
+    },
+    {
+      "epoch": 7.402900566206417,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00020777192808185162,
+      "loss": 0.4833,
+      "step": 149050
+    },
+    {
+      "epoch": 7.403397238502036,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00020773219429820206,
+      "loss": 0.4826,
+      "step": 149060
+    },
+    {
+      "epoch": 7.403893910797656,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00020769246051455248,
+      "loss": 0.5,
+      "step": 149070
+    },
+    {
+      "epoch": 7.404390583093275,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00020765272673090298,
+      "loss": 0.479,
+      "step": 149080
+    },
+    {
+      "epoch": 7.4048872553888945,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002076129929472534,
+      "loss": 0.4818,
+      "step": 149090
+    },
+    {
+      "epoch": 7.405383927684514,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0002075732591636039,
+      "loss": 0.498,
+      "step": 149100
+    },
+    {
+      "epoch": 7.405880599980133,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002075335253799543,
+      "loss": 0.4568,
+      "step": 149110
+    },
+    {
+      "epoch": 7.406377272275752,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00020749379159630475,
+      "loss": 0.483,
+      "step": 149120
+    },
+    {
+      "epoch": 7.406873944571371,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00020745405781265523,
+      "loss": 0.4718,
+      "step": 149130
+    },
+    {
+      "epoch": 7.4073706168669915,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00020741432402900567,
+      "loss": 0.5066,
+      "step": 149140
+    },
+    {
+      "epoch": 7.407867289162611,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00020737459024535611,
+      "loss": 0.4941,
+      "step": 149150
+    },
+    {
+      "epoch": 7.40836396145823,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00020733485646170659,
+      "loss": 0.4757,
+      "step": 149160
+    },
+    {
+      "epoch": 7.408860633753849,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00020729512267805703,
+      "loss": 0.5042,
+      "step": 149170
+    },
+    {
+      "epoch": 7.409357306049468,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002072553888944075,
+      "loss": 0.5143,
+      "step": 149180
+    },
+    {
+      "epoch": 7.409853978345088,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00020721565511075795,
+      "loss": 0.4808,
+      "step": 149190
+    },
+    {
+      "epoch": 7.410350650640707,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00020717592132710836,
+      "loss": 0.5003,
+      "step": 149200
+    },
+    {
+      "epoch": 7.410847322936327,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00020713618754345886,
+      "loss": 0.4878,
+      "step": 149210
+    },
+    {
+      "epoch": 7.411343995231946,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00020709645375980928,
+      "loss": 0.504,
+      "step": 149220
+    },
+    {
+      "epoch": 7.411840667527565,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00020705671997615972,
+      "loss": 0.481,
+      "step": 149230
+    },
+    {
+      "epoch": 7.412337339823185,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0002070169861925102,
+      "loss": 0.4732,
+      "step": 149240
+    },
+    {
+      "epoch": 7.412834012118804,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00020697725240886064,
+      "loss": 0.5001,
+      "step": 149250
+    },
+    {
+      "epoch": 7.413330684414423,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0002069375186252111,
+      "loss": 0.4797,
+      "step": 149260
+    },
+    {
+      "epoch": 7.413827356710042,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00020689778484156155,
+      "loss": 0.5091,
+      "step": 149270
+    },
+    {
+      "epoch": 7.4143240290056625,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00020685805105791197,
+      "loss": 0.4502,
+      "step": 149280
+    },
+    {
+      "epoch": 7.414820701301282,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00020681831727426247,
+      "loss": 0.4787,
+      "step": 149290
+    },
+    {
+      "epoch": 7.415317373596901,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0002067785834906129,
+      "loss": 0.5072,
+      "step": 149300
+    },
+    {
+      "epoch": 7.41581404589252,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00020673884970696333,
+      "loss": 0.5042,
+      "step": 149310
+    },
+    {
+      "epoch": 7.416310718188139,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0002066991159233138,
+      "loss": 0.4981,
+      "step": 149320
+    },
+    {
+      "epoch": 7.416807390483759,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00020665938213966425,
+      "loss": 0.4738,
+      "step": 149330
+    },
+    {
+      "epoch": 7.417304062779378,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00020661964835601472,
+      "loss": 0.4689,
+      "step": 149340
+    },
+    {
+      "epoch": 7.417800735074998,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00020657991457236516,
+      "loss": 0.4857,
+      "step": 149350
+    },
+    {
+      "epoch": 7.418297407370617,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002065401807887156,
+      "loss": 0.5077,
+      "step": 149360
+    },
+    {
+      "epoch": 7.418794079666236,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00020650044700506608,
+      "loss": 0.4643,
+      "step": 149370
+    },
+    {
+      "epoch": 7.419290751961856,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00020646071322141652,
+      "loss": 0.4682,
+      "step": 149380
+    },
+    {
+      "epoch": 7.419787424257475,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.000206420979437767,
+      "loss": 0.4662,
+      "step": 149390
+    },
+    {
+      "epoch": 7.420284096553094,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00020638124565411744,
+      "loss": 0.4841,
+      "step": 149400
+    },
+    {
+      "epoch": 7.420780768848713,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00020634151187046785,
+      "loss": 0.4925,
+      "step": 149410
+    },
+    {
+      "epoch": 7.421277441144333,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00020630177808681835,
+      "loss": 0.4395,
+      "step": 149420
+    },
+    {
+      "epoch": 7.421774113439953,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00020626204430316877,
+      "loss": 0.4531,
+      "step": 149430
+    },
+    {
+      "epoch": 7.422270785735572,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00020622231051951921,
+      "loss": 0.4508,
+      "step": 149440
+    },
+    {
+      "epoch": 7.422767458031191,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00020618257673586969,
+      "loss": 0.4889,
+      "step": 149450
+    },
+    {
+      "epoch": 7.42326413032681,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00020614284295222013,
+      "loss": 0.5129,
+      "step": 149460
+    },
+    {
+      "epoch": 7.4237608026224295,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0002061031091685706,
+      "loss": 0.5116,
+      "step": 149470
+    },
+    {
+      "epoch": 7.424257474918049,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00020606337538492105,
+      "loss": 0.4896,
+      "step": 149480
+    },
+    {
+      "epoch": 7.424754147213669,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0002060236416012715,
+      "loss": 0.5167,
+      "step": 149490
+    },
+    {
+      "epoch": 7.425250819509288,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00020598390781762196,
+      "loss": 0.4773,
+      "step": 149500
+    },
+    {
+      "epoch": 7.425747491804907,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002059441740339724,
+      "loss": 0.4978,
+      "step": 149510
+    },
+    {
+      "epoch": 7.426244164100527,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00020590444025032282,
+      "loss": 0.4872,
+      "step": 149520
+    },
+    {
+      "epoch": 7.426740836396146,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0002058647064666733,
+      "loss": 0.5036,
+      "step": 149530
+    },
+    {
+      "epoch": 7.427237508691765,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00020582497268302374,
+      "loss": 0.4777,
+      "step": 149540
+    },
+    {
+      "epoch": 7.427734180987384,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002057852388993742,
+      "loss": 0.4929,
+      "step": 149550
+    },
+    {
+      "epoch": 7.4282308532830035,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00020574550511572465,
+      "loss": 0.503,
+      "step": 149560
+    },
+    {
+      "epoch": 7.428727525578624,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002057057713320751,
+      "loss": 0.4705,
+      "step": 149570
+    },
+    {
+      "epoch": 7.429224197874243,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00020566603754842557,
+      "loss": 0.4755,
+      "step": 149580
+    },
+    {
+      "epoch": 7.429720870169862,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00020562630376477601,
+      "loss": 0.4845,
+      "step": 149590
+    },
+    {
+      "epoch": 7.430217542465481,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00020558656998112643,
+      "loss": 0.5069,
+      "step": 149600
+    },
+    {
+      "epoch": 7.4307142147611005,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.00020554683619747693,
+      "loss": 0.4874,
+      "step": 149610
+    },
+    {
+      "epoch": 7.43121088705672,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00020550710241382735,
+      "loss": 0.482,
+      "step": 149620
+    },
+    {
+      "epoch": 7.431707559352339,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00020546736863017785,
+      "loss": 0.4415,
+      "step": 149630
+    },
+    {
+      "epoch": 7.432204231647959,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00020542763484652826,
+      "loss": 0.4703,
+      "step": 149640
+    },
+    {
+      "epoch": 7.432700903943578,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002053879010628787,
+      "loss": 0.4841,
+      "step": 149650
+    },
+    {
+      "epoch": 7.4331975762391975,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00020534816727922918,
+      "loss": 0.5322,
+      "step": 149660
+    },
+    {
+      "epoch": 7.433694248534817,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00020530843349557962,
+      "loss": 0.4818,
+      "step": 149670
+    },
+    {
+      "epoch": 7.434190920830436,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00020526869971193007,
+      "loss": 0.4964,
+      "step": 149680
+    },
+    {
+      "epoch": 7.434687593126055,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00020522896592828054,
+      "loss": 0.4843,
+      "step": 149690
+    },
+    {
+      "epoch": 7.435184265421674,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00020518923214463098,
+      "loss": 0.4871,
+      "step": 149700
+    },
+    {
+      "epoch": 7.435680937717294,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00020514949836098145,
+      "loss": 0.5375,
+      "step": 149710
+    },
+    {
+      "epoch": 7.436177610012914,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002051097645773319,
+      "loss": 0.4897,
+      "step": 149720
+    },
+    {
+      "epoch": 7.436674282308533,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00020507003079368231,
+      "loss": 0.4603,
+      "step": 149730
+    },
+    {
+      "epoch": 7.437170954604152,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0002050302970100328,
+      "loss": 0.473,
+      "step": 149740
+    },
+    {
+      "epoch": 7.4376676268997715,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00020499056322638323,
+      "loss": 0.5023,
+      "step": 149750
+    },
+    {
+      "epoch": 7.438164299195391,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00020495082944273373,
+      "loss": 0.475,
+      "step": 149760
+    },
+    {
+      "epoch": 7.43866097149101,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00020491109565908415,
+      "loss": 0.4797,
+      "step": 149770
+    },
+    {
+      "epoch": 7.439157643786629,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002048713618754346,
+      "loss": 0.4915,
+      "step": 149780
+    },
+    {
+      "epoch": 7.439654316082249,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00020483162809178506,
+      "loss": 0.4932,
+      "step": 149790
+    },
+    {
+      "epoch": 7.4401509883778685,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0002047918943081355,
+      "loss": 0.4973,
+      "step": 149800
+    },
+    {
+      "epoch": 7.440647660673488,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00020475216052448592,
+      "loss": 0.4873,
+      "step": 149810
+    },
+    {
+      "epoch": 7.441144332969107,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00020471242674083642,
+      "loss": 0.474,
+      "step": 149820
+    },
+    {
+      "epoch": 7.441641005264726,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00020467269295718684,
+      "loss": 0.4897,
+      "step": 149830
+    },
+    {
+      "epoch": 7.442137677560345,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00020463295917353734,
+      "loss": 0.4911,
+      "step": 149840
+    },
+    {
+      "epoch": 7.442634349855965,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00020459322538988775,
+      "loss": 0.5272,
+      "step": 149850
+    },
+    {
+      "epoch": 7.443131022151585,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0002045534916062382,
+      "loss": 0.511,
+      "step": 149860
+    },
+    {
+      "epoch": 7.443627694447204,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00020451375782258867,
+      "loss": 0.5126,
+      "step": 149870
+    },
+    {
+      "epoch": 7.444124366742823,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00020447402403893911,
+      "loss": 0.4711,
+      "step": 149880
+    },
+    {
+      "epoch": 7.444621039038442,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00020443429025528956,
+      "loss": 0.4834,
+      "step": 149890
+    },
+    {
+      "epoch": 7.445117711334062,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00020439455647164003,
+      "loss": 0.5153,
+      "step": 149900
+    },
+    {
+      "epoch": 7.445614383629681,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00020435482268799047,
+      "loss": 0.4782,
+      "step": 149910
+    },
+    {
+      "epoch": 7.4461110559253,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00020431508890434095,
+      "loss": 0.468,
+      "step": 149920
+    },
+    {
+      "epoch": 7.44660772822092,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0002042753551206914,
+      "loss": 0.4964,
+      "step": 149930
+    },
+    {
+      "epoch": 7.447104400516539,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0002042356213370418,
+      "loss": 0.5064,
+      "step": 149940
+    },
+    {
+      "epoch": 7.447601072812159,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0002041958875533923,
+      "loss": 0.4847,
+      "step": 149950
+    },
+    {
+      "epoch": 7.448097745107778,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00020415615376974272,
+      "loss": 0.4819,
+      "step": 149960
+    },
+    {
+      "epoch": 7.448594417403397,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00020411641998609317,
+      "loss": 0.4937,
+      "step": 149970
+    },
+    {
+      "epoch": 7.449091089699016,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00020407668620244364,
+      "loss": 0.4924,
+      "step": 149980
+    },
+    {
+      "epoch": 7.449587761994636,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00020403695241879408,
+      "loss": 0.4743,
+      "step": 149990
+    },
+    {
+      "epoch": 7.450084434290256,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00020399721863514455,
+      "loss": 0.4696,
+      "step": 150000
+    },
+    {
+      "epoch": 7.450581106585875,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.000203957484851495,
+      "loss": 0.5106,
+      "step": 150010
+    },
+    {
+      "epoch": 7.451077778881494,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00020391775106784544,
+      "loss": 0.4905,
+      "step": 150020
+    },
+    {
+      "epoch": 7.451574451177113,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0002038780172841959,
+      "loss": 0.496,
+      "step": 150030
+    },
+    {
+      "epoch": 7.452071123472733,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00020383828350054636,
+      "loss": 0.4744,
+      "step": 150040
+    },
+    {
+      "epoch": 7.452567795768352,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00020379854971689677,
+      "loss": 0.4868,
+      "step": 150050
+    },
+    {
+      "epoch": 7.453064468063971,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00020375881593324727,
+      "loss": 0.4942,
+      "step": 150060
+    },
+    {
+      "epoch": 7.453561140359591,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002037190821495977,
+      "loss": 0.497,
+      "step": 150070
+    },
+    {
+      "epoch": 7.45405781265521,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00020367934836594816,
+      "loss": 0.4981,
+      "step": 150080
+    },
+    {
+      "epoch": 7.45455448495083,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002036396145822986,
+      "loss": 0.4988,
+      "step": 150090
+    },
+    {
+      "epoch": 7.455051157246449,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00020359988079864905,
+      "loss": 0.479,
+      "step": 150100
+    },
+    {
+      "epoch": 7.455547829542068,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00020356014701499952,
+      "loss": 0.4834,
+      "step": 150110
+    },
+    {
+      "epoch": 7.456044501837687,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00020352041323134997,
+      "loss": 0.4881,
+      "step": 150120
+    },
+    {
+      "epoch": 7.4565411741333065,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00020348067944770038,
+      "loss": 0.4548,
+      "step": 150130
+    },
+    {
+      "epoch": 7.457037846428927,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00020344094566405088,
+      "loss": 0.4853,
+      "step": 150140
+    },
+    {
+      "epoch": 7.457534518724546,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0002034012118804013,
+      "loss": 0.5291,
+      "step": 150150
+    },
+    {
+      "epoch": 7.458031191020165,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0002033614780967518,
+      "loss": 0.4977,
+      "step": 150160
+    },
+    {
+      "epoch": 7.458527863315784,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00020332174431310221,
+      "loss": 0.482,
+      "step": 150170
+    },
+    {
+      "epoch": 7.4590245356114036,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00020328201052945266,
+      "loss": 0.463,
+      "step": 150180
+    },
+    {
+      "epoch": 7.459521207907023,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00020324227674580313,
+      "loss": 0.504,
+      "step": 150190
+    },
+    {
+      "epoch": 7.460017880202642,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00020320254296215357,
+      "loss": 0.5334,
+      "step": 150200
+    },
+    {
+      "epoch": 7.460514552498261,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00020316280917850405,
+      "loss": 0.4615,
+      "step": 150210
+    },
+    {
+      "epoch": 7.461011224793881,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002031230753948545,
+      "loss": 0.4766,
+      "step": 150220
+    },
+    {
+      "epoch": 7.461507897089501,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.00020308334161120493,
+      "loss": 0.4841,
+      "step": 150230
+    },
+    {
+      "epoch": 7.46200456938512,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0002030436078275554,
+      "loss": 0.4853,
+      "step": 150240
+    },
+    {
+      "epoch": 7.462501241680739,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00020300387404390585,
+      "loss": 0.4894,
+      "step": 150250
+    },
+    {
+      "epoch": 7.462997913976358,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00020296414026025627,
+      "loss": 0.4651,
+      "step": 150260
+    },
+    {
+      "epoch": 7.4634945862719775,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00020292440647660677,
+      "loss": 0.4855,
+      "step": 150270
+    },
+    {
+      "epoch": 7.463991258567597,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00020288467269295718,
+      "loss": 0.4818,
+      "step": 150280
+    },
+    {
+      "epoch": 7.464487930863217,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00020284493890930768,
+      "loss": 0.4988,
+      "step": 150290
+    },
+    {
+      "epoch": 7.464984603158836,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002028052051256581,
+      "loss": 0.487,
+      "step": 150300
+    },
+    {
+      "epoch": 7.465481275454455,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00020276547134200854,
+      "loss": 0.4859,
+      "step": 150310
+    },
+    {
+      "epoch": 7.4659779477500745,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00020272573755835901,
+      "loss": 0.4853,
+      "step": 150320
+    },
+    {
+      "epoch": 7.466474620045694,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00020268600377470946,
+      "loss": 0.4654,
+      "step": 150330
+    },
+    {
+      "epoch": 7.466971292341313,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0002026462699910599,
+      "loss": 0.4836,
+      "step": 150340
+    },
+    {
+      "epoch": 7.467467964636932,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00020260653620741037,
+      "loss": 0.4624,
+      "step": 150350
+    },
+    {
+      "epoch": 7.467964636932552,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00020256680242376082,
+      "loss": 0.4759,
+      "step": 150360
+    },
+    {
+      "epoch": 7.4684613092281715,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002025270686401113,
+      "loss": 0.4909,
+      "step": 150370
+    },
+    {
+      "epoch": 7.468957981523791,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0002024873348564617,
+      "loss": 0.4793,
+      "step": 150380
+    },
+    {
+      "epoch": 7.46945465381941,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00020244760107281215,
+      "loss": 0.5134,
+      "step": 150390
+    },
+    {
+      "epoch": 7.469951326115029,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00020240786728916262,
+      "loss": 0.448,
+      "step": 150400
+    },
+    {
+      "epoch": 7.470447998410648,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00020236813350551307,
+      "loss": 0.4737,
+      "step": 150410
+    },
+    {
+      "epoch": 7.470944670706268,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0002023283997218635,
+      "loss": 0.4889,
+      "step": 150420
+    },
+    {
+      "epoch": 7.471441343001887,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00020228866593821398,
+      "loss": 0.4734,
+      "step": 150430
+    },
+    {
+      "epoch": 7.471938015297507,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00020224893215456443,
+      "loss": 0.457,
+      "step": 150440
+    },
+    {
+      "epoch": 7.472434687593126,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002022091983709149,
+      "loss": 0.5119,
+      "step": 150450
+    },
+    {
+      "epoch": 7.4729313598887455,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00020216946458726534,
+      "loss": 0.4664,
+      "step": 150460
+    },
+    {
+      "epoch": 7.473428032184365,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00020212973080361576,
+      "loss": 0.4786,
+      "step": 150470
+    },
+    {
+      "epoch": 7.473924704479984,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00020208999701996626,
+      "loss": 0.4814,
+      "step": 150480
+    },
+    {
+      "epoch": 7.474421376775603,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00020205026323631667,
+      "loss": 0.4903,
+      "step": 150490
+    },
+    {
+      "epoch": 7.474918049071222,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00020201052945266712,
+      "loss": 0.4863,
+      "step": 150500
+    },
+    {
+      "epoch": 7.4754147213668425,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002019707956690176,
+      "loss": 0.5055,
+      "step": 150510
+    },
+    {
+      "epoch": 7.475911393662462,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00020193106188536803,
+      "loss": 0.5019,
+      "step": 150520
+    },
+    {
+      "epoch": 7.476408065958081,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0002018913281017185,
+      "loss": 0.4812,
+      "step": 150530
+    },
+    {
+      "epoch": 7.4769047382537,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00020185159431806895,
+      "loss": 0.5075,
+      "step": 150540
+    },
+    {
+      "epoch": 7.477401410549319,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0002018118605344194,
+      "loss": 0.4877,
+      "step": 150550
+    },
+    {
+      "epoch": 7.477898082844939,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00020177212675076987,
+      "loss": 0.4877,
+      "step": 150560
+    },
+    {
+      "epoch": 7.478394755140558,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0002017323929671203,
+      "loss": 0.4917,
+      "step": 150570
+    },
+    {
+      "epoch": 7.478891427436178,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00020169265918347073,
+      "loss": 0.4761,
+      "step": 150580
+    },
+    {
+      "epoch": 7.479388099731797,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00020165292539982123,
+      "loss": 0.5167,
+      "step": 150590
+    },
+    {
+      "epoch": 7.479884772027416,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00020161319161617164,
+      "loss": 0.4985,
+      "step": 150600
+    },
+    {
+      "epoch": 7.480381444323036,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00020157345783252214,
+      "loss": 0.4823,
+      "step": 150610
+    },
+    {
+      "epoch": 7.480878116618655,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00020153372404887256,
+      "loss": 0.4774,
+      "step": 150620
+    },
+    {
+      "epoch": 7.481374788914274,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.000201493990265223,
+      "loss": 0.5035,
+      "step": 150630
+    },
+    {
+      "epoch": 7.481871461209893,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00020145425648157347,
+      "loss": 0.4775,
+      "step": 150640
+    },
+    {
+      "epoch": 7.482368133505513,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00020141452269792392,
+      "loss": 0.4709,
+      "step": 150650
+    },
+    {
+      "epoch": 7.482864805801133,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0002013747889142744,
+      "loss": 0.5085,
+      "step": 150660
+    },
+    {
+      "epoch": 7.483361478096752,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00020133505513062483,
+      "loss": 0.4913,
+      "step": 150670
+    },
+    {
+      "epoch": 7.483858150392371,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00020129532134697525,
+      "loss": 0.5026,
+      "step": 150680
+    },
+    {
+      "epoch": 7.48435482268799,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00020125558756332575,
+      "loss": 0.5076,
+      "step": 150690
+    },
+    {
+      "epoch": 7.48485149498361,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00020121585377967617,
+      "loss": 0.4643,
+      "step": 150700
+    },
+    {
+      "epoch": 7.485348167279229,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0002011761199960266,
+      "loss": 0.4904,
+      "step": 150710
+    },
+    {
+      "epoch": 7.485844839574849,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00020113638621237708,
+      "loss": 0.4684,
+      "step": 150720
+    },
+    {
+      "epoch": 7.486341511870468,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00020109665242872753,
+      "loss": 0.4995,
+      "step": 150730
+    },
+    {
+      "epoch": 7.486838184166087,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.000201056918645078,
+      "loss": 0.4827,
+      "step": 150740
+    },
+    {
+      "epoch": 7.487334856461707,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00020101718486142844,
+      "loss": 0.4878,
+      "step": 150750
+    },
+    {
+      "epoch": 7.487831528757326,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00020097745107777889,
+      "loss": 0.5121,
+      "step": 150760
+    },
+    {
+      "epoch": 7.488328201052945,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00020093771729412936,
+      "loss": 0.5013,
+      "step": 150770
+    },
+    {
+      "epoch": 7.488824873348564,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0002008979835104798,
+      "loss": 0.4898,
+      "step": 150780
+    },
+    {
+      "epoch": 7.489321545644184,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00020085824972683022,
+      "loss": 0.5036,
+      "step": 150790
+    },
+    {
+      "epoch": 7.489818217939804,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00020081851594318072,
+      "loss": 0.4944,
+      "step": 150800
+    },
+    {
+      "epoch": 7.490314890235423,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00020077878215953113,
+      "loss": 0.5034,
+      "step": 150810
+    },
+    {
+      "epoch": 7.490811562531042,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00020073904837588163,
+      "loss": 0.5013,
+      "step": 150820
+    },
+    {
+      "epoch": 7.491308234826661,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00020069931459223205,
+      "loss": 0.4785,
+      "step": 150830
+    },
+    {
+      "epoch": 7.4918049071222805,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0002006595808085825,
+      "loss": 0.4658,
+      "step": 150840
+    },
+    {
+      "epoch": 7.4923015794179,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00020061984702493297,
+      "loss": 0.4785,
+      "step": 150850
+    },
+    {
+      "epoch": 7.49279825171352,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0002005801132412834,
+      "loss": 0.4904,
+      "step": 150860
+    },
+    {
+      "epoch": 7.493294924009139,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00020054037945763385,
+      "loss": 0.4657,
+      "step": 150870
+    },
+    {
+      "epoch": 7.493791596304758,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00020050064567398433,
+      "loss": 0.5018,
+      "step": 150880
+    },
+    {
+      "epoch": 7.494288268600378,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00020046091189033477,
+      "loss": 0.5251,
+      "step": 150890
+    },
+    {
+      "epoch": 7.494784940895997,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00020042117810668524,
+      "loss": 0.481,
+      "step": 150900
+    },
+    {
+      "epoch": 7.495281613191616,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00020038144432303569,
+      "loss": 0.4986,
+      "step": 150910
+    },
+    {
+      "epoch": 7.495778285487235,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002003417105393861,
+      "loss": 0.5211,
+      "step": 150920
+    },
+    {
+      "epoch": 7.4962749577828545,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00020030197675573657,
+      "loss": 0.4996,
+      "step": 150930
+    },
+    {
+      "epoch": 7.496771630078475,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00020026224297208702,
+      "loss": 0.4629,
+      "step": 150940
+    },
+    {
+      "epoch": 7.497268302374094,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00020022250918843746,
+      "loss": 0.5066,
+      "step": 150950
+    },
+    {
+      "epoch": 7.497764974669713,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00020018277540478793,
+      "loss": 0.4895,
+      "step": 150960
+    },
+    {
+      "epoch": 7.498261646965332,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00020014304162113838,
+      "loss": 0.4812,
+      "step": 150970
+    },
+    {
+      "epoch": 7.4987583192609515,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00020010330783748885,
+      "loss": 0.4761,
+      "step": 150980
+    },
+    {
+      "epoch": 7.499254991556571,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0002000635740538393,
+      "loss": 0.5024,
+      "step": 150990
+    },
+    {
+      "epoch": 7.49975166385219,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0002000238402701897,
+      "loss": 0.4823,
+      "step": 151000
+    },
+    {
+      "epoch": 7.50024833614781,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001999841064865402,
+      "loss": 0.4834,
+      "step": 151010
+    },
+    {
+      "epoch": 7.500745008443429,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019994437270289063,
+      "loss": 0.4755,
+      "step": 151020
+    },
+    {
+      "epoch": 7.5012416807390485,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001999046389192411,
+      "loss": 0.4558,
+      "step": 151030
+    },
+    {
+      "epoch": 7.501738353034668,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019986490513559154,
+      "loss": 0.5269,
+      "step": 151040
+    },
+    {
+      "epoch": 7.502235025330287,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.000199825171351942,
+      "loss": 0.5016,
+      "step": 151050
+    },
+    {
+      "epoch": 7.502731697625906,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019978543756829246,
+      "loss": 0.4781,
+      "step": 151060
+    },
+    {
+      "epoch": 7.503228369921525,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001997457037846429,
+      "loss": 0.5139,
+      "step": 151070
+    },
+    {
+      "epoch": 7.503725042217145,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019970597000099335,
+      "loss": 0.4934,
+      "step": 151080
+    },
+    {
+      "epoch": 7.504221714512765,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00019966623621734382,
+      "loss": 0.5015,
+      "step": 151090
+    },
+    {
+      "epoch": 7.504718386808384,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019962650243369426,
+      "loss": 0.4734,
+      "step": 151100
+    },
+    {
+      "epoch": 7.505215059104003,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0001995867686500447,
+      "loss": 0.4978,
+      "step": 151110
+    },
+    {
+      "epoch": 7.505711731399622,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00019954703486639518,
+      "loss": 0.5078,
+      "step": 151120
+    },
+    {
+      "epoch": 7.506208403695242,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019950730108274562,
+      "loss": 0.5149,
+      "step": 151130
+    },
+    {
+      "epoch": 7.506705075990861,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00019946756729909607,
+      "loss": 0.4712,
+      "step": 151140
+    },
+    {
+      "epoch": 7.50720174828648,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0001994278335154465,
+      "loss": 0.4818,
+      "step": 151150
+    },
+    {
+      "epoch": 7.5076984205821,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00019938809973179698,
+      "loss": 0.5044,
+      "step": 151160
+    },
+    {
+      "epoch": 7.5081950928777195,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00019934836594814743,
+      "loss": 0.4519,
+      "step": 151170
+    },
+    {
+      "epoch": 7.508691765173339,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00019930863216449787,
+      "loss": 0.5022,
+      "step": 151180
+    },
+    {
+      "epoch": 7.509188437468958,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019926889838084831,
+      "loss": 0.4742,
+      "step": 151190
+    },
+    {
+      "epoch": 7.509685109764577,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019922916459719879,
+      "loss": 0.5147,
+      "step": 151200
+    },
+    {
+      "epoch": 7.510181782060196,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00019918943081354923,
+      "loss": 0.5044,
+      "step": 151210
+    },
+    {
+      "epoch": 7.510678454355816,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019914969702989967,
+      "loss": 0.5168,
+      "step": 151220
+    },
+    {
+      "epoch": 7.511175126651436,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019910996324625012,
+      "loss": 0.4839,
+      "step": 151230
+    },
+    {
+      "epoch": 7.511671798947055,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001990702294626006,
+      "loss": 0.4637,
+      "step": 151240
+    },
+    {
+      "epoch": 7.512168471242674,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019903049567895103,
+      "loss": 0.4912,
+      "step": 151250
+    },
+    {
+      "epoch": 7.512665143538293,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00019899076189530148,
+      "loss": 0.4905,
+      "step": 151260
+    },
+    {
+      "epoch": 7.513161815833913,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019895102811165195,
+      "loss": 0.4811,
+      "step": 151270
+    },
+    {
+      "epoch": 7.513658488129532,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0001989112943280024,
+      "loss": 0.4526,
+      "step": 151280
+    },
+    {
+      "epoch": 7.514155160425151,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019887156054435286,
+      "loss": 0.4655,
+      "step": 151290
+    },
+    {
+      "epoch": 7.514651832720771,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019883182676070328,
+      "loss": 0.4852,
+      "step": 151300
+    },
+    {
+      "epoch": 7.51514850501639,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019879209297705375,
+      "loss": 0.4661,
+      "step": 151310
+    },
+    {
+      "epoch": 7.51564517731201,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0001987523591934042,
+      "loss": 0.4959,
+      "step": 151320
+    },
+    {
+      "epoch": 7.516141849607629,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00019871262540975467,
+      "loss": 0.4951,
+      "step": 151330
+    },
+    {
+      "epoch": 7.516638521903248,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019867289162610509,
+      "loss": 0.4933,
+      "step": 151340
+    },
+    {
+      "epoch": 7.517135194198867,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00019863315784245556,
+      "loss": 0.4918,
+      "step": 151350
+    },
+    {
+      "epoch": 7.517631866494487,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.000198593424058806,
+      "loss": 0.4697,
+      "step": 151360
+    },
+    {
+      "epoch": 7.518128538790107,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00019855369027515647,
+      "loss": 0.4969,
+      "step": 151370
+    },
+    {
+      "epoch": 7.518625211085726,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0001985139564915069,
+      "loss": 0.5117,
+      "step": 151380
+    },
+    {
+      "epoch": 7.519121883381345,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019847422270785736,
+      "loss": 0.5019,
+      "step": 151390
+    },
+    {
+      "epoch": 7.519618555676964,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001984344889242078,
+      "loss": 0.5078,
+      "step": 151400
+    },
+    {
+      "epoch": 7.520115227972584,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019839475514055828,
+      "loss": 0.4668,
+      "step": 151410
+    },
+    {
+      "epoch": 7.520611900268203,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019835502135690872,
+      "loss": 0.4808,
+      "step": 151420
+    },
+    {
+      "epoch": 7.521108572563822,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019831528757325917,
+      "loss": 0.4847,
+      "step": 151430
+    },
+    {
+      "epoch": 7.521605244859442,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00019827555378960964,
+      "loss": 0.4861,
+      "step": 151440
+    },
+    {
+      "epoch": 7.522101917155061,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00019823582000596008,
+      "loss": 0.4836,
+      "step": 151450
+    },
+    {
+      "epoch": 7.522598589450681,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00019819608622231055,
+      "loss": 0.4662,
+      "step": 151460
+    },
+    {
+      "epoch": 7.5230952617463,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019815635243866097,
+      "loss": 0.5157,
+      "step": 151470
+    },
+    {
+      "epoch": 7.523591934041919,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019811661865501144,
+      "loss": 0.4875,
+      "step": 151480
+    },
+    {
+      "epoch": 7.524088606337538,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00019807688487136189,
+      "loss": 0.4465,
+      "step": 151490
+    },
+    {
+      "epoch": 7.5245852786331575,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019803715108771236,
+      "loss": 0.4683,
+      "step": 151500
+    },
+    {
+      "epoch": 7.525081950928778,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00019799741730406277,
+      "loss": 0.4887,
+      "step": 151510
+    },
+    {
+      "epoch": 7.525578623224397,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.00019795768352041325,
+      "loss": 0.5133,
+      "step": 151520
+    },
+    {
+      "epoch": 7.526075295520016,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001979179497367637,
+      "loss": 0.5277,
+      "step": 151530
+    },
+    {
+      "epoch": 7.526571967815635,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00019787821595311416,
+      "loss": 0.4574,
+      "step": 151540
+    },
+    {
+      "epoch": 7.5270686401112545,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00019783848216946458,
+      "loss": 0.5024,
+      "step": 151550
+    },
+    {
+      "epoch": 7.527565312406874,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019779874838581505,
+      "loss": 0.4973,
+      "step": 151560
+    },
+    {
+      "epoch": 7.528061984702493,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0001977590146021655,
+      "loss": 0.496,
+      "step": 151570
+    },
+    {
+      "epoch": 7.528558656998113,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019771928081851596,
+      "loss": 0.4705,
+      "step": 151580
+    },
+    {
+      "epoch": 7.529055329293732,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001976795470348664,
+      "loss": 0.4755,
+      "step": 151590
+    },
+    {
+      "epoch": 7.529552001589352,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019763981325121685,
+      "loss": 0.4684,
+      "step": 151600
+    },
+    {
+      "epoch": 7.530048673884971,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019760007946756732,
+      "loss": 0.4966,
+      "step": 151610
+    },
+    {
+      "epoch": 7.53054534618059,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019756034568391777,
+      "loss": 0.5127,
+      "step": 151620
+    },
+    {
+      "epoch": 7.531042018476209,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001975206119002682,
+      "loss": 0.4807,
+      "step": 151630
+    },
+    {
+      "epoch": 7.5315386907718285,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00019748087811661866,
+      "loss": 0.4676,
+      "step": 151640
+    },
+    {
+      "epoch": 7.532035363067449,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019744114433296913,
+      "loss": 0.4945,
+      "step": 151650
+    },
+    {
+      "epoch": 7.532532035363068,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00019740141054931957,
+      "loss": 0.4862,
+      "step": 151660
+    },
+    {
+      "epoch": 7.533028707658687,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019736167676567002,
+      "loss": 0.4681,
+      "step": 151670
+    },
+    {
+      "epoch": 7.533525379954306,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019732194298202046,
+      "loss": 0.4649,
+      "step": 151680
+    },
+    {
+      "epoch": 7.5340220522499255,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019728220919837093,
+      "loss": 0.5175,
+      "step": 151690
+    },
+    {
+      "epoch": 7.534518724545545,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019724247541472138,
+      "loss": 0.4863,
+      "step": 151700
+    },
+    {
+      "epoch": 7.535015396841164,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019720274163107182,
+      "loss": 0.4879,
+      "step": 151710
+    },
+    {
+      "epoch": 7.535512069136783,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019716300784742227,
+      "loss": 0.4997,
+      "step": 151720
+    },
+    {
+      "epoch": 7.536008741432402,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019712327406377274,
+      "loss": 0.4874,
+      "step": 151730
+    },
+    {
+      "epoch": 7.5365054137280225,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019708354028012318,
+      "loss": 0.4761,
+      "step": 151740
+    },
+    {
+      "epoch": 7.537002086023642,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00019704380649647363,
+      "loss": 0.4938,
+      "step": 151750
+    },
+    {
+      "epoch": 7.537498758319261,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001970040727128241,
+      "loss": 0.4834,
+      "step": 151760
+    },
+    {
+      "epoch": 7.53799543061488,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00019696433892917454,
+      "loss": 0.4996,
+      "step": 151770
+    },
+    {
+      "epoch": 7.538492102910499,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00019692460514552499,
+      "loss": 0.4935,
+      "step": 151780
+    },
+    {
+      "epoch": 7.538988775206119,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00019688487136187543,
+      "loss": 0.4975,
+      "step": 151790
+    },
+    {
+      "epoch": 7.539485447501738,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001968451375782259,
+      "loss": 0.4957,
+      "step": 151800
+    },
+    {
+      "epoch": 7.539982119797358,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019680540379457635,
+      "loss": 0.484,
+      "step": 151810
+    },
+    {
+      "epoch": 7.540478792092977,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00019676567001092682,
+      "loss": 0.5024,
+      "step": 151820
+    },
+    {
+      "epoch": 7.5409754643885964,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019672593622727723,
+      "loss": 0.4973,
+      "step": 151830
+    },
+    {
+      "epoch": 7.541472136684216,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001966862024436277,
+      "loss": 0.5094,
+      "step": 151840
+    },
+    {
+      "epoch": 7.541968808979835,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.00019664646865997815,
+      "loss": 0.4825,
+      "step": 151850
+    },
+    {
+      "epoch": 7.542465481275454,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019660673487632862,
+      "loss": 0.4917,
+      "step": 151860
+    },
+    {
+      "epoch": 7.542962153571073,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019656700109267907,
+      "loss": 0.4823,
+      "step": 151870
+    },
+    {
+      "epoch": 7.5434588258666935,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0001965272673090295,
+      "loss": 0.4685,
+      "step": 151880
+    },
+    {
+      "epoch": 7.543955498162313,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00019648753352537995,
+      "loss": 0.4929,
+      "step": 151890
+    },
+    {
+      "epoch": 7.544452170457932,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019644779974173042,
+      "loss": 0.498,
+      "step": 151900
+    },
+    {
+      "epoch": 7.544948842753551,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00019640806595808087,
+      "loss": 0.5111,
+      "step": 151910
+    },
+    {
+      "epoch": 7.54544551504917,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001963683321744313,
+      "loss": 0.4959,
+      "step": 151920
+    },
+    {
+      "epoch": 7.54594218734479,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019632859839078176,
+      "loss": 0.4935,
+      "step": 151930
+    },
+    {
+      "epoch": 7.546438859640409,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019628886460713223,
+      "loss": 0.5074,
+      "step": 151940
+    },
+    {
+      "epoch": 7.546935531936029,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00019624913082348267,
+      "loss": 0.4886,
+      "step": 151950
+    },
+    {
+      "epoch": 7.547432204231648,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00019620939703983312,
+      "loss": 0.4911,
+      "step": 151960
+    },
+    {
+      "epoch": 7.547928876527267,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001961696632561836,
+      "loss": 0.4812,
+      "step": 151970
+    },
+    {
+      "epoch": 7.548425548822887,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00019612992947253403,
+      "loss": 0.498,
+      "step": 151980
+    },
+    {
+      "epoch": 7.548922221118506,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001960901956888845,
+      "loss": 0.4915,
+      "step": 151990
+    },
+    {
+      "epoch": 7.549418893414125,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00019605046190523492,
+      "loss": 0.5126,
+      "step": 152000
+    },
+    {
+      "epoch": 7.549915565709744,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0001960107281215854,
+      "loss": 0.4934,
+      "step": 152010
+    },
+    {
+      "epoch": 7.550412238005364,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019597099433793584,
+      "loss": 0.5036,
+      "step": 152020
+    },
+    {
+      "epoch": 7.550908910300984,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001959312605542863,
+      "loss": 0.5045,
+      "step": 152030
+    },
+    {
+      "epoch": 7.551405582596603,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019589152677063673,
+      "loss": 0.5075,
+      "step": 152040
+    },
+    {
+      "epoch": 7.551902254892222,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0001958517929869872,
+      "loss": 0.4903,
+      "step": 152050
+    },
+    {
+      "epoch": 7.552398927187841,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019581205920333764,
+      "loss": 0.5088,
+      "step": 152060
+    },
+    {
+      "epoch": 7.552895599483461,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001957723254196881,
+      "loss": 0.4865,
+      "step": 152070
+    },
+    {
+      "epoch": 7.55339227177908,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00019573259163603853,
+      "loss": 0.4959,
+      "step": 152080
+    },
+    {
+      "epoch": 7.5538889440747,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.000195692857852389,
+      "loss": 0.4927,
+      "step": 152090
+    },
+    {
+      "epoch": 7.554385616370319,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019565312406873945,
+      "loss": 0.4903,
+      "step": 152100
+    },
+    {
+      "epoch": 7.554882288665938,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019561339028508992,
+      "loss": 0.4768,
+      "step": 152110
+    },
+    {
+      "epoch": 7.555378960961558,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00019557365650144036,
+      "loss": 0.466,
+      "step": 152120
+    },
+    {
+      "epoch": 7.555875633257177,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001955339227177908,
+      "loss": 0.4701,
+      "step": 152130
+    },
+    {
+      "epoch": 7.556372305552796,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019549418893414128,
+      "loss": 0.4693,
+      "step": 152140
+    },
+    {
+      "epoch": 7.556868977848415,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019545445515049172,
+      "loss": 0.5133,
+      "step": 152150
+    },
+    {
+      "epoch": 7.557365650144035,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00019541472136684217,
+      "loss": 0.4872,
+      "step": 152160
+    },
+    {
+      "epoch": 7.557862322439655,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001953749875831926,
+      "loss": 0.4954,
+      "step": 152170
+    },
+    {
+      "epoch": 7.558358994735274,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019533525379954308,
+      "loss": 0.486,
+      "step": 152180
+    },
+    {
+      "epoch": 7.558855667030893,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00019529552001589353,
+      "loss": 0.5108,
+      "step": 152190
+    },
+    {
+      "epoch": 7.559352339326512,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00019525578623224397,
+      "loss": 0.4682,
+      "step": 152200
+    },
+    {
+      "epoch": 7.5598490116221315,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0001952160524485944,
+      "loss": 0.4867,
+      "step": 152210
+    },
+    {
+      "epoch": 7.560345683917751,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.00019517631866494488,
+      "loss": 0.4885,
+      "step": 152220
+    },
+    {
+      "epoch": 7.560842356213371,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00019513658488129533,
+      "loss": 0.5012,
+      "step": 152230
+    },
+    {
+      "epoch": 7.56133902850899,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00019509685109764577,
+      "loss": 0.4948,
+      "step": 152240
+    },
+    {
+      "epoch": 7.561835700804609,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019505711731399622,
+      "loss": 0.5236,
+      "step": 152250
+    },
+    {
+      "epoch": 7.5623323731002285,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0001950173835303467,
+      "loss": 0.4801,
+      "step": 152260
+    },
+    {
+      "epoch": 7.562829045395848,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019497764974669713,
+      "loss": 0.5022,
+      "step": 152270
+    },
+    {
+      "epoch": 7.563325717691467,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0001949379159630476,
+      "loss": 0.4811,
+      "step": 152280
+    },
+    {
+      "epoch": 7.563822389987086,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019489818217939805,
+      "loss": 0.5052,
+      "step": 152290
+    },
+    {
+      "epoch": 7.564319062282706,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001948584483957485,
+      "loss": 0.5147,
+      "step": 152300
+    },
+    {
+      "epoch": 7.564815734578326,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019481871461209896,
+      "loss": 0.4936,
+      "step": 152310
+    },
+    {
+      "epoch": 7.565312406873945,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001947789808284494,
+      "loss": 0.4627,
+      "step": 152320
+    },
+    {
+      "epoch": 7.565809079169564,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019473924704479985,
+      "loss": 0.4492,
+      "step": 152330
+    },
+    {
+      "epoch": 7.566305751465183,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001946995132611503,
+      "loss": 0.4764,
+      "step": 152340
+    },
+    {
+      "epoch": 7.5668024237608025,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019465977947750077,
+      "loss": 0.5023,
+      "step": 152350
+    },
+    {
+      "epoch": 7.567299096056422,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0001946200456938512,
+      "loss": 0.5026,
+      "step": 152360
+    },
+    {
+      "epoch": 7.567795768352042,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00019458031191020166,
+      "loss": 0.4972,
+      "step": 152370
+    },
+    {
+      "epoch": 7.568292440647661,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0001945405781265521,
+      "loss": 0.4831,
+      "step": 152380
+    },
+    {
+      "epoch": 7.56878911294328,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019450084434290257,
+      "loss": 0.4975,
+      "step": 152390
+    },
+    {
+      "epoch": 7.5692857852388995,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00019446111055925302,
+      "loss": 0.4622,
+      "step": 152400
+    },
+    {
+      "epoch": 7.569782457534519,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019442137677560346,
+      "loss": 0.5062,
+      "step": 152410
+    },
+    {
+      "epoch": 7.570279129830138,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001943816429919539,
+      "loss": 0.4771,
+      "step": 152420
+    },
+    {
+      "epoch": 7.570775802125757,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019434190920830438,
+      "loss": 0.4967,
+      "step": 152430
+    },
+    {
+      "epoch": 7.571272474421376,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00019430217542465482,
+      "loss": 0.492,
+      "step": 152440
+    },
+    {
+      "epoch": 7.571769146716996,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.00019426244164100527,
+      "loss": 0.5033,
+      "step": 152450
+    },
+    {
+      "epoch": 7.572265819012616,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00019422270785735574,
+      "loss": 0.4958,
+      "step": 152460
+    },
+    {
+      "epoch": 7.572762491308235,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019418297407370618,
+      "loss": 0.4741,
+      "step": 152470
+    },
+    {
+      "epoch": 7.573259163603854,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019414324029005663,
+      "loss": 0.4976,
+      "step": 152480
+    },
+    {
+      "epoch": 7.573755835899473,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019410350650640707,
+      "loss": 0.4931,
+      "step": 152490
+    },
+    {
+      "epoch": 7.574252508195093,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00019406377272275754,
+      "loss": 0.487,
+      "step": 152500
+    },
+    {
+      "epoch": 7.574749180490712,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019402403893910798,
+      "loss": 0.4996,
+      "step": 152510
+    },
+    {
+      "epoch": 7.575245852786331,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019398430515545846,
+      "loss": 0.4514,
+      "step": 152520
+    },
+    {
+      "epoch": 7.575742525081951,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019394457137180887,
+      "loss": 0.4785,
+      "step": 152530
+    },
+    {
+      "epoch": 7.5762391973775705,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00019390483758815934,
+      "loss": 0.499,
+      "step": 152540
+    },
+    {
+      "epoch": 7.57673586967319,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0001938651038045098,
+      "loss": 0.4918,
+      "step": 152550
+    },
+    {
+      "epoch": 7.577232541968809,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019382537002086026,
+      "loss": 0.47,
+      "step": 152560
+    },
+    {
+      "epoch": 7.577729214264428,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019378563623721068,
+      "loss": 0.4788,
+      "step": 152570
+    },
+    {
+      "epoch": 7.578225886560047,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019374590245356115,
+      "loss": 0.4927,
+      "step": 152580
+    },
+    {
+      "epoch": 7.578722558855667,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001937061686699116,
+      "loss": 0.4955,
+      "step": 152590
+    },
+    {
+      "epoch": 7.579219231151287,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00019366643488626206,
+      "loss": 0.4595,
+      "step": 152600
+    },
+    {
+      "epoch": 7.579715903446906,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001936267011026125,
+      "loss": 0.482,
+      "step": 152610
+    },
+    {
+      "epoch": 7.580212575742525,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019358696731896295,
+      "loss": 0.4559,
+      "step": 152620
+    },
+    {
+      "epoch": 7.580709248038144,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001935472335353134,
+      "loss": 0.4868,
+      "step": 152630
+    },
+    {
+      "epoch": 7.581205920333764,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019350749975166387,
+      "loss": 0.498,
+      "step": 152640
+    },
+    {
+      "epoch": 7.581702592629383,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001934677659680143,
+      "loss": 0.5297,
+      "step": 152650
+    },
+    {
+      "epoch": 7.582199264925002,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019342803218436476,
+      "loss": 0.4542,
+      "step": 152660
+    },
+    {
+      "epoch": 7.582695937220622,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019338829840071523,
+      "loss": 0.4785,
+      "step": 152670
+    },
+    {
+      "epoch": 7.583192609516241,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019334856461706567,
+      "loss": 0.4759,
+      "step": 152680
+    },
+    {
+      "epoch": 7.583689281811861,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019330883083341612,
+      "loss": 0.5101,
+      "step": 152690
+    },
+    {
+      "epoch": 7.58418595410748,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019326909704976656,
+      "loss": 0.4911,
+      "step": 152700
+    },
+    {
+      "epoch": 7.584682626403099,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00019322936326611703,
+      "loss": 0.4462,
+      "step": 152710
+    },
+    {
+      "epoch": 7.585179298698718,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00019318962948246748,
+      "loss": 0.4964,
+      "step": 152720
+    },
+    {
+      "epoch": 7.5856759709943375,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00019314989569881795,
+      "loss": 0.4997,
+      "step": 152730
+    },
+    {
+      "epoch": 7.586172643289958,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019311016191516837,
+      "loss": 0.4711,
+      "step": 152740
+    },
+    {
+      "epoch": 7.586669315585577,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019307042813151884,
+      "loss": 0.4773,
+      "step": 152750
+    },
+    {
+      "epoch": 7.587165987881196,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00019303069434786928,
+      "loss": 0.4628,
+      "step": 152760
+    },
+    {
+      "epoch": 7.587662660176815,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019299096056421975,
+      "loss": 0.4686,
+      "step": 152770
+    },
+    {
+      "epoch": 7.588159332472435,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019295122678057017,
+      "loss": 0.4697,
+      "step": 152780
+    },
+    {
+      "epoch": 7.588656004768054,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00019291149299692064,
+      "loss": 0.4738,
+      "step": 152790
+    },
+    {
+      "epoch": 7.589152677063673,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00019287175921327109,
+      "loss": 0.4983,
+      "step": 152800
+    },
+    {
+      "epoch": 7.589649349359293,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00019283202542962156,
+      "loss": 0.4926,
+      "step": 152810
+    },
+    {
+      "epoch": 7.590146021654912,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.000192792291645972,
+      "loss": 0.4742,
+      "step": 152820
+    },
+    {
+      "epoch": 7.590642693950532,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019275255786232244,
+      "loss": 0.4904,
+      "step": 152830
+    },
+    {
+      "epoch": 7.591139366246151,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019271282407867292,
+      "loss": 0.4825,
+      "step": 152840
+    },
+    {
+      "epoch": 7.59163603854177,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019267309029502336,
+      "loss": 0.4604,
+      "step": 152850
+    },
+    {
+      "epoch": 7.592132710837389,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0001926333565113738,
+      "loss": 0.4852,
+      "step": 152860
+    },
+    {
+      "epoch": 7.5926293831330085,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00019259362272772425,
+      "loss": 0.4775,
+      "step": 152870
+    },
+    {
+      "epoch": 7.593126055428629,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00019255388894407472,
+      "loss": 0.4789,
+      "step": 152880
+    },
+    {
+      "epoch": 7.593622727724248,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019251415516042516,
+      "loss": 0.4874,
+      "step": 152890
+    },
+    {
+      "epoch": 7.594119400019867,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0001924744213767756,
+      "loss": 0.4828,
+      "step": 152900
+    },
+    {
+      "epoch": 7.594616072315486,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019243468759312605,
+      "loss": 0.4898,
+      "step": 152910
+    },
+    {
+      "epoch": 7.5951127446111055,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019239495380947652,
+      "loss": 0.4894,
+      "step": 152920
+    },
+    {
+      "epoch": 7.595609416906725,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019235522002582697,
+      "loss": 0.5232,
+      "step": 152930
+    },
+    {
+      "epoch": 7.596106089202344,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001923154862421774,
+      "loss": 0.4549,
+      "step": 152940
+    },
+    {
+      "epoch": 7.596602761497964,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00019227575245852786,
+      "loss": 0.469,
+      "step": 152950
+    },
+    {
+      "epoch": 7.597099433793583,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019223601867487833,
+      "loss": 0.517,
+      "step": 152960
+    },
+    {
+      "epoch": 7.5975961060892026,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019219628489122877,
+      "loss": 0.4966,
+      "step": 152970
+    },
+    {
+      "epoch": 7.598092778384822,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00019215655110757922,
+      "loss": 0.4698,
+      "step": 152980
+    },
+    {
+      "epoch": 7.598589450680441,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001921168173239297,
+      "loss": 0.4923,
+      "step": 152990
+    },
+    {
+      "epoch": 7.59908612297606,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019207708354028013,
+      "loss": 0.4954,
+      "step": 153000
+    },
+    {
+      "epoch": 7.5995827952716795,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0001920373497566306,
+      "loss": 0.5022,
+      "step": 153010
+    },
+    {
+      "epoch": 7.6000794675673,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019199761597298102,
+      "loss": 0.4913,
+      "step": 153020
+    },
+    {
+      "epoch": 7.600576139862919,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001919578821893315,
+      "loss": 0.5118,
+      "step": 153030
+    },
+    {
+      "epoch": 7.601072812158538,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00019191814840568194,
+      "loss": 0.4902,
+      "step": 153040
+    },
+    {
+      "epoch": 7.601569484454157,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001918784146220324,
+      "loss": 0.4935,
+      "step": 153050
+    },
+    {
+      "epoch": 7.6020661567497765,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00019183868083838283,
+      "loss": 0.498,
+      "step": 153060
+    },
+    {
+      "epoch": 7.602562829045396,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0001917989470547333,
+      "loss": 0.4913,
+      "step": 153070
+    },
+    {
+      "epoch": 7.603059501341015,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019175921327108374,
+      "loss": 0.477,
+      "step": 153080
+    },
+    {
+      "epoch": 7.603556173636634,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001917194794874342,
+      "loss": 0.4654,
+      "step": 153090
+    },
+    {
+      "epoch": 7.604052845932254,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00019167974570378463,
+      "loss": 0.4991,
+      "step": 153100
+    },
+    {
+      "epoch": 7.6045495182278735,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0001916400119201351,
+      "loss": 0.4697,
+      "step": 153110
+    },
+    {
+      "epoch": 7.605046190523493,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019160027813648555,
+      "loss": 0.4818,
+      "step": 153120
+    },
+    {
+      "epoch": 7.605542862819112,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019156054435283602,
+      "loss": 0.4834,
+      "step": 153130
+    },
+    {
+      "epoch": 7.606039535114731,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019152081056918646,
+      "loss": 0.4731,
+      "step": 153140
+    },
+    {
+      "epoch": 7.60653620741035,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0001914810767855369,
+      "loss": 0.4613,
+      "step": 153150
+    },
+    {
+      "epoch": 7.60703287970597,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00019144134300188738,
+      "loss": 0.5042,
+      "step": 153160
+    },
+    {
+      "epoch": 7.607529552001589,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00019140160921823782,
+      "loss": 0.5001,
+      "step": 153170
+    },
+    {
+      "epoch": 7.608026224297209,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019136187543458826,
+      "loss": 0.484,
+      "step": 153180
+    },
+    {
+      "epoch": 7.608522896592828,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001913221416509387,
+      "loss": 0.4745,
+      "step": 153190
+    },
+    {
+      "epoch": 7.609019568888447,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00019128240786728918,
+      "loss": 0.4809,
+      "step": 153200
+    },
+    {
+      "epoch": 7.609516241184067,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019124267408363962,
+      "loss": 0.4721,
+      "step": 153210
+    },
+    {
+      "epoch": 7.610012913479686,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0001912029402999901,
+      "loss": 0.4792,
+      "step": 153220
+    },
+    {
+      "epoch": 7.610509585775305,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0001911632065163405,
+      "loss": 0.4953,
+      "step": 153230
+    },
+    {
+      "epoch": 7.611006258070924,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00019112347273269098,
+      "loss": 0.4718,
+      "step": 153240
+    },
+    {
+      "epoch": 7.6115029303665445,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019108373894904143,
+      "loss": 0.5329,
+      "step": 153250
+    },
+    {
+      "epoch": 7.611999602662164,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001910440051653919,
+      "loss": 0.5006,
+      "step": 153260
+    },
+    {
+      "epoch": 7.612496274957783,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019100427138174232,
+      "loss": 0.4735,
+      "step": 153270
+    },
+    {
+      "epoch": 7.612992947253402,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0001909645375980928,
+      "loss": 0.5047,
+      "step": 153280
+    },
+    {
+      "epoch": 7.613489619549021,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019092480381444323,
+      "loss": 0.5039,
+      "step": 153290
+    },
+    {
+      "epoch": 7.613986291844641,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0001908850700307937,
+      "loss": 0.4499,
+      "step": 153300
+    },
+    {
+      "epoch": 7.61448296414026,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00019084533624714415,
+      "loss": 0.5066,
+      "step": 153310
+    },
+    {
+      "epoch": 7.61497963643588,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001908056024634946,
+      "loss": 0.4999,
+      "step": 153320
+    },
+    {
+      "epoch": 7.615476308731499,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019076586867984504,
+      "loss": 0.5119,
+      "step": 153330
+    },
+    {
+      "epoch": 7.615972981027118,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001907261348961955,
+      "loss": 0.489,
+      "step": 153340
+    },
+    {
+      "epoch": 7.616469653322738,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00019068640111254595,
+      "loss": 0.4963,
+      "step": 153350
+    },
+    {
+      "epoch": 7.616966325618357,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001906466673288964,
+      "loss": 0.5032,
+      "step": 153360
+    },
+    {
+      "epoch": 7.617462997913976,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019060693354524687,
+      "loss": 0.4976,
+      "step": 153370
+    },
+    {
+      "epoch": 7.617959670209595,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001905671997615973,
+      "loss": 0.4896,
+      "step": 153380
+    },
+    {
+      "epoch": 7.618456342505215,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00019052746597794776,
+      "loss": 0.5082,
+      "step": 153390
+    },
+    {
+      "epoch": 7.618953014800835,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001904877321942982,
+      "loss": 0.495,
+      "step": 153400
+    },
+    {
+      "epoch": 7.619449687096454,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00019044799841064867,
+      "loss": 0.5106,
+      "step": 153410
+    },
+    {
+      "epoch": 7.619946359392073,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00019040826462699912,
+      "loss": 0.4946,
+      "step": 153420
+    },
+    {
+      "epoch": 7.620443031687692,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019036853084334956,
+      "loss": 0.4776,
+      "step": 153430
+    },
+    {
+      "epoch": 7.6209397039833116,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0001903287970597,
+      "loss": 0.5085,
+      "step": 153440
+    },
+    {
+      "epoch": 7.621436376278931,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019028906327605048,
+      "loss": 0.5323,
+      "step": 153450
+    },
+    {
+      "epoch": 7.621933048574551,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019024932949240092,
+      "loss": 0.4859,
+      "step": 153460
+    },
+    {
+      "epoch": 7.62242972087017,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00019020959570875136,
+      "loss": 0.5257,
+      "step": 153470
+    },
+    {
+      "epoch": 7.622926393165789,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001901698619251018,
+      "loss": 0.4878,
+      "step": 153480
+    },
+    {
+      "epoch": 7.623423065461409,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019013012814145228,
+      "loss": 0.4682,
+      "step": 153490
+    },
+    {
+      "epoch": 7.623919737757028,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00019009039435780272,
+      "loss": 0.4855,
+      "step": 153500
+    },
+    {
+      "epoch": 7.624416410052647,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019005066057415317,
+      "loss": 0.4881,
+      "step": 153510
+    },
+    {
+      "epoch": 7.624913082348266,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019001092679050364,
+      "loss": 0.4935,
+      "step": 153520
+    },
+    {
+      "epoch": 7.625409754643886,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00018997119300685408,
+      "loss": 0.4882,
+      "step": 153530
+    },
+    {
+      "epoch": 7.625906426939506,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00018993145922320456,
+      "loss": 0.5026,
+      "step": 153540
+    },
+    {
+      "epoch": 7.626403099235125,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.000189891725439555,
+      "loss": 0.4697,
+      "step": 153550
+    },
+    {
+      "epoch": 7.626899771530744,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018985199165590544,
+      "loss": 0.493,
+      "step": 153560
+    },
+    {
+      "epoch": 7.627396443826363,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001898122578722559,
+      "loss": 0.5156,
+      "step": 153570
+    },
+    {
+      "epoch": 7.6278931161219825,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018977252408860636,
+      "loss": 0.5061,
+      "step": 153580
+    },
+    {
+      "epoch": 7.628389788417602,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001897327903049568,
+      "loss": 0.4553,
+      "step": 153590
+    },
+    {
+      "epoch": 7.628886460713222,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00018969305652130725,
+      "loss": 0.4785,
+      "step": 153600
+    },
+    {
+      "epoch": 7.629383133008841,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001896533227376577,
+      "loss": 0.4759,
+      "step": 153610
+    },
+    {
+      "epoch": 7.62987980530446,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00018961358895400816,
+      "loss": 0.4862,
+      "step": 153620
+    },
+    {
+      "epoch": 7.6303764776000795,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001895738551703586,
+      "loss": 0.5018,
+      "step": 153630
+    },
+    {
+      "epoch": 7.630873149895699,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00018953412138670905,
+      "loss": 0.484,
+      "step": 153640
+    },
+    {
+      "epoch": 7.631369822191318,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001894943876030595,
+      "loss": 0.5266,
+      "step": 153650
+    },
+    {
+      "epoch": 7.631866494486937,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00018945465381940997,
+      "loss": 0.4716,
+      "step": 153660
+    },
+    {
+      "epoch": 7.632363166782557,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0001894149200357604,
+      "loss": 0.4972,
+      "step": 153670
+    },
+    {
+      "epoch": 7.632859839078177,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00018937518625211086,
+      "loss": 0.4969,
+      "step": 153680
+    },
+    {
+      "epoch": 7.633356511373796,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00018933545246846133,
+      "loss": 0.5109,
+      "step": 153690
+    },
+    {
+      "epoch": 7.633853183669415,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018929571868481177,
+      "loss": 0.4701,
+      "step": 153700
+    },
+    {
+      "epoch": 7.634349855965034,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00018925598490116224,
+      "loss": 0.4861,
+      "step": 153710
+    },
+    {
+      "epoch": 7.6348465282606535,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00018921625111751266,
+      "loss": 0.508,
+      "step": 153720
+    },
+    {
+      "epoch": 7.635343200556273,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018917651733386313,
+      "loss": 0.4445,
+      "step": 153730
+    },
+    {
+      "epoch": 7.635839872851893,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00018913678355021358,
+      "loss": 0.5065,
+      "step": 153740
+    },
+    {
+      "epoch": 7.636336545147512,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00018909704976656405,
+      "loss": 0.5,
+      "step": 153750
+    },
+    {
+      "epoch": 7.636833217443131,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00018905731598291446,
+      "loss": 0.4926,
+      "step": 153760
+    },
+    {
+      "epoch": 7.6373298897387505,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018901758219926494,
+      "loss": 0.501,
+      "step": 153770
+    },
+    {
+      "epoch": 7.63782656203437,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00018897784841561538,
+      "loss": 0.4738,
+      "step": 153780
+    },
+    {
+      "epoch": 7.638323234329989,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00018893811463196585,
+      "loss": 0.4654,
+      "step": 153790
+    },
+    {
+      "epoch": 7.638819906625608,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00018889838084831627,
+      "loss": 0.4611,
+      "step": 153800
+    },
+    {
+      "epoch": 7.639316578921227,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00018885864706466674,
+      "loss": 0.5073,
+      "step": 153810
+    },
+    {
+      "epoch": 7.6398132512168475,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00018881891328101718,
+      "loss": 0.5262,
+      "step": 153820
+    },
+    {
+      "epoch": 7.640309923512467,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00018877917949736766,
+      "loss": 0.4801,
+      "step": 153830
+    },
+    {
+      "epoch": 7.640806595808086,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001887394457137181,
+      "loss": 0.4817,
+      "step": 153840
+    },
+    {
+      "epoch": 7.641303268103705,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00018869971193006854,
+      "loss": 0.4774,
+      "step": 153850
+    },
+    {
+      "epoch": 7.641799940399324,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018865997814641902,
+      "loss": 0.5228,
+      "step": 153860
+    },
+    {
+      "epoch": 7.642296612694944,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00018862024436276946,
+      "loss": 0.4823,
+      "step": 153870
+    },
+    {
+      "epoch": 7.642793284990563,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001885805105791199,
+      "loss": 0.491,
+      "step": 153880
+    },
+    {
+      "epoch": 7.643289957286182,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00018854077679547035,
+      "loss": 0.4891,
+      "step": 153890
+    },
+    {
+      "epoch": 7.643786629581802,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00018850104301182082,
+      "loss": 0.4896,
+      "step": 153900
+    },
+    {
+      "epoch": 7.644283301877421,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00018846130922817126,
+      "loss": 0.4967,
+      "step": 153910
+    },
+    {
+      "epoch": 7.644779974173041,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0001884215754445217,
+      "loss": 0.4747,
+      "step": 153920
+    },
+    {
+      "epoch": 7.64527664646866,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00018838184166087215,
+      "loss": 0.4724,
+      "step": 153930
+    },
+    {
+      "epoch": 7.645773318764279,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00018834210787722262,
+      "loss": 0.5126,
+      "step": 153940
+    },
+    {
+      "epoch": 7.646269991059898,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00018830237409357307,
+      "loss": 0.4926,
+      "step": 153950
+    },
+    {
+      "epoch": 7.646766663355518,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00018826264030992354,
+      "loss": 0.5141,
+      "step": 153960
+    },
+    {
+      "epoch": 7.647263335651138,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00018822290652627396,
+      "loss": 0.4958,
+      "step": 153970
+    },
+    {
+      "epoch": 7.647760007946757,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00018818317274262443,
+      "loss": 0.488,
+      "step": 153980
+    },
+    {
+      "epoch": 7.648256680242376,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00018814343895897487,
+      "loss": 0.5178,
+      "step": 153990
+    },
+    {
+      "epoch": 7.648753352537995,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00018810370517532534,
+      "loss": 0.4964,
+      "step": 154000
+    },
+    {
+      "epoch": 7.649250024833615,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001880639713916758,
+      "loss": 0.4906,
+      "step": 154010
+    },
+    {
+      "epoch": 7.649746697129234,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00018802423760802623,
+      "loss": 0.4911,
+      "step": 154020
+    },
+    {
+      "epoch": 7.650243369424853,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00018798450382437668,
+      "loss": 0.5231,
+      "step": 154030
+    },
+    {
+      "epoch": 7.650740041720473,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00018794477004072715,
+      "loss": 0.485,
+      "step": 154040
+    },
+    {
+      "epoch": 7.651236714016092,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001879050362570776,
+      "loss": 0.4854,
+      "step": 154050
+    },
+    {
+      "epoch": 7.651733386311712,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018786530247342804,
+      "loss": 0.5198,
+      "step": 154060
+    },
+    {
+      "epoch": 7.652230058607331,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001878255686897785,
+      "loss": 0.5184,
+      "step": 154070
+    },
+    {
+      "epoch": 7.65272673090295,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018778583490612895,
+      "loss": 0.5059,
+      "step": 154080
+    },
+    {
+      "epoch": 7.653223403198569,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0001877461011224794,
+      "loss": 0.4949,
+      "step": 154090
+    },
+    {
+      "epoch": 7.6537200754941885,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00018770636733882984,
+      "loss": 0.4948,
+      "step": 154100
+    },
+    {
+      "epoch": 7.654216747789809,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0001876666335551803,
+      "loss": 0.462,
+      "step": 154110
+    },
+    {
+      "epoch": 7.654713420085428,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00018762689977153076,
+      "loss": 0.4503,
+      "step": 154120
+    },
+    {
+      "epoch": 7.655210092381047,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001875871659878812,
+      "loss": 0.4983,
+      "step": 154130
+    },
+    {
+      "epoch": 7.655706764676666,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018754743220423164,
+      "loss": 0.4772,
+      "step": 154140
+    },
+    {
+      "epoch": 7.656203436972286,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00018750769842058212,
+      "loss": 0.5236,
+      "step": 154150
+    },
+    {
+      "epoch": 7.656700109267905,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00018746796463693256,
+      "loss": 0.4789,
+      "step": 154160
+    },
+    {
+      "epoch": 7.657196781563524,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.000187428230853283,
+      "loss": 0.4866,
+      "step": 154170
+    },
+    {
+      "epoch": 7.657693453859144,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00018738849706963345,
+      "loss": 0.4892,
+      "step": 154180
+    },
+    {
+      "epoch": 7.658190126154763,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00018734876328598392,
+      "loss": 0.506,
+      "step": 154190
+    },
+    {
+      "epoch": 7.658686798450383,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00018730902950233436,
+      "loss": 0.4922,
+      "step": 154200
+    },
+    {
+      "epoch": 7.659183470746002,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001872692957186848,
+      "loss": 0.4795,
+      "step": 154210
+    },
+    {
+      "epoch": 7.659680143041621,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00018722956193503528,
+      "loss": 0.5183,
+      "step": 154220
+    },
+    {
+      "epoch": 7.66017681533724,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018718982815138572,
+      "loss": 0.4634,
+      "step": 154230
+    },
+    {
+      "epoch": 7.6606734876328595,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0001871500943677362,
+      "loss": 0.5115,
+      "step": 154240
+    },
+    {
+      "epoch": 7.66117015992848,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001871103605840866,
+      "loss": 0.5029,
+      "step": 154250
+    },
+    {
+      "epoch": 7.661666832224099,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00018707062680043708,
+      "loss": 0.5018,
+      "step": 154260
+    },
+    {
+      "epoch": 7.662163504519718,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.00018703089301678753,
+      "loss": 0.4781,
+      "step": 154270
+    },
+    {
+      "epoch": 7.662660176815337,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.000186991159233138,
+      "loss": 0.4766,
+      "step": 154280
+    },
+    {
+      "epoch": 7.6631568491109565,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00018695142544948842,
+      "loss": 0.4757,
+      "step": 154290
+    },
+    {
+      "epoch": 7.663653521406576,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0001869116916658389,
+      "loss": 0.5078,
+      "step": 154300
+    },
+    {
+      "epoch": 7.664150193702195,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.00018687195788218933,
+      "loss": 0.4786,
+      "step": 154310
+    },
+    {
+      "epoch": 7.664646865997815,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001868322240985398,
+      "loss": 0.5029,
+      "step": 154320
+    },
+    {
+      "epoch": 7.665143538293434,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00018679249031489022,
+      "loss": 0.4848,
+      "step": 154330
+    },
+    {
+      "epoch": 7.6656402105890535,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001867527565312407,
+      "loss": 0.499,
+      "step": 154340
+    },
+    {
+      "epoch": 7.666136882884673,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00018671302274759114,
+      "loss": 0.4852,
+      "step": 154350
+    },
+    {
+      "epoch": 7.666633555180292,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001866732889639416,
+      "loss": 0.4741,
+      "step": 154360
+    },
+    {
+      "epoch": 7.667130227475911,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00018663355518029205,
+      "loss": 0.4989,
+      "step": 154370
+    },
+    {
+      "epoch": 7.6676268997715304,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0001865938213966425,
+      "loss": 0.5169,
+      "step": 154380
+    },
+    {
+      "epoch": 7.668123572067151,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00018655408761299297,
+      "loss": 0.4676,
+      "step": 154390
+    },
+    {
+      "epoch": 7.66862024436277,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0001865143538293434,
+      "loss": 0.4761,
+      "step": 154400
+    },
+    {
+      "epoch": 7.669116916658389,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00018647462004569388,
+      "loss": 0.5073,
+      "step": 154410
+    },
+    {
+      "epoch": 7.669613588954008,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0001864348862620443,
+      "loss": 0.4982,
+      "step": 154420
+    },
+    {
+      "epoch": 7.6701102612496275,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00018639515247839477,
+      "loss": 0.4747,
+      "step": 154430
+    },
+    {
+      "epoch": 7.670606933545247,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00018635541869474522,
+      "loss": 0.4954,
+      "step": 154440
+    },
+    {
+      "epoch": 7.671103605840866,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001863156849110957,
+      "loss": 0.4677,
+      "step": 154450
+    },
+    {
+      "epoch": 7.671600278136486,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001862759511274461,
+      "loss": 0.5054,
+      "step": 154460
+    },
+    {
+      "epoch": 7.672096950432105,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00018623621734379658,
+      "loss": 0.4732,
+      "step": 154470
+    },
+    {
+      "epoch": 7.6725936227277245,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00018619648356014702,
+      "loss": 0.486,
+      "step": 154480
+    },
+    {
+      "epoch": 7.673090295023344,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0001861567497764975,
+      "loss": 0.5007,
+      "step": 154490
+    },
+    {
+      "epoch": 7.673586967318963,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001861170159928479,
+      "loss": 0.4551,
+      "step": 154500
+    },
+    {
+      "epoch": 7.674083639614582,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018607728220919838,
+      "loss": 0.512,
+      "step": 154510
+    },
+    {
+      "epoch": 7.674580311910201,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00018603754842554882,
+      "loss": 0.5094,
+      "step": 154520
+    },
+    {
+      "epoch": 7.675076984205821,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001859978146418993,
+      "loss": 0.4922,
+      "step": 154530
+    },
+    {
+      "epoch": 7.67557365650144,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00018595808085824974,
+      "loss": 0.4668,
+      "step": 154540
+    },
+    {
+      "epoch": 7.67607032879706,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00018591834707460018,
+      "loss": 0.4956,
+      "step": 154550
+    },
+    {
+      "epoch": 7.676567001092679,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00018587861329095066,
+      "loss": 0.5197,
+      "step": 154560
+    },
+    {
+      "epoch": 7.677063673388298,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001858388795073011,
+      "loss": 0.466,
+      "step": 154570
+    },
+    {
+      "epoch": 7.677560345683918,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00018579914572365154,
+      "loss": 0.4589,
+      "step": 154580
+    },
+    {
+      "epoch": 7.678057017979537,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.000185759411940002,
+      "loss": 0.4978,
+      "step": 154590
+    },
+    {
+      "epoch": 7.678553690275156,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00018571967815635246,
+      "loss": 0.4784,
+      "step": 154600
+    },
+    {
+      "epoch": 7.679050362570775,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001856799443727029,
+      "loss": 0.4704,
+      "step": 154610
+    },
+    {
+      "epoch": 7.6795470348663954,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00018564021058905335,
+      "loss": 0.4839,
+      "step": 154620
+    },
+    {
+      "epoch": 7.680043707162015,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001856004768054038,
+      "loss": 0.4889,
+      "step": 154630
+    },
+    {
+      "epoch": 7.680540379457634,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00018556074302175426,
+      "loss": 0.4979,
+      "step": 154640
+    },
+    {
+      "epoch": 7.681037051753253,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001855210092381047,
+      "loss": 0.5,
+      "step": 154650
+    },
+    {
+      "epoch": 7.681533724048872,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00018548127545445515,
+      "loss": 0.504,
+      "step": 154660
+    },
+    {
+      "epoch": 7.682030396344492,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001854415416708056,
+      "loss": 0.4719,
+      "step": 154670
+    },
+    {
+      "epoch": 7.682527068640111,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00018540180788715607,
+      "loss": 0.4961,
+      "step": 154680
+    },
+    {
+      "epoch": 7.683023740935731,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001853620741035065,
+      "loss": 0.4992,
+      "step": 154690
+    },
+    {
+      "epoch": 7.68352041323135,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00018532234031985696,
+      "loss": 0.4729,
+      "step": 154700
+    },
+    {
+      "epoch": 7.684017085526969,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00018528260653620743,
+      "loss": 0.5064,
+      "step": 154710
+    },
+    {
+      "epoch": 7.684513757822589,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00018524287275255787,
+      "loss": 0.4831,
+      "step": 154720
+    },
+    {
+      "epoch": 7.685010430118208,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018520313896890832,
+      "loss": 0.4868,
+      "step": 154730
+    },
+    {
+      "epoch": 7.685507102413827,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018516340518525876,
+      "loss": 0.4677,
+      "step": 154740
+    },
+    {
+      "epoch": 7.686003774709446,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00018512367140160923,
+      "loss": 0.5095,
+      "step": 154750
+    },
+    {
+      "epoch": 7.686500447005066,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00018508393761795968,
+      "loss": 0.5201,
+      "step": 154760
+    },
+    {
+      "epoch": 7.686997119300686,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00018504420383431015,
+      "loss": 0.4974,
+      "step": 154770
+    },
+    {
+      "epoch": 7.687493791596305,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00018500447005066056,
+      "loss": 0.479,
+      "step": 154780
+    },
+    {
+      "epoch": 7.687990463891924,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00018496473626701104,
+      "loss": 0.5237,
+      "step": 154790
+    },
+    {
+      "epoch": 7.688487136187543,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00018492500248336148,
+      "loss": 0.4972,
+      "step": 154800
+    },
+    {
+      "epoch": 7.6889838084831625,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00018488526869971195,
+      "loss": 0.5134,
+      "step": 154810
+    },
+    {
+      "epoch": 7.689480480778782,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001848455349160624,
+      "loss": 0.4614,
+      "step": 154820
+    },
+    {
+      "epoch": 7.689977153074402,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00018480580113241284,
+      "loss": 0.508,
+      "step": 154830
+    },
+    {
+      "epoch": 7.690473825370021,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00018476606734876328,
+      "loss": 0.4976,
+      "step": 154840
+    },
+    {
+      "epoch": 7.69097049766564,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018472633356511376,
+      "loss": 0.4854,
+      "step": 154850
+    },
+    {
+      "epoch": 7.69146716996126,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001846865997814642,
+      "loss": 0.4516,
+      "step": 154860
+    },
+    {
+      "epoch": 7.691963842256879,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00018464686599781464,
+      "loss": 0.4901,
+      "step": 154870
+    },
+    {
+      "epoch": 7.692460514552498,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001846071322141651,
+      "loss": 0.484,
+      "step": 154880
+    },
+    {
+      "epoch": 7.692957186848117,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018456739843051556,
+      "loss": 0.446,
+      "step": 154890
+    },
+    {
+      "epoch": 7.693453859143737,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.000184527664646866,
+      "loss": 0.4732,
+      "step": 154900
+    },
+    {
+      "epoch": 7.693950531439357,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00018448793086321645,
+      "loss": 0.518,
+      "step": 154910
+    },
+    {
+      "epoch": 7.694447203734976,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00018444819707956692,
+      "loss": 0.4795,
+      "step": 154920
+    },
+    {
+      "epoch": 7.694943876030595,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00018440846329591736,
+      "loss": 0.5062,
+      "step": 154930
+    },
+    {
+      "epoch": 7.695440548326214,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00018436872951226784,
+      "loss": 0.4908,
+      "step": 154940
+    },
+    {
+      "epoch": 7.6959372206218335,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00018432899572861825,
+      "loss": 0.4745,
+      "step": 154950
+    },
+    {
+      "epoch": 7.696433892917453,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00018428926194496872,
+      "loss": 0.5064,
+      "step": 154960
+    },
+    {
+      "epoch": 7.696930565213073,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00018424952816131917,
+      "loss": 0.4763,
+      "step": 154970
+    },
+    {
+      "epoch": 7.697427237508692,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018420979437766964,
+      "loss": 0.5162,
+      "step": 154980
+    },
+    {
+      "epoch": 7.697923909804311,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018417006059402006,
+      "loss": 0.4906,
+      "step": 154990
+    },
+    {
+      "epoch": 7.6984205820999305,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00018413032681037053,
+      "loss": 0.4674,
+      "step": 155000
+    },
+    {
+      "epoch": 7.69891725439555,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00018409059302672097,
+      "loss": 0.5068,
+      "step": 155010
+    },
+    {
+      "epoch": 7.699413926691169,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00018405085924307144,
+      "loss": 0.4977,
+      "step": 155020
+    },
+    {
+      "epoch": 7.699910598986788,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00018401112545942186,
+      "loss": 0.4931,
+      "step": 155030
+    },
+    {
+      "epoch": 7.700407271282408,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00018397139167577233,
+      "loss": 0.5055,
+      "step": 155040
+    },
+    {
+      "epoch": 7.7009039435780275,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00018393165789212278,
+      "loss": 0.4862,
+      "step": 155050
+    },
+    {
+      "epoch": 7.701400615873647,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00018389192410847325,
+      "loss": 0.4901,
+      "step": 155060
+    },
+    {
+      "epoch": 7.701897288169266,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001838521903248237,
+      "loss": 0.4955,
+      "step": 155070
+    },
+    {
+      "epoch": 7.702393960464885,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018381245654117414,
+      "loss": 0.4724,
+      "step": 155080
+    },
+    {
+      "epoch": 7.7028906327605045,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0001837727227575246,
+      "loss": 0.4718,
+      "step": 155090
+    },
+    {
+      "epoch": 7.703387305056124,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00018373298897387505,
+      "loss": 0.484,
+      "step": 155100
+    },
+    {
+      "epoch": 7.703883977351744,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001836932551902255,
+      "loss": 0.4619,
+      "step": 155110
+    },
+    {
+      "epoch": 7.704380649647363,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018365352140657594,
+      "loss": 0.5065,
+      "step": 155120
+    },
+    {
+      "epoch": 7.704877321942982,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0001836137876229264,
+      "loss": 0.4808,
+      "step": 155130
+    },
+    {
+      "epoch": 7.7053739942386015,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00018357405383927686,
+      "loss": 0.4847,
+      "step": 155140
+    },
+    {
+      "epoch": 7.705870666534221,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001835343200556273,
+      "loss": 0.4682,
+      "step": 155150
+    },
+    {
+      "epoch": 7.70636733882984,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00018349458627197774,
+      "loss": 0.504,
+      "step": 155160
+    },
+    {
+      "epoch": 7.706864011125459,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018345485248832822,
+      "loss": 0.4837,
+      "step": 155170
+    },
+    {
+      "epoch": 7.707360683421078,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018341511870467866,
+      "loss": 0.4956,
+      "step": 155180
+    },
+    {
+      "epoch": 7.7078573557166985,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001833753849210291,
+      "loss": 0.5038,
+      "step": 155190
+    },
+    {
+      "epoch": 7.708354028012318,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.00018333565113737955,
+      "loss": 0.4812,
+      "step": 155200
+    },
+    {
+      "epoch": 7.708850700307937,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00018329591735373002,
+      "loss": 0.5041,
+      "step": 155210
+    },
+    {
+      "epoch": 7.709347372603556,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00018325618357008046,
+      "loss": 0.4985,
+      "step": 155220
+    },
+    {
+      "epoch": 7.709844044899175,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00018321644978643094,
+      "loss": 0.4971,
+      "step": 155230
+    },
+    {
+      "epoch": 7.710340717194795,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018317671600278138,
+      "loss": 0.481,
+      "step": 155240
+    },
+    {
+      "epoch": 7.710837389490414,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00018313698221913182,
+      "loss": 0.4365,
+      "step": 155250
+    },
+    {
+      "epoch": 7.711334061786033,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001830972484354823,
+      "loss": 0.4678,
+      "step": 155260
+    },
+    {
+      "epoch": 7.711830734081653,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00018305751465183274,
+      "loss": 0.466,
+      "step": 155270
+    },
+    {
+      "epoch": 7.712327406377272,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018301778086818318,
+      "loss": 0.4843,
+      "step": 155280
+    },
+    {
+      "epoch": 7.712824078672892,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00018297804708453363,
+      "loss": 0.4702,
+      "step": 155290
+    },
+    {
+      "epoch": 7.713320750968511,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001829383133008841,
+      "loss": 0.4972,
+      "step": 155300
+    },
+    {
+      "epoch": 7.71381742326413,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00018289857951723454,
+      "loss": 0.5,
+      "step": 155310
+    },
+    {
+      "epoch": 7.714314095559749,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.000182858845733585,
+      "loss": 0.5008,
+      "step": 155320
+    },
+    {
+      "epoch": 7.714810767855369,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00018281911194993543,
+      "loss": 0.4864,
+      "step": 155330
+    },
+    {
+      "epoch": 7.715307440150989,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001827793781662859,
+      "loss": 0.4841,
+      "step": 155340
+    },
+    {
+      "epoch": 7.715804112446608,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00018273964438263635,
+      "loss": 0.4886,
+      "step": 155350
+    },
+    {
+      "epoch": 7.716300784742227,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001826999105989868,
+      "loss": 0.5069,
+      "step": 155360
+    },
+    {
+      "epoch": 7.716797457037846,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00018266017681533724,
+      "loss": 0.4672,
+      "step": 155370
+    },
+    {
+      "epoch": 7.717294129333466,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001826204430316877,
+      "loss": 0.4969,
+      "step": 155380
+    },
+    {
+      "epoch": 7.717790801629085,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00018258070924803815,
+      "loss": 0.5053,
+      "step": 155390
+    },
+    {
+      "epoch": 7.718287473924704,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0001825409754643886,
+      "loss": 0.4804,
+      "step": 155400
+    },
+    {
+      "epoch": 7.718784146220324,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00018250124168073907,
+      "loss": 0.4871,
+      "step": 155410
+    },
+    {
+      "epoch": 7.719280818515943,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001824615078970895,
+      "loss": 0.5047,
+      "step": 155420
+    },
+    {
+      "epoch": 7.719777490811563,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00018242177411343996,
+      "loss": 0.4768,
+      "step": 155430
+    },
+    {
+      "epoch": 7.720274163107182,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0001823820403297904,
+      "loss": 0.4859,
+      "step": 155440
+    },
+    {
+      "epoch": 7.720770835402801,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00018234230654614087,
+      "loss": 0.4843,
+      "step": 155450
+    },
+    {
+      "epoch": 7.72126750769842,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00018230257276249132,
+      "loss": 0.4914,
+      "step": 155460
+    },
+    {
+      "epoch": 7.7217641799940395,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001822628389788418,
+      "loss": 0.4985,
+      "step": 155470
+    },
+    {
+      "epoch": 7.72226085228966,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001822231051951922,
+      "loss": 0.4893,
+      "step": 155480
+    },
+    {
+      "epoch": 7.722757524585279,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00018218337141154268,
+      "loss": 0.5087,
+      "step": 155490
+    },
+    {
+      "epoch": 7.723254196880898,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00018214363762789312,
+      "loss": 0.4684,
+      "step": 155500
+    },
+    {
+      "epoch": 7.723750869176517,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0001821039038442436,
+      "loss": 0.492,
+      "step": 155510
+    },
+    {
+      "epoch": 7.7242475414721365,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.000182064170060594,
+      "loss": 0.4771,
+      "step": 155520
+    },
+    {
+      "epoch": 7.724744213767756,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018202443627694448,
+      "loss": 0.4561,
+      "step": 155530
+    },
+    {
+      "epoch": 7.725240886063375,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00018198470249329492,
+      "loss": 0.4831,
+      "step": 155540
+    },
+    {
+      "epoch": 7.725737558358995,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001819449687096454,
+      "loss": 0.4935,
+      "step": 155550
+    },
+    {
+      "epoch": 7.726234230654614,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00018190523492599584,
+      "loss": 0.4966,
+      "step": 155560
+    },
+    {
+      "epoch": 7.726730902950234,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018186550114234628,
+      "loss": 0.48,
+      "step": 155570
+    },
+    {
+      "epoch": 7.727227575245853,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00018182576735869673,
+      "loss": 0.5067,
+      "step": 155580
+    },
+    {
+      "epoch": 7.727724247541472,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001817860335750472,
+      "loss": 0.4902,
+      "step": 155590
+    },
+    {
+      "epoch": 7.728220919837091,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00018174629979139764,
+      "loss": 0.4832,
+      "step": 155600
+    },
+    {
+      "epoch": 7.7287175921327105,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001817065660077481,
+      "loss": 0.5113,
+      "step": 155610
+    },
+    {
+      "epoch": 7.729214264428331,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00018166683222409856,
+      "loss": 0.4901,
+      "step": 155620
+    },
+    {
+      "epoch": 7.72971093672395,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.000181627098440449,
+      "loss": 0.4785,
+      "step": 155630
+    },
+    {
+      "epoch": 7.730207609019569,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018158736465679947,
+      "loss": 0.4777,
+      "step": 155640
+    },
+    {
+      "epoch": 7.730704281315188,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0001815476308731499,
+      "loss": 0.4751,
+      "step": 155650
+    },
+    {
+      "epoch": 7.7312009536108075,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00018150789708950036,
+      "loss": 0.4931,
+      "step": 155660
+    },
+    {
+      "epoch": 7.731697625906427,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001814681633058508,
+      "loss": 0.5069,
+      "step": 155670
+    },
+    {
+      "epoch": 7.732194298202046,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00018142842952220128,
+      "loss": 0.5336,
+      "step": 155680
+    },
+    {
+      "epoch": 7.732690970497666,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001813886957385517,
+      "loss": 0.4794,
+      "step": 155690
+    },
+    {
+      "epoch": 7.733187642793285,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00018134896195490217,
+      "loss": 0.4496,
+      "step": 155700
+    },
+    {
+      "epoch": 7.7336843150889045,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0001813092281712526,
+      "loss": 0.506,
+      "step": 155710
+    },
+    {
+      "epoch": 7.734180987384524,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00018126949438760308,
+      "loss": 0.4618,
+      "step": 155720
+    },
+    {
+      "epoch": 7.734677659680143,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001812297606039535,
+      "loss": 0.4909,
+      "step": 155730
+    },
+    {
+      "epoch": 7.735174331975762,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018119002682030397,
+      "loss": 0.5024,
+      "step": 155740
+    },
+    {
+      "epoch": 7.735671004271381,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00018115029303665442,
+      "loss": 0.5077,
+      "step": 155750
+    },
+    {
+      "epoch": 7.7361676765670015,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001811105592530049,
+      "loss": 0.4669,
+      "step": 155760
+    },
+    {
+      "epoch": 7.736664348862621,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018107082546935533,
+      "loss": 0.5059,
+      "step": 155770
+    },
+    {
+      "epoch": 7.73716102115824,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00018103109168570578,
+      "loss": 0.49,
+      "step": 155780
+    },
+    {
+      "epoch": 7.737657693453859,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00018099135790205625,
+      "loss": 0.5105,
+      "step": 155790
+    },
+    {
+      "epoch": 7.7381543657494785,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001809516241184067,
+      "loss": 0.4722,
+      "step": 155800
+    },
+    {
+      "epoch": 7.738651038045098,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00018091189033475714,
+      "loss": 0.484,
+      "step": 155810
+    },
+    {
+      "epoch": 7.739147710340717,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00018087215655110758,
+      "loss": 0.4742,
+      "step": 155820
+    },
+    {
+      "epoch": 7.739644382636337,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00018083242276745805,
+      "loss": 0.5169,
+      "step": 155830
+    },
+    {
+      "epoch": 7.740141054931956,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001807926889838085,
+      "loss": 0.462,
+      "step": 155840
+    },
+    {
+      "epoch": 7.7406377272275755,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00018075295520015894,
+      "loss": 0.4676,
+      "step": 155850
+    },
+    {
+      "epoch": 7.741134399523195,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00018071322141650938,
+      "loss": 0.5185,
+      "step": 155860
+    },
+    {
+      "epoch": 7.741631071818814,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00018067348763285986,
+      "loss": 0.4933,
+      "step": 155870
+    },
+    {
+      "epoch": 7.742127744114433,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001806337538492103,
+      "loss": 0.4851,
+      "step": 155880
+    },
+    {
+      "epoch": 7.742624416410052,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00018059402006556074,
+      "loss": 0.498,
+      "step": 155890
+    },
+    {
+      "epoch": 7.743121088705672,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001805542862819112,
+      "loss": 0.4972,
+      "step": 155900
+    },
+    {
+      "epoch": 7.743617761001292,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00018051455249826166,
+      "loss": 0.4893,
+      "step": 155910
+    },
+    {
+      "epoch": 7.744114433296911,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0001804748187146121,
+      "loss": 0.491,
+      "step": 155920
+    },
+    {
+      "epoch": 7.74461110559253,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00018043508493096255,
+      "loss": 0.5067,
+      "step": 155930
+    },
+    {
+      "epoch": 7.745107777888149,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00018039535114731302,
+      "loss": 0.4909,
+      "step": 155940
+    },
+    {
+      "epoch": 7.745604450183769,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00018035561736366346,
+      "loss": 0.5073,
+      "step": 155950
+    },
+    {
+      "epoch": 7.746101122479388,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.00018031588358001393,
+      "loss": 0.4952,
+      "step": 155960
+    },
+    {
+      "epoch": 7.746597794775007,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00018027614979636435,
+      "loss": 0.485,
+      "step": 155970
+    },
+    {
+      "epoch": 7.747094467070626,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00018023641601271482,
+      "loss": 0.4798,
+      "step": 155980
+    },
+    {
+      "epoch": 7.747591139366246,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00018019668222906527,
+      "loss": 0.4683,
+      "step": 155990
+    },
+    {
+      "epoch": 7.748087811661866,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018015694844541574,
+      "loss": 0.4895,
+      "step": 156000
+    },
+    {
+      "epoch": 7.748584483957485,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018011721466176616,
+      "loss": 0.4628,
+      "step": 156010
+    },
+    {
+      "epoch": 7.749081156253104,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00018007748087811663,
+      "loss": 0.4879,
+      "step": 156020
+    },
+    {
+      "epoch": 7.749577828548723,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00018003774709446707,
+      "loss": 0.4913,
+      "step": 156030
+    },
+    {
+      "epoch": 7.750074500844343,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00017999801331081754,
+      "loss": 0.4876,
+      "step": 156040
+    },
+    {
+      "epoch": 7.750571173139962,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00017995827952716796,
+      "loss": 0.5229,
+      "step": 156050
+    },
+    {
+      "epoch": 7.751067845435582,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00017991854574351843,
+      "loss": 0.4982,
+      "step": 156060
+    },
+    {
+      "epoch": 7.751564517731201,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017987881195986888,
+      "loss": 0.4882,
+      "step": 156070
+    },
+    {
+      "epoch": 7.75206119002682,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00017983907817621935,
+      "loss": 0.5045,
+      "step": 156080
+    },
+    {
+      "epoch": 7.75255786232244,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001797993443925698,
+      "loss": 0.456,
+      "step": 156090
+    },
+    {
+      "epoch": 7.753054534618059,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017975961060892024,
+      "loss": 0.5028,
+      "step": 156100
+    },
+    {
+      "epoch": 7.753551206913678,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001797198768252707,
+      "loss": 0.5156,
+      "step": 156110
+    },
+    {
+      "epoch": 7.754047879209297,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017968014304162115,
+      "loss": 0.4691,
+      "step": 156120
+    },
+    {
+      "epoch": 7.754544551504917,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00017964040925797162,
+      "loss": 0.4989,
+      "step": 156130
+    },
+    {
+      "epoch": 7.755041223800537,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00017960067547432204,
+      "loss": 0.4972,
+      "step": 156140
+    },
+    {
+      "epoch": 7.755537896096156,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001795609416906725,
+      "loss": 0.4985,
+      "step": 156150
+    },
+    {
+      "epoch": 7.756034568391775,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00017952120790702296,
+      "loss": 0.5022,
+      "step": 156160
+    },
+    {
+      "epoch": 7.756531240687394,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00017948147412337343,
+      "loss": 0.5003,
+      "step": 156170
+    },
+    {
+      "epoch": 7.7570279129830135,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00017944174033972384,
+      "loss": 0.5014,
+      "step": 156180
+    },
+    {
+      "epoch": 7.757524585278633,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00017940200655607432,
+      "loss": 0.4945,
+      "step": 156190
+    },
+    {
+      "epoch": 7.758021257574253,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00017936227277242476,
+      "loss": 0.5087,
+      "step": 156200
+    },
+    {
+      "epoch": 7.758517929869872,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00017932253898877523,
+      "loss": 0.4933,
+      "step": 156210
+    },
+    {
+      "epoch": 7.759014602165491,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00017928280520512565,
+      "loss": 0.5283,
+      "step": 156220
+    },
+    {
+      "epoch": 7.7595112744611106,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00017924307142147612,
+      "loss": 0.4793,
+      "step": 156230
+    },
+    {
+      "epoch": 7.76000794675673,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00017920333763782656,
+      "loss": 0.4575,
+      "step": 156240
+    },
+    {
+      "epoch": 7.760504619052349,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00017916360385417703,
+      "loss": 0.5082,
+      "step": 156250
+    },
+    {
+      "epoch": 7.761001291347968,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00017912387007052748,
+      "loss": 0.4818,
+      "step": 156260
+    },
+    {
+      "epoch": 7.761497963643588,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00017908413628687792,
+      "loss": 0.4817,
+      "step": 156270
+    },
+    {
+      "epoch": 7.761994635939208,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00017904440250322837,
+      "loss": 0.4939,
+      "step": 156280
+    },
+    {
+      "epoch": 7.762491308234827,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00017900466871957884,
+      "loss": 0.4987,
+      "step": 156290
+    },
+    {
+      "epoch": 7.762987980530446,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00017896493493592928,
+      "loss": 0.4516,
+      "step": 156300
+    },
+    {
+      "epoch": 7.763484652826065,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017892520115227973,
+      "loss": 0.4825,
+      "step": 156310
+    },
+    {
+      "epoch": 7.7639813251216845,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001788854673686302,
+      "loss": 0.5048,
+      "step": 156320
+    },
+    {
+      "epoch": 7.764477997417304,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00017884573358498064,
+      "loss": 0.5065,
+      "step": 156330
+    },
+    {
+      "epoch": 7.764974669712924,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001788059998013311,
+      "loss": 0.4905,
+      "step": 156340
+    },
+    {
+      "epoch": 7.765471342008543,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00017876626601768153,
+      "loss": 0.4914,
+      "step": 156350
+    },
+    {
+      "epoch": 7.765968014304162,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.000178726532234032,
+      "loss": 0.5184,
+      "step": 156360
+    },
+    {
+      "epoch": 7.7664646865997815,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017868679845038245,
+      "loss": 0.4882,
+      "step": 156370
+    },
+    {
+      "epoch": 7.766961358895401,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001786470646667329,
+      "loss": 0.501,
+      "step": 156380
+    },
+    {
+      "epoch": 7.76745803119102,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00017860733088308334,
+      "loss": 0.5198,
+      "step": 156390
+    },
+    {
+      "epoch": 7.767954703486639,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001785675970994338,
+      "loss": 0.4973,
+      "step": 156400
+    },
+    {
+      "epoch": 7.768451375782259,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00017852786331578425,
+      "loss": 0.5214,
+      "step": 156410
+    },
+    {
+      "epoch": 7.7689480480778785,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0001784881295321347,
+      "loss": 0.4826,
+      "step": 156420
+    },
+    {
+      "epoch": 7.769444720373498,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00017844839574848514,
+      "loss": 0.4648,
+      "step": 156430
+    },
+    {
+      "epoch": 7.769941392669117,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001784086619648356,
+      "loss": 0.5546,
+      "step": 156440
+    },
+    {
+      "epoch": 7.770438064964736,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017836892818118606,
+      "loss": 0.502,
+      "step": 156450
+    },
+    {
+      "epoch": 7.770934737260355,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001783291943975365,
+      "loss": 0.4913,
+      "step": 156460
+    },
+    {
+      "epoch": 7.771431409555975,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00017828946061388697,
+      "loss": 0.4959,
+      "step": 156470
+    },
+    {
+      "epoch": 7.771928081851595,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00017824972683023742,
+      "loss": 0.4986,
+      "step": 156480
+    },
+    {
+      "epoch": 7.772424754147214,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001782099930465879,
+      "loss": 0.4853,
+      "step": 156490
+    },
+    {
+      "epoch": 7.772921426442833,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017817025926293833,
+      "loss": 0.4826,
+      "step": 156500
+    },
+    {
+      "epoch": 7.7734180987384525,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00017813052547928878,
+      "loss": 0.5112,
+      "step": 156510
+    },
+    {
+      "epoch": 7.773914771034072,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00017809079169563922,
+      "loss": 0.4555,
+      "step": 156520
+    },
+    {
+      "epoch": 7.774411443329691,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001780510579119897,
+      "loss": 0.5022,
+      "step": 156530
+    },
+    {
+      "epoch": 7.77490811562531,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00017801132412834014,
+      "loss": 0.4864,
+      "step": 156540
+    },
+    {
+      "epoch": 7.77540478792093,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00017797159034469058,
+      "loss": 0.5203,
+      "step": 156550
+    },
+    {
+      "epoch": 7.7759014602165495,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00017793185656104102,
+      "loss": 0.4664,
+      "step": 156560
+    },
+    {
+      "epoch": 7.776398132512169,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0001778921227773915,
+      "loss": 0.4759,
+      "step": 156570
+    },
+    {
+      "epoch": 7.776894804807788,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00017785238899374194,
+      "loss": 0.5151,
+      "step": 156580
+    },
+    {
+      "epoch": 7.777391477103407,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017781265521009238,
+      "loss": 0.5092,
+      "step": 156590
+    },
+    {
+      "epoch": 7.777888149399026,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00017777292142644283,
+      "loss": 0.4774,
+      "step": 156600
+    },
+    {
+      "epoch": 7.778384821694646,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0001777331876427933,
+      "loss": 0.481,
+      "step": 156610
+    },
+    {
+      "epoch": 7.778881493990265,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00017769345385914374,
+      "loss": 0.4945,
+      "step": 156620
+    },
+    {
+      "epoch": 7.779378166285884,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001776537200754942,
+      "loss": 0.4764,
+      "step": 156630
+    },
+    {
+      "epoch": 7.779874838581504,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00017761398629184466,
+      "loss": 0.4602,
+      "step": 156640
+    },
+    {
+      "epoch": 7.780371510877123,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001775742525081951,
+      "loss": 0.5032,
+      "step": 156650
+    },
+    {
+      "epoch": 7.780868183172743,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017753451872454557,
+      "loss": 0.4916,
+      "step": 156660
+    },
+    {
+      "epoch": 7.781364855468362,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.000177494784940896,
+      "loss": 0.4989,
+      "step": 156670
+    },
+    {
+      "epoch": 7.781861527763981,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00017745505115724646,
+      "loss": 0.5072,
+      "step": 156680
+    },
+    {
+      "epoch": 7.7823582000596,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001774153173735969,
+      "loss": 0.4669,
+      "step": 156690
+    },
+    {
+      "epoch": 7.7828548723552196,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017737558358994738,
+      "loss": 0.4916,
+      "step": 156700
+    },
+    {
+      "epoch": 7.78335154465084,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001773358498062978,
+      "loss": 0.5019,
+      "step": 156710
+    },
+    {
+      "epoch": 7.783848216946459,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00017729611602264827,
+      "loss": 0.5173,
+      "step": 156720
+    },
+    {
+      "epoch": 7.784344889242078,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0001772563822389987,
+      "loss": 0.4668,
+      "step": 156730
+    },
+    {
+      "epoch": 7.784841561537697,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00017721664845534918,
+      "loss": 0.4851,
+      "step": 156740
+    },
+    {
+      "epoch": 7.785338233833317,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001771769146716996,
+      "loss": 0.4818,
+      "step": 156750
+    },
+    {
+      "epoch": 7.785834906128936,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00017713718088805007,
+      "loss": 0.5166,
+      "step": 156760
+    },
+    {
+      "epoch": 7.786331578424555,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00017709744710440052,
+      "loss": 0.4916,
+      "step": 156770
+    },
+    {
+      "epoch": 7.786828250720175,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.000177057713320751,
+      "loss": 0.483,
+      "step": 156780
+    },
+    {
+      "epoch": 7.787324923015794,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00017701797953710143,
+      "loss": 0.4881,
+      "step": 156790
+    },
+    {
+      "epoch": 7.787821595311414,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017697824575345188,
+      "loss": 0.493,
+      "step": 156800
+    },
+    {
+      "epoch": 7.788318267607033,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00017693851196980235,
+      "loss": 0.4836,
+      "step": 156810
+    },
+    {
+      "epoch": 7.788814939902652,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001768987781861528,
+      "loss": 0.4688,
+      "step": 156820
+    },
+    {
+      "epoch": 7.789311612198271,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017685904440250324,
+      "loss": 0.5103,
+      "step": 156830
+    },
+    {
+      "epoch": 7.7898082844938905,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00017681931061885368,
+      "loss": 0.5353,
+      "step": 156840
+    },
+    {
+      "epoch": 7.790304956789511,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017677957683520415,
+      "loss": 0.493,
+      "step": 156850
+    },
+    {
+      "epoch": 7.79080162908513,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0001767398430515546,
+      "loss": 0.488,
+      "step": 156860
+    },
+    {
+      "epoch": 7.791298301380749,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00017670010926790504,
+      "loss": 0.4948,
+      "step": 156870
+    },
+    {
+      "epoch": 7.791794973676368,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00017666037548425548,
+      "loss": 0.5081,
+      "step": 156880
+    },
+    {
+      "epoch": 7.7922916459719875,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00017662064170060595,
+      "loss": 0.483,
+      "step": 156890
+    },
+    {
+      "epoch": 7.792788318267607,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0001765809079169564,
+      "loss": 0.5121,
+      "step": 156900
+    },
+    {
+      "epoch": 7.793284990563226,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00017654117413330687,
+      "loss": 0.4898,
+      "step": 156910
+    },
+    {
+      "epoch": 7.793781662858846,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001765014403496573,
+      "loss": 0.4897,
+      "step": 156920
+    },
+    {
+      "epoch": 7.794278335154465,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00017646170656600776,
+      "loss": 0.4827,
+      "step": 156930
+    },
+    {
+      "epoch": 7.794775007450085,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001764219727823582,
+      "loss": 0.4754,
+      "step": 156940
+    },
+    {
+      "epoch": 7.795271679745704,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00017638223899870867,
+      "loss": 0.5168,
+      "step": 156950
+    },
+    {
+      "epoch": 7.795768352041323,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00017634250521505912,
+      "loss": 0.4789,
+      "step": 156960
+    },
+    {
+      "epoch": 7.796265024336942,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017630277143140956,
+      "loss": 0.4859,
+      "step": 156970
+    },
+    {
+      "epoch": 7.7967616966325615,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00017626303764776003,
+      "loss": 0.5107,
+      "step": 156980
+    },
+    {
+      "epoch": 7.797258368928182,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017622330386411048,
+      "loss": 0.4651,
+      "step": 156990
+    },
+    {
+      "epoch": 7.797755041223801,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00017618357008046092,
+      "loss": 0.468,
+      "step": 157000
+    },
+    {
+      "epoch": 7.79825171351942,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017614383629681137,
+      "loss": 0.4886,
+      "step": 157010
+    },
+    {
+      "epoch": 7.798748385815039,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00017610410251316184,
+      "loss": 0.4947,
+      "step": 157020
+    },
+    {
+      "epoch": 7.7992450581106585,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00017606436872951228,
+      "loss": 0.5017,
+      "step": 157030
+    },
+    {
+      "epoch": 7.799741730406278,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00017602463494586273,
+      "loss": 0.4941,
+      "step": 157040
+    },
+    {
+      "epoch": 7.800238402701897,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00017598490116221317,
+      "loss": 0.4735,
+      "step": 157050
+    },
+    {
+      "epoch": 7.800735074997517,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00017594516737856364,
+      "loss": 0.4988,
+      "step": 157060
+    },
+    {
+      "epoch": 7.801231747293136,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001759054335949141,
+      "loss": 0.4839,
+      "step": 157070
+    },
+    {
+      "epoch": 7.8017284195887555,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00017586569981126453,
+      "loss": 0.4763,
+      "step": 157080
+    },
+    {
+      "epoch": 7.802225091884375,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017582596602761498,
+      "loss": 0.5064,
+      "step": 157090
+    },
+    {
+      "epoch": 7.802721764179994,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00017578623224396545,
+      "loss": 0.4735,
+      "step": 157100
+    },
+    {
+      "epoch": 7.803218436475613,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001757464984603159,
+      "loss": 0.483,
+      "step": 157110
+    },
+    {
+      "epoch": 7.803715108771232,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00017570676467666634,
+      "loss": 0.4756,
+      "step": 157120
+    },
+    {
+      "epoch": 7.8042117810668525,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00017566703089301678,
+      "loss": 0.4639,
+      "step": 157130
+    },
+    {
+      "epoch": 7.804708453362472,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00017562729710936725,
+      "loss": 0.4743,
+      "step": 157140
+    },
+    {
+      "epoch": 7.805205125658091,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001755875633257177,
+      "loss": 0.5498,
+      "step": 157150
+    },
+    {
+      "epoch": 7.80570179795371,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00017554782954206814,
+      "loss": 0.4827,
+      "step": 157160
+    },
+    {
+      "epoch": 7.8061984702493294,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001755080957584186,
+      "loss": 0.4982,
+      "step": 157170
+    },
+    {
+      "epoch": 7.806695142544949,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00017546836197476905,
+      "loss": 0.5006,
+      "step": 157180
+    },
+    {
+      "epoch": 7.807191814840568,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00017542862819111953,
+      "loss": 0.4829,
+      "step": 157190
+    },
+    {
+      "epoch": 7.807688487136188,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00017538889440746994,
+      "loss": 0.4748,
+      "step": 157200
+    },
+    {
+      "epoch": 7.808185159431807,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00017534916062382041,
+      "loss": 0.4986,
+      "step": 157210
+    },
+    {
+      "epoch": 7.8086818317274265,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017530942684017086,
+      "loss": 0.4668,
+      "step": 157220
+    },
+    {
+      "epoch": 7.809178504023046,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00017526969305652133,
+      "loss": 0.484,
+      "step": 157230
+    },
+    {
+      "epoch": 7.809675176318665,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00017522995927287175,
+      "loss": 0.4552,
+      "step": 157240
+    },
+    {
+      "epoch": 7.810171848614284,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00017519022548922222,
+      "loss": 0.4988,
+      "step": 157250
+    },
+    {
+      "epoch": 7.810668520909903,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017515049170557266,
+      "loss": 0.4594,
+      "step": 157260
+    },
+    {
+      "epoch": 7.8111651932055235,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00017511075792192313,
+      "loss": 0.4824,
+      "step": 157270
+    },
+    {
+      "epoch": 7.811661865501143,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017507102413827355,
+      "loss": 0.5047,
+      "step": 157280
+    },
+    {
+      "epoch": 7.812158537796762,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00017503129035462402,
+      "loss": 0.5078,
+      "step": 157290
+    },
+    {
+      "epoch": 7.812655210092381,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00017499155657097447,
+      "loss": 0.4975,
+      "step": 157300
+    },
+    {
+      "epoch": 7.813151882388,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00017495182278732494,
+      "loss": 0.4623,
+      "step": 157310
+    },
+    {
+      "epoch": 7.81364855468362,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00017491208900367538,
+      "loss": 0.4954,
+      "step": 157320
+    },
+    {
+      "epoch": 7.814145226979239,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00017487235522002583,
+      "loss": 0.4881,
+      "step": 157330
+    },
+    {
+      "epoch": 7.814641899274858,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001748326214363763,
+      "loss": 0.4988,
+      "step": 157340
+    },
+    {
+      "epoch": 7.815138571570477,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00017479288765272674,
+      "loss": 0.4848,
+      "step": 157350
+    },
+    {
+      "epoch": 7.815635243866097,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00017475315386907721,
+      "loss": 0.4845,
+      "step": 157360
+    },
+    {
+      "epoch": 7.816131916161717,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017471342008542763,
+      "loss": 0.503,
+      "step": 157370
+    },
+    {
+      "epoch": 7.816628588457336,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001746736863017781,
+      "loss": 0.478,
+      "step": 157380
+    },
+    {
+      "epoch": 7.817125260752955,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00017463395251812855,
+      "loss": 0.4855,
+      "step": 157390
+    },
+    {
+      "epoch": 7.817621933048574,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017459421873447902,
+      "loss": 0.4919,
+      "step": 157400
+    },
+    {
+      "epoch": 7.818118605344194,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00017455448495082944,
+      "loss": 0.493,
+      "step": 157410
+    },
+    {
+      "epoch": 7.818615277639813,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001745147511671799,
+      "loss": 0.4962,
+      "step": 157420
+    },
+    {
+      "epoch": 7.819111949935433,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017447501738353035,
+      "loss": 0.4884,
+      "step": 157430
+    },
+    {
+      "epoch": 7.819608622231052,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017443528359988082,
+      "loss": 0.5151,
+      "step": 157440
+    },
+    {
+      "epoch": 7.820105294526671,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00017439554981623124,
+      "loss": 0.4866,
+      "step": 157450
+    },
+    {
+      "epoch": 7.820601966822291,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001743558160325817,
+      "loss": 0.5049,
+      "step": 157460
+    },
+    {
+      "epoch": 7.82109863911791,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00017431608224893216,
+      "loss": 0.4865,
+      "step": 157470
+    },
+    {
+      "epoch": 7.821595311413529,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017427634846528263,
+      "loss": 0.5304,
+      "step": 157480
+    },
+    {
+      "epoch": 7.822091983709148,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00017423661468163307,
+      "loss": 0.4809,
+      "step": 157490
+    },
+    {
+      "epoch": 7.822588656004768,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00017419688089798351,
+      "loss": 0.5039,
+      "step": 157500
+    },
+    {
+      "epoch": 7.823085328300388,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00017415714711433399,
+      "loss": 0.4733,
+      "step": 157510
+    },
+    {
+      "epoch": 7.823582000596007,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00017411741333068443,
+      "loss": 0.4945,
+      "step": 157520
+    },
+    {
+      "epoch": 7.824078672891626,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00017407767954703487,
+      "loss": 0.4967,
+      "step": 157530
+    },
+    {
+      "epoch": 7.824575345187245,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017403794576338532,
+      "loss": 0.4992,
+      "step": 157540
+    },
+    {
+      "epoch": 7.8250720174828645,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001739982119797358,
+      "loss": 0.4455,
+      "step": 157550
+    },
+    {
+      "epoch": 7.825568689778484,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00017395847819608623,
+      "loss": 0.4939,
+      "step": 157560
+    },
+    {
+      "epoch": 7.826065362074104,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00017391874441243668,
+      "loss": 0.4764,
+      "step": 157570
+    },
+    {
+      "epoch": 7.826562034369723,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00017387901062878712,
+      "loss": 0.4906,
+      "step": 157580
+    },
+    {
+      "epoch": 7.827058706665342,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001738392768451376,
+      "loss": 0.5121,
+      "step": 157590
+    },
+    {
+      "epoch": 7.8275553789609615,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017379954306148804,
+      "loss": 0.4899,
+      "step": 157600
+    },
+    {
+      "epoch": 7.828052051256581,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017375980927783848,
+      "loss": 0.5032,
+      "step": 157610
+    },
+    {
+      "epoch": 7.8285487235522,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00017372007549418893,
+      "loss": 0.5003,
+      "step": 157620
+    },
+    {
+      "epoch": 7.829045395847819,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001736803417105394,
+      "loss": 0.5219,
+      "step": 157630
+    },
+    {
+      "epoch": 7.829542068143439,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00017364060792688984,
+      "loss": 0.4936,
+      "step": 157640
+    },
+    {
+      "epoch": 7.830038740439059,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0001736008741432403,
+      "loss": 0.4803,
+      "step": 157650
+    },
+    {
+      "epoch": 7.830535412734678,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017356114035959076,
+      "loss": 0.5206,
+      "step": 157660
+    },
+    {
+      "epoch": 7.831032085030297,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001735214065759412,
+      "loss": 0.5155,
+      "step": 157670
+    },
+    {
+      "epoch": 7.831528757325916,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00017348167279229167,
+      "loss": 0.4948,
+      "step": 157680
+    },
+    {
+      "epoch": 7.8320254296215355,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001734419390086421,
+      "loss": 0.4792,
+      "step": 157690
+    },
+    {
+      "epoch": 7.832522101917155,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00017340220522499256,
+      "loss": 0.4773,
+      "step": 157700
+    },
+    {
+      "epoch": 7.833018774212775,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.000173362471441343,
+      "loss": 0.4869,
+      "step": 157710
+    },
+    {
+      "epoch": 7.833515446508394,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00017332273765769348,
+      "loss": 0.4808,
+      "step": 157720
+    },
+    {
+      "epoch": 7.834012118804013,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001732830038740439,
+      "loss": 0.4866,
+      "step": 157730
+    },
+    {
+      "epoch": 7.8345087910996325,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00017324327009039437,
+      "loss": 0.4477,
+      "step": 157740
+    },
+    {
+      "epoch": 7.835005463395252,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001732035363067448,
+      "loss": 0.483,
+      "step": 157750
+    },
+    {
+      "epoch": 7.835502135690871,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00017316380252309528,
+      "loss": 0.5277,
+      "step": 157760
+    },
+    {
+      "epoch": 7.83599880798649,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00017312406873944573,
+      "loss": 0.4935,
+      "step": 157770
+    },
+    {
+      "epoch": 7.83649548028211,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00017308433495579617,
+      "loss": 0.4675,
+      "step": 157780
+    },
+    {
+      "epoch": 7.8369921525777295,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00017304460117214662,
+      "loss": 0.4856,
+      "step": 157790
+    },
+    {
+      "epoch": 7.837488824873349,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00017300486738849709,
+      "loss": 0.4605,
+      "step": 157800
+    },
+    {
+      "epoch": 7.837985497168968,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00017296513360484753,
+      "loss": 0.4991,
+      "step": 157810
+    },
+    {
+      "epoch": 7.838482169464587,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00017292539982119797,
+      "loss": 0.5,
+      "step": 157820
+    },
+    {
+      "epoch": 7.838978841760206,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00017288566603754845,
+      "loss": 0.4949,
+      "step": 157830
+    },
+    {
+      "epoch": 7.839475514055826,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001728459322538989,
+      "loss": 0.4982,
+      "step": 157840
+    },
+    {
+      "epoch": 7.839972186351446,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00017280619847024933,
+      "loss": 0.4761,
+      "step": 157850
+    },
+    {
+      "epoch": 7.840468858647065,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00017276646468659978,
+      "loss": 0.4898,
+      "step": 157860
+    },
+    {
+      "epoch": 7.840965530942684,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00017272673090295025,
+      "loss": 0.5283,
+      "step": 157870
+    },
+    {
+      "epoch": 7.8414622032383035,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0001726869971193007,
+      "loss": 0.4675,
+      "step": 157880
+    },
+    {
+      "epoch": 7.841958875533923,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00017264726333565117,
+      "loss": 0.4695,
+      "step": 157890
+    },
+    {
+      "epoch": 7.842455547829542,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017260752955200158,
+      "loss": 0.4885,
+      "step": 157900
+    },
+    {
+      "epoch": 7.842952220125161,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00017256779576835205,
+      "loss": 0.4774,
+      "step": 157910
+    },
+    {
+      "epoch": 7.843448892420781,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001725280619847025,
+      "loss": 0.489,
+      "step": 157920
+    },
+    {
+      "epoch": 7.8439455647164005,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017248832820105297,
+      "loss": 0.4955,
+      "step": 157930
+    },
+    {
+      "epoch": 7.84444223701202,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001724485944174034,
+      "loss": 0.4899,
+      "step": 157940
+    },
+    {
+      "epoch": 7.844938909307639,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00017240886063375386,
+      "loss": 0.5077,
+      "step": 157950
+    },
+    {
+      "epoch": 7.845435581603258,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0001723691268501043,
+      "loss": 0.5141,
+      "step": 157960
+    },
+    {
+      "epoch": 7.845932253898877,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00017232939306645477,
+      "loss": 0.486,
+      "step": 157970
+    },
+    {
+      "epoch": 7.846428926194497,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001722896592828052,
+      "loss": 0.4688,
+      "step": 157980
+    },
+    {
+      "epoch": 7.846925598490116,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00017224992549915566,
+      "loss": 0.4759,
+      "step": 157990
+    },
+    {
+      "epoch": 7.847422270785736,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001722101917155061,
+      "loss": 0.5069,
+      "step": 158000
+    },
+    {
+      "epoch": 7.847918943081355,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00017217045793185658,
+      "loss": 0.4946,
+      "step": 158010
+    },
+    {
+      "epoch": 7.848415615376974,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00017213072414820702,
+      "loss": 0.515,
+      "step": 158020
+    },
+    {
+      "epoch": 7.848912287672594,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00017209099036455747,
+      "loss": 0.4999,
+      "step": 158030
+    },
+    {
+      "epoch": 7.849408959968213,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00017205125658090794,
+      "loss": 0.4776,
+      "step": 158040
+    },
+    {
+      "epoch": 7.849905632263832,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00017201152279725838,
+      "loss": 0.487,
+      "step": 158050
+    },
+    {
+      "epoch": 7.850402304559451,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00017197178901360883,
+      "loss": 0.5035,
+      "step": 158060
+    },
+    {
+      "epoch": 7.8508989768550705,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00017193205522995927,
+      "loss": 0.488,
+      "step": 158070
+    },
+    {
+      "epoch": 7.851395649150691,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017189232144630974,
+      "loss": 0.5068,
+      "step": 158080
+    },
+    {
+      "epoch": 7.85189232144631,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0001718525876626602,
+      "loss": 0.4852,
+      "step": 158090
+    },
+    {
+      "epoch": 7.852388993741929,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00017181285387901063,
+      "loss": 0.5238,
+      "step": 158100
+    },
+    {
+      "epoch": 7.852885666037548,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00017177312009536107,
+      "loss": 0.498,
+      "step": 158110
+    },
+    {
+      "epoch": 7.853382338333168,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00017173338631171155,
+      "loss": 0.485,
+      "step": 158120
+    },
+    {
+      "epoch": 7.853879010628787,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.000171693652528062,
+      "loss": 0.4411,
+      "step": 158130
+    },
+    {
+      "epoch": 7.854375682924406,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00017165391874441243,
+      "loss": 0.5216,
+      "step": 158140
+    },
+    {
+      "epoch": 7.854872355220026,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00017161418496076288,
+      "loss": 0.494,
+      "step": 158150
+    },
+    {
+      "epoch": 7.855369027515645,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00017157445117711335,
+      "loss": 0.5004,
+      "step": 158160
+    },
+    {
+      "epoch": 7.855865699811265,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001715347173934638,
+      "loss": 0.4854,
+      "step": 158170
+    },
+    {
+      "epoch": 7.856362372106884,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00017149498360981427,
+      "loss": 0.467,
+      "step": 158180
+    },
+    {
+      "epoch": 7.856859044402503,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001714552498261647,
+      "loss": 0.4838,
+      "step": 158190
+    },
+    {
+      "epoch": 7.857355716698122,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00017141551604251515,
+      "loss": 0.5075,
+      "step": 158200
+    },
+    {
+      "epoch": 7.8578523889937415,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00017137578225886563,
+      "loss": 0.4666,
+      "step": 158210
+    },
+    {
+      "epoch": 7.858349061289362,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00017133604847521607,
+      "loss": 0.495,
+      "step": 158220
+    },
+    {
+      "epoch": 7.858845733584981,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00017129631469156651,
+      "loss": 0.4623,
+      "step": 158230
+    },
+    {
+      "epoch": 7.8593424058806,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00017125658090791696,
+      "loss": 0.4929,
+      "step": 158240
+    },
+    {
+      "epoch": 7.859839078176219,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00017121684712426743,
+      "loss": 0.4998,
+      "step": 158250
+    },
+    {
+      "epoch": 7.8603357504718385,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00017117711334061787,
+      "loss": 0.465,
+      "step": 158260
+    },
+    {
+      "epoch": 7.860832422767458,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017113737955696832,
+      "loss": 0.5087,
+      "step": 158270
+    },
+    {
+      "epoch": 7.861329095063077,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00017109764577331876,
+      "loss": 0.4851,
+      "step": 158280
+    },
+    {
+      "epoch": 7.861825767358697,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00017105791198966923,
+      "loss": 0.4744,
+      "step": 158290
+    },
+    {
+      "epoch": 7.862322439654316,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00017101817820601968,
+      "loss": 0.505,
+      "step": 158300
+    },
+    {
+      "epoch": 7.8628191119499355,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00017097844442237012,
+      "loss": 0.4845,
+      "step": 158310
+    },
+    {
+      "epoch": 7.863315784245555,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00017093871063872057,
+      "loss": 0.5029,
+      "step": 158320
+    },
+    {
+      "epoch": 7.863812456541174,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00017089897685507104,
+      "loss": 0.4545,
+      "step": 158330
+    },
+    {
+      "epoch": 7.864309128836793,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00017085924307142148,
+      "loss": 0.4649,
+      "step": 158340
+    },
+    {
+      "epoch": 7.8648058011324125,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00017081950928777193,
+      "loss": 0.5021,
+      "step": 158350
+    },
+    {
+      "epoch": 7.865302473428033,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001707797755041224,
+      "loss": 0.4561,
+      "step": 158360
+    },
+    {
+      "epoch": 7.865799145723652,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00017074004172047284,
+      "loss": 0.5059,
+      "step": 158370
+    },
+    {
+      "epoch": 7.866295818019271,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00017070030793682331,
+      "loss": 0.5086,
+      "step": 158380
+    },
+    {
+      "epoch": 7.86679249031489,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00017066057415317373,
+      "loss": 0.4865,
+      "step": 158390
+    },
+    {
+      "epoch": 7.8672891626105095,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001706208403695242,
+      "loss": 0.4936,
+      "step": 158400
+    },
+    {
+      "epoch": 7.867785834906129,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00017058110658587465,
+      "loss": 0.4558,
+      "step": 158410
+    },
+    {
+      "epoch": 7.868282507201748,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00017054137280222512,
+      "loss": 0.5057,
+      "step": 158420
+    },
+    {
+      "epoch": 7.868779179497368,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00017050163901857553,
+      "loss": 0.4906,
+      "step": 158430
+    },
+    {
+      "epoch": 7.869275851792987,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.000170461905234926,
+      "loss": 0.5208,
+      "step": 158440
+    },
+    {
+      "epoch": 7.8697725240886065,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017042217145127645,
+      "loss": 0.4733,
+      "step": 158450
+    },
+    {
+      "epoch": 7.870269196384226,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00017038243766762692,
+      "loss": 0.4874,
+      "step": 158460
+    },
+    {
+      "epoch": 7.870765868679845,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00017034270388397734,
+      "loss": 0.4597,
+      "step": 158470
+    },
+    {
+      "epoch": 7.871262540975464,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001703029701003278,
+      "loss": 0.5403,
+      "step": 158480
+    },
+    {
+      "epoch": 7.871759213271083,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00017026323631667825,
+      "loss": 0.525,
+      "step": 158490
+    },
+    {
+      "epoch": 7.8722558855667035,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00017022350253302873,
+      "loss": 0.4927,
+      "step": 158500
+    },
+    {
+      "epoch": 7.872752557862323,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00017018376874937917,
+      "loss": 0.5222,
+      "step": 158510
+    },
+    {
+      "epoch": 7.873249230157942,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00017014403496572961,
+      "loss": 0.4882,
+      "step": 158520
+    },
+    {
+      "epoch": 7.873745902453561,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00017010430118208009,
+      "loss": 0.4862,
+      "step": 158530
+    },
+    {
+      "epoch": 7.87424257474918,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00017006456739843053,
+      "loss": 0.4564,
+      "step": 158540
+    },
+    {
+      "epoch": 7.8747392470448,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017002483361478097,
+      "loss": 0.5056,
+      "step": 158550
+    },
+    {
+      "epoch": 7.875235919340419,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.00016998509983113142,
+      "loss": 0.49,
+      "step": 158560
+    },
+    {
+      "epoch": 7.875732591636039,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001699453660474819,
+      "loss": 0.4968,
+      "step": 158570
+    },
+    {
+      "epoch": 7.876229263931658,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00016990563226383233,
+      "loss": 0.4868,
+      "step": 158580
+    },
+    {
+      "epoch": 7.8767259362272775,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001698658984801828,
+      "loss": 0.4979,
+      "step": 158590
+    },
+    {
+      "epoch": 7.877222608522897,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00016982616469653322,
+      "loss": 0.4879,
+      "step": 158600
+    },
+    {
+      "epoch": 7.877719280818516,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001697864309128837,
+      "loss": 0.5487,
+      "step": 158610
+    },
+    {
+      "epoch": 7.878215953114135,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00016974669712923414,
+      "loss": 0.4763,
+      "step": 158620
+    },
+    {
+      "epoch": 7.878712625409754,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001697069633455846,
+      "loss": 0.4907,
+      "step": 158630
+    },
+    {
+      "epoch": 7.8792092977053745,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016966722956193503,
+      "loss": 0.4819,
+      "step": 158640
+    },
+    {
+      "epoch": 7.879705970000994,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001696274957782855,
+      "loss": 0.4651,
+      "step": 158650
+    },
+    {
+      "epoch": 7.880202642296613,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00016958776199463594,
+      "loss": 0.5158,
+      "step": 158660
+    },
+    {
+      "epoch": 7.880699314592232,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00016954802821098641,
+      "loss": 0.4985,
+      "step": 158670
+    },
+    {
+      "epoch": 7.881195986887851,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00016950829442733683,
+      "loss": 0.4626,
+      "step": 158680
+    },
+    {
+      "epoch": 7.881692659183471,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001694685606436873,
+      "loss": 0.4984,
+      "step": 158690
+    },
+    {
+      "epoch": 7.88218933147909,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00016942882686003775,
+      "loss": 0.4647,
+      "step": 158700
+    },
+    {
+      "epoch": 7.882686003774709,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00016938909307638822,
+      "loss": 0.4911,
+      "step": 158710
+    },
+    {
+      "epoch": 7.883182676070328,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00016934935929273866,
+      "loss": 0.5046,
+      "step": 158720
+    },
+    {
+      "epoch": 7.883679348365948,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001693096255090891,
+      "loss": 0.4652,
+      "step": 158730
+    },
+    {
+      "epoch": 7.884176020661568,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00016926989172543958,
+      "loss": 0.5216,
+      "step": 158740
+    },
+    {
+      "epoch": 7.884672692957187,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00016923015794179002,
+      "loss": 0.4934,
+      "step": 158750
+    },
+    {
+      "epoch": 7.885169365252806,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00016919042415814047,
+      "loss": 0.5015,
+      "step": 158760
+    },
+    {
+      "epoch": 7.885666037548425,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001691506903744909,
+      "loss": 0.5578,
+      "step": 158770
+    },
+    {
+      "epoch": 7.8861627098440445,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00016911095659084138,
+      "loss": 0.4995,
+      "step": 158780
+    },
+    {
+      "epoch": 7.886659382139664,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00016907122280719183,
+      "loss": 0.5216,
+      "step": 158790
+    },
+    {
+      "epoch": 7.887156054435284,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00016903148902354227,
+      "loss": 0.4803,
+      "step": 158800
+    },
+    {
+      "epoch": 7.887652726730903,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00016899175523989271,
+      "loss": 0.4816,
+      "step": 158810
+    },
+    {
+      "epoch": 7.888149399026522,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00016895202145624319,
+      "loss": 0.4879,
+      "step": 158820
+    },
+    {
+      "epoch": 7.888646071322142,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00016891228767259363,
+      "loss": 0.4714,
+      "step": 158830
+    },
+    {
+      "epoch": 7.889142743617761,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00016887255388894407,
+      "loss": 0.4949,
+      "step": 158840
+    },
+    {
+      "epoch": 7.88963941591338,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00016883282010529452,
+      "loss": 0.5038,
+      "step": 158850
+    },
+    {
+      "epoch": 7.890136088208999,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.000168793086321645,
+      "loss": 0.4899,
+      "step": 158860
+    },
+    {
+      "epoch": 7.890632760504619,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00016875335253799543,
+      "loss": 0.5325,
+      "step": 158870
+    },
+    {
+      "epoch": 7.891129432800239,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00016871361875434588,
+      "loss": 0.4795,
+      "step": 158880
+    },
+    {
+      "epoch": 7.891626105095858,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00016867388497069635,
+      "loss": 0.5102,
+      "step": 158890
+    },
+    {
+      "epoch": 7.892122777391477,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0001686341511870468,
+      "loss": 0.4802,
+      "step": 158900
+    },
+    {
+      "epoch": 7.892619449687096,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00016859441740339727,
+      "loss": 0.4854,
+      "step": 158910
+    },
+    {
+      "epoch": 7.8931161219827155,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00016855468361974768,
+      "loss": 0.4677,
+      "step": 158920
+    },
+    {
+      "epoch": 7.893612794278335,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00016851494983609815,
+      "loss": 0.4833,
+      "step": 158930
+    },
+    {
+      "epoch": 7.894109466573955,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001684752160524486,
+      "loss": 0.4998,
+      "step": 158940
+    },
+    {
+      "epoch": 7.894606138869574,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00016843548226879907,
+      "loss": 0.4769,
+      "step": 158950
+    },
+    {
+      "epoch": 7.895102811165193,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0001683957484851495,
+      "loss": 0.4937,
+      "step": 158960
+    },
+    {
+      "epoch": 7.8955994834608125,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00016835601470149996,
+      "loss": 0.487,
+      "step": 158970
+    },
+    {
+      "epoch": 7.896096155756432,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0001683162809178504,
+      "loss": 0.5119,
+      "step": 158980
+    },
+    {
+      "epoch": 7.896592828052051,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00016827654713420087,
+      "loss": 0.4928,
+      "step": 158990
+    },
+    {
+      "epoch": 7.89708950034767,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001682368133505513,
+      "loss": 0.4702,
+      "step": 159000
+    },
+    {
+      "epoch": 7.89758617264329,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00016819707956690176,
+      "loss": 0.481,
+      "step": 159010
+    },
+    {
+      "epoch": 7.8980828449389096,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0001681573457832522,
+      "loss": 0.5237,
+      "step": 159020
+    },
+    {
+      "epoch": 7.898579517234529,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00016811761199960268,
+      "loss": 0.4982,
+      "step": 159030
+    },
+    {
+      "epoch": 7.899076189530148,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00016807787821595312,
+      "loss": 0.4826,
+      "step": 159040
+    },
+    {
+      "epoch": 7.899572861825767,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00016803814443230357,
+      "loss": 0.4898,
+      "step": 159050
+    },
+    {
+      "epoch": 7.9000695341213865,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016799841064865404,
+      "loss": 0.479,
+      "step": 159060
+    },
+    {
+      "epoch": 7.900566206417006,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00016795867686500448,
+      "loss": 0.4809,
+      "step": 159070
+    },
+    {
+      "epoch": 7.901062878712626,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00016791894308135495,
+      "loss": 0.4751,
+      "step": 159080
+    },
+    {
+      "epoch": 7.901559551008245,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00016787920929770537,
+      "loss": 0.5325,
+      "step": 159090
+    },
+    {
+      "epoch": 7.902056223303864,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00016783947551405584,
+      "loss": 0.4959,
+      "step": 159100
+    },
+    {
+      "epoch": 7.9025528955994835,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00016779974173040629,
+      "loss": 0.4888,
+      "step": 159110
+    },
+    {
+      "epoch": 7.903049567895103,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00016776000794675676,
+      "loss": 0.5317,
+      "step": 159120
+    },
+    {
+      "epoch": 7.903546240190722,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00016772027416310717,
+      "loss": 0.4846,
+      "step": 159130
+    },
+    {
+      "epoch": 7.904042912486341,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00016768054037945765,
+      "loss": 0.5025,
+      "step": 159140
+    },
+    {
+      "epoch": 7.904539584781961,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001676408065958081,
+      "loss": 0.5048,
+      "step": 159150
+    },
+    {
+      "epoch": 7.9050362570775805,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00016760107281215856,
+      "loss": 0.5299,
+      "step": 159160
+    },
+    {
+      "epoch": 7.9055329293732,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00016756133902850898,
+      "loss": 0.4928,
+      "step": 159170
+    },
+    {
+      "epoch": 7.906029601668819,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00016752160524485945,
+      "loss": 0.491,
+      "step": 159180
+    },
+    {
+      "epoch": 7.906526273964438,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001674818714612099,
+      "loss": 0.4857,
+      "step": 159190
+    },
+    {
+      "epoch": 7.907022946260057,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00016744213767756037,
+      "loss": 0.4391,
+      "step": 159200
+    },
+    {
+      "epoch": 7.907519618555677,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001674024038939108,
+      "loss": 0.5006,
+      "step": 159210
+    },
+    {
+      "epoch": 7.908016290851297,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00016736267011026125,
+      "loss": 0.5145,
+      "step": 159220
+    },
+    {
+      "epoch": 7.908512963146916,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00016732293632661173,
+      "loss": 0.4822,
+      "step": 159230
+    },
+    {
+      "epoch": 7.909009635442535,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00016728320254296217,
+      "loss": 0.4696,
+      "step": 159240
+    },
+    {
+      "epoch": 7.909506307738154,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00016724346875931261,
+      "loss": 0.4848,
+      "step": 159250
+    },
+    {
+      "epoch": 7.910002980033774,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00016720373497566306,
+      "loss": 0.4845,
+      "step": 159260
+    },
+    {
+      "epoch": 7.910499652329393,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00016716400119201353,
+      "loss": 0.5213,
+      "step": 159270
+    },
+    {
+      "epoch": 7.910996324625012,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00016712426740836397,
+      "loss": 0.4884,
+      "step": 159280
+    },
+    {
+      "epoch": 7.911492996920632,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00016708453362471442,
+      "loss": 0.4857,
+      "step": 159290
+    },
+    {
+      "epoch": 7.9119896692162515,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00016704479984106486,
+      "loss": 0.4944,
+      "step": 159300
+    },
+    {
+      "epoch": 7.912486341511871,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016700506605741533,
+      "loss": 0.5093,
+      "step": 159310
+    },
+    {
+      "epoch": 7.91298301380749,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00016696533227376578,
+      "loss": 0.4598,
+      "step": 159320
+    },
+    {
+      "epoch": 7.913479686103109,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00016692559849011622,
+      "loss": 0.5011,
+      "step": 159330
+    },
+    {
+      "epoch": 7.913976358398728,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00016688586470646667,
+      "loss": 0.4991,
+      "step": 159340
+    },
+    {
+      "epoch": 7.914473030694348,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00016684613092281714,
+      "loss": 0.4935,
+      "step": 159350
+    },
+    {
+      "epoch": 7.914969702989968,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016680639713916758,
+      "loss": 0.4924,
+      "step": 159360
+    },
+    {
+      "epoch": 7.915466375285587,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016676666335551803,
+      "loss": 0.503,
+      "step": 159370
+    },
+    {
+      "epoch": 7.915963047581206,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0001667269295718685,
+      "loss": 0.4942,
+      "step": 159380
+    },
+    {
+      "epoch": 7.916459719876825,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00016668719578821894,
+      "loss": 0.4976,
+      "step": 159390
+    },
+    {
+      "epoch": 7.916956392172445,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00016664746200456939,
+      "loss": 0.4774,
+      "step": 159400
+    },
+    {
+      "epoch": 7.917453064468064,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.00016660772822091983,
+      "loss": 0.4851,
+      "step": 159410
+    },
+    {
+      "epoch": 7.917949736763683,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0001665679944372703,
+      "loss": 0.4923,
+      "step": 159420
+    },
+    {
+      "epoch": 7.918446409059302,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00016652826065362075,
+      "loss": 0.4775,
+      "step": 159430
+    },
+    {
+      "epoch": 7.9189430813549215,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00016648852686997122,
+      "loss": 0.5103,
+      "step": 159440
+    },
+    {
+      "epoch": 7.919439753650542,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00016644879308632166,
+      "loss": 0.5057,
+      "step": 159450
+    },
+    {
+      "epoch": 7.919936425946161,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0001664090593026721,
+      "loss": 0.4869,
+      "step": 159460
+    },
+    {
+      "epoch": 7.92043309824178,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00016636932551902255,
+      "loss": 0.4947,
+      "step": 159470
+    },
+    {
+      "epoch": 7.920929770537399,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00016632959173537302,
+      "loss": 0.4713,
+      "step": 159480
+    },
+    {
+      "epoch": 7.9214264428330186,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00016628985795172347,
+      "loss": 0.4921,
+      "step": 159490
+    },
+    {
+      "epoch": 7.921923115128638,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001662501241680739,
+      "loss": 0.4989,
+      "step": 159500
+    },
+    {
+      "epoch": 7.922419787424257,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00016621039038442435,
+      "loss": 0.5064,
+      "step": 159510
+    },
+    {
+      "epoch": 7.922916459719877,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00016617065660077483,
+      "loss": 0.4692,
+      "step": 159520
+    },
+    {
+      "epoch": 7.923413132015496,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00016613092281712527,
+      "loss": 0.5094,
+      "step": 159530
+    },
+    {
+      "epoch": 7.923909804311116,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00016609118903347571,
+      "loss": 0.5177,
+      "step": 159540
+    },
+    {
+      "epoch": 7.924406476606735,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00016605145524982616,
+      "loss": 0.5124,
+      "step": 159550
+    },
+    {
+      "epoch": 7.924903148902354,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00016601172146617663,
+      "loss": 0.5021,
+      "step": 159560
+    },
+    {
+      "epoch": 7.925399821197973,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00016597198768252707,
+      "loss": 0.5097,
+      "step": 159570
+    },
+    {
+      "epoch": 7.9258964934935925,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00016593225389887752,
+      "loss": 0.4946,
+      "step": 159580
+    },
+    {
+      "epoch": 7.926393165789213,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.000165892520115228,
+      "loss": 0.5066,
+      "step": 159590
+    },
+    {
+      "epoch": 7.926889838084832,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00016585278633157843,
+      "loss": 0.5075,
+      "step": 159600
+    },
+    {
+      "epoch": 7.927386510380451,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001658130525479289,
+      "loss": 0.4872,
+      "step": 159610
+    },
+    {
+      "epoch": 7.92788318267607,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00016577331876427932,
+      "loss": 0.4864,
+      "step": 159620
+    },
+    {
+      "epoch": 7.9283798549716895,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001657335849806298,
+      "loss": 0.4786,
+      "step": 159630
+    },
+    {
+      "epoch": 7.928876527267309,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00016569385119698024,
+      "loss": 0.511,
+      "step": 159640
+    },
+    {
+      "epoch": 7.929373199562928,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001656541174133307,
+      "loss": 0.4909,
+      "step": 159650
+    },
+    {
+      "epoch": 7.929869871858548,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016561438362968113,
+      "loss": 0.4978,
+      "step": 159660
+    },
+    {
+      "epoch": 7.930366544154167,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001655746498460316,
+      "loss": 0.475,
+      "step": 159670
+    },
+    {
+      "epoch": 7.9308632164497865,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00016553491606238204,
+      "loss": 0.4994,
+      "step": 159680
+    },
+    {
+      "epoch": 7.931359888745406,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001654951822787325,
+      "loss": 0.5013,
+      "step": 159690
+    },
+    {
+      "epoch": 7.931856561041025,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00016545544849508293,
+      "loss": 0.5227,
+      "step": 159700
+    },
+    {
+      "epoch": 7.932353233336644,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001654157147114334,
+      "loss": 0.4901,
+      "step": 159710
+    },
+    {
+      "epoch": 7.932849905632263,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00016537598092778385,
+      "loss": 0.5077,
+      "step": 159720
+    },
+    {
+      "epoch": 7.933346577927884,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00016533624714413432,
+      "loss": 0.4844,
+      "step": 159730
+    },
+    {
+      "epoch": 7.933843250223503,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00016529651336048476,
+      "loss": 0.4831,
+      "step": 159740
+    },
+    {
+      "epoch": 7.934339922519122,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001652567795768352,
+      "loss": 0.4687,
+      "step": 159750
+    },
+    {
+      "epoch": 7.934836594814741,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00016521704579318568,
+      "loss": 0.4882,
+      "step": 159760
+    },
+    {
+      "epoch": 7.9353332671103605,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00016517731200953612,
+      "loss": 0.5135,
+      "step": 159770
+    },
+    {
+      "epoch": 7.93582993940598,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00016513757822588657,
+      "loss": 0.5015,
+      "step": 159780
+    },
+    {
+      "epoch": 7.936326611701599,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.000165097844442237,
+      "loss": 0.4614,
+      "step": 159790
+    },
+    {
+      "epoch": 7.936823283997219,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00016505811065858748,
+      "loss": 0.4798,
+      "step": 159800
+    },
+    {
+      "epoch": 7.937319956292838,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00016501837687493793,
+      "loss": 0.4945,
+      "step": 159810
+    },
+    {
+      "epoch": 7.9378166285884575,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00016497864309128837,
+      "loss": 0.4811,
+      "step": 159820
+    },
+    {
+      "epoch": 7.938313300884077,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00016493890930763881,
+      "loss": 0.497,
+      "step": 159830
+    },
+    {
+      "epoch": 7.938809973179696,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00016489917552398929,
+      "loss": 0.5042,
+      "step": 159840
+    },
+    {
+      "epoch": 7.939306645475315,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00016485944174033973,
+      "loss": 0.496,
+      "step": 159850
+    },
+    {
+      "epoch": 7.939803317770934,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001648197079566902,
+      "loss": 0.5029,
+      "step": 159860
+    },
+    {
+      "epoch": 7.9402999900665545,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00016477997417304062,
+      "loss": 0.479,
+      "step": 159870
+    },
+    {
+      "epoch": 7.940796662362174,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001647402403893911,
+      "loss": 0.5002,
+      "step": 159880
+    },
+    {
+      "epoch": 7.941293334657793,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00016470050660574153,
+      "loss": 0.4777,
+      "step": 159890
+    },
+    {
+      "epoch": 7.941790006953412,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.000164660772822092,
+      "loss": 0.4911,
+      "step": 159900
+    },
+    {
+      "epoch": 7.942286679249031,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00016462103903844245,
+      "loss": 0.4854,
+      "step": 159910
+    },
+    {
+      "epoch": 7.942783351544651,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0001645813052547929,
+      "loss": 0.5135,
+      "step": 159920
+    },
+    {
+      "epoch": 7.94328002384027,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00016454157147114337,
+      "loss": 0.4886,
+      "step": 159930
+    },
+    {
+      "epoch": 7.94377669613589,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001645018376874938,
+      "loss": 0.4596,
+      "step": 159940
+    },
+    {
+      "epoch": 7.944273368431509,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00016446210390384425,
+      "loss": 0.4739,
+      "step": 159950
+    },
+    {
+      "epoch": 7.9447700407271284,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001644223701201947,
+      "loss": 0.4765,
+      "step": 159960
+    },
+    {
+      "epoch": 7.945266713022748,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00016438263633654517,
+      "loss": 0.4598,
+      "step": 159970
+    },
+    {
+      "epoch": 7.945763385318367,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001643429025528956,
+      "loss": 0.5043,
+      "step": 159980
+    },
+    {
+      "epoch": 7.946260057613986,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00016430316876924606,
+      "loss": 0.514,
+      "step": 159990
+    },
+    {
+      "epoch": 7.946756729909605,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001642634349855965,
+      "loss": 0.4984,
+      "step": 160000
+    },
+    {
+      "epoch": 7.9472534022052255,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00016422370120194697,
+      "loss": 0.498,
+      "step": 160010
+    },
+    {
+      "epoch": 7.947750074500845,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00016418396741829742,
+      "loss": 0.51,
+      "step": 160020
+    },
+    {
+      "epoch": 7.948246746796464,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00016414423363464786,
+      "loss": 0.5087,
+      "step": 160030
+    },
+    {
+      "epoch": 7.948743419092083,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001641044998509983,
+      "loss": 0.5029,
+      "step": 160040
+    },
+    {
+      "epoch": 7.949240091387702,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00016406476606734878,
+      "loss": 0.5312,
+      "step": 160050
+    },
+    {
+      "epoch": 7.949736763683322,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00016402503228369922,
+      "loss": 0.4919,
+      "step": 160060
+    },
+    {
+      "epoch": 7.950233435978941,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00016398529850004967,
+      "loss": 0.4937,
+      "step": 160070
+    },
+    {
+      "epoch": 7.95073010827456,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00016394556471640014,
+      "loss": 0.475,
+      "step": 160080
+    },
+    {
+      "epoch": 7.95122678057018,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00016390583093275058,
+      "loss": 0.4837,
+      "step": 160090
+    },
+    {
+      "epoch": 7.951723452865799,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00016386609714910103,
+      "loss": 0.4909,
+      "step": 160100
+    },
+    {
+      "epoch": 7.952220125161419,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00016382636336545147,
+      "loss": 0.5178,
+      "step": 160110
+    },
+    {
+      "epoch": 7.952716797457038,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00016378662958180194,
+      "loss": 0.5154,
+      "step": 160120
+    },
+    {
+      "epoch": 7.953213469752657,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00016374689579815239,
+      "loss": 0.4721,
+      "step": 160130
+    },
+    {
+      "epoch": 7.953710142048276,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00016370716201450286,
+      "loss": 0.5122,
+      "step": 160140
+    },
+    {
+      "epoch": 7.9542068143438955,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00016366742823085327,
+      "loss": 0.4571,
+      "step": 160150
+    },
+    {
+      "epoch": 7.954703486639515,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00016362769444720375,
+      "loss": 0.4744,
+      "step": 160160
+    },
+    {
+      "epoch": 7.955200158935135,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001635879606635542,
+      "loss": 0.498,
+      "step": 160170
+    },
+    {
+      "epoch": 7.955696831230754,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00016354822687990466,
+      "loss": 0.4578,
+      "step": 160180
+    },
+    {
+      "epoch": 7.956193503526373,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00016350849309625508,
+      "loss": 0.4896,
+      "step": 160190
+    },
+    {
+      "epoch": 7.956690175821993,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00016346875931260555,
+      "loss": 0.4738,
+      "step": 160200
+    },
+    {
+      "epoch": 7.957186848117612,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.000163429025528956,
+      "loss": 0.5022,
+      "step": 160210
+    },
+    {
+      "epoch": 7.957683520413231,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00016338929174530647,
+      "loss": 0.4943,
+      "step": 160220
+    },
+    {
+      "epoch": 7.95818019270885,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001633495579616569,
+      "loss": 0.4896,
+      "step": 160230
+    },
+    {
+      "epoch": 7.95867686500447,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016330982417800735,
+      "loss": 0.492,
+      "step": 160240
+    },
+    {
+      "epoch": 7.95917353730009,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001632700903943578,
+      "loss": 0.5101,
+      "step": 160250
+    },
+    {
+      "epoch": 7.959670209595709,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00016323035661070827,
+      "loss": 0.5329,
+      "step": 160260
+    },
+    {
+      "epoch": 7.960166881891328,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00016319062282705871,
+      "loss": 0.476,
+      "step": 160270
+    },
+    {
+      "epoch": 7.960663554186947,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00016315088904340916,
+      "loss": 0.4737,
+      "step": 160280
+    },
+    {
+      "epoch": 7.9611602264825665,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00016311115525975963,
+      "loss": 0.4942,
+      "step": 160290
+    },
+    {
+      "epoch": 7.961656898778186,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00016307142147611007,
+      "loss": 0.5531,
+      "step": 160300
+    },
+    {
+      "epoch": 7.962153571073806,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00016303168769246054,
+      "loss": 0.5119,
+      "step": 160310
+    },
+    {
+      "epoch": 7.962650243369425,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00016299195390881096,
+      "loss": 0.5132,
+      "step": 160320
+    },
+    {
+      "epoch": 7.963146915665044,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00016295222012516143,
+      "loss": 0.479,
+      "step": 160330
+    },
+    {
+      "epoch": 7.9636435879606635,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00016291248634151188,
+      "loss": 0.5006,
+      "step": 160340
+    },
+    {
+      "epoch": 7.964140260256283,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00016287275255786235,
+      "loss": 0.4917,
+      "step": 160350
+    },
+    {
+      "epoch": 7.964636932551902,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00016283301877421277,
+      "loss": 0.5246,
+      "step": 160360
+    },
+    {
+      "epoch": 7.965133604847521,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00016279328499056324,
+      "loss": 0.4925,
+      "step": 160370
+    },
+    {
+      "epoch": 7.965630277143141,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00016275355120691368,
+      "loss": 0.474,
+      "step": 160380
+    },
+    {
+      "epoch": 7.9661269494387605,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00016271381742326415,
+      "loss": 0.4957,
+      "step": 160390
+    },
+    {
+      "epoch": 7.96662362173438,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00016267408363961457,
+      "loss": 0.4688,
+      "step": 160400
+    },
+    {
+      "epoch": 7.967120294029999,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016263434985596504,
+      "loss": 0.4886,
+      "step": 160410
+    },
+    {
+      "epoch": 7.967616966325618,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00016259461607231549,
+      "loss": 0.4947,
+      "step": 160420
+    },
+    {
+      "epoch": 7.9681136386212374,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016255488228866596,
+      "loss": 0.4888,
+      "step": 160430
+    },
+    {
+      "epoch": 7.968610310916857,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001625151485050164,
+      "loss": 0.4831,
+      "step": 160440
+    },
+    {
+      "epoch": 7.969106983212477,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00016247541472136685,
+      "loss": 0.5021,
+      "step": 160450
+    },
+    {
+      "epoch": 7.969603655508096,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00016243568093771732,
+      "loss": 0.4852,
+      "step": 160460
+    },
+    {
+      "epoch": 7.970100327803715,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00016239594715406776,
+      "loss": 0.4923,
+      "step": 160470
+    },
+    {
+      "epoch": 7.9705970000993345,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001623562133704182,
+      "loss": 0.4684,
+      "step": 160480
+    },
+    {
+      "epoch": 7.971093672394954,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016231647958676865,
+      "loss": 0.4737,
+      "step": 160490
+    },
+    {
+      "epoch": 7.971590344690573,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00016227674580311912,
+      "loss": 0.521,
+      "step": 160500
+    },
+    {
+      "epoch": 7.972087016986192,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00016223701201946957,
+      "loss": 0.468,
+      "step": 160510
+    },
+    {
+      "epoch": 7.972583689281812,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00016219727823582,
+      "loss": 0.5087,
+      "step": 160520
+    },
+    {
+      "epoch": 7.9730803615774315,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00016215754445217045,
+      "loss": 0.4756,
+      "step": 160530
+    },
+    {
+      "epoch": 7.973577033873051,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00016211781066852093,
+      "loss": 0.5339,
+      "step": 160540
+    },
+    {
+      "epoch": 7.97407370616867,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00016207807688487137,
+      "loss": 0.4721,
+      "step": 160550
+    },
+    {
+      "epoch": 7.974570378464289,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00016203834310122181,
+      "loss": 0.5024,
+      "step": 160560
+    },
+    {
+      "epoch": 7.975067050759908,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00016199860931757226,
+      "loss": 0.4412,
+      "step": 160570
+    },
+    {
+      "epoch": 7.975563723055528,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00016195887553392273,
+      "loss": 0.4727,
+      "step": 160580
+    },
+    {
+      "epoch": 7.976060395351148,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00016191914175027317,
+      "loss": 0.5233,
+      "step": 160590
+    },
+    {
+      "epoch": 7.976557067646767,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00016187940796662362,
+      "loss": 0.4815,
+      "step": 160600
+    },
+    {
+      "epoch": 7.977053739942386,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001618396741829741,
+      "loss": 0.4791,
+      "step": 160610
+    },
+    {
+      "epoch": 7.977550412238005,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00016179994039932453,
+      "loss": 0.5128,
+      "step": 160620
+    },
+    {
+      "epoch": 7.978047084533625,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.000161760206615675,
+      "loss": 0.4934,
+      "step": 160630
+    },
+    {
+      "epoch": 7.978543756829244,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00016172047283202542,
+      "loss": 0.5397,
+      "step": 160640
+    },
+    {
+      "epoch": 7.979040429124863,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001616807390483759,
+      "loss": 0.4757,
+      "step": 160650
+    },
+    {
+      "epoch": 7.979537101420483,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00016164100526472634,
+      "loss": 0.4689,
+      "step": 160660
+    },
+    {
+      "epoch": 7.9800337737161025,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001616012714810768,
+      "loss": 0.5168,
+      "step": 160670
+    },
+    {
+      "epoch": 7.980530446011722,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00016156153769742723,
+      "loss": 0.4812,
+      "step": 160680
+    },
+    {
+      "epoch": 7.981027118307341,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001615218039137777,
+      "loss": 0.5067,
+      "step": 160690
+    },
+    {
+      "epoch": 7.98152379060296,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00016148207013012814,
+      "loss": 0.4903,
+      "step": 160700
+    },
+    {
+      "epoch": 7.982020462898579,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0001614423363464786,
+      "loss": 0.4981,
+      "step": 160710
+    },
+    {
+      "epoch": 7.982517135194199,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00016140260256282906,
+      "loss": 0.5052,
+      "step": 160720
+    },
+    {
+      "epoch": 7.983013807489819,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0001613628687791795,
+      "loss": 0.4815,
+      "step": 160730
+    },
+    {
+      "epoch": 7.983510479785438,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00016132313499552995,
+      "loss": 0.5113,
+      "step": 160740
+    },
+    {
+      "epoch": 7.984007152081057,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00016128340121188042,
+      "loss": 0.492,
+      "step": 160750
+    },
+    {
+      "epoch": 7.984503824376676,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00016124366742823086,
+      "loss": 0.5066,
+      "step": 160760
+    },
+    {
+      "epoch": 7.985000496672296,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001612039336445813,
+      "loss": 0.497,
+      "step": 160770
+    },
+    {
+      "epoch": 7.985497168967915,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00016116419986093178,
+      "loss": 0.5048,
+      "step": 160780
+    },
+    {
+      "epoch": 7.985993841263534,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00016112446607728222,
+      "loss": 0.4885,
+      "step": 160790
+    },
+    {
+      "epoch": 7.986490513559153,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00016108473229363267,
+      "loss": 0.4969,
+      "step": 160800
+    },
+    {
+      "epoch": 7.986987185854773,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001610449985099831,
+      "loss": 0.4893,
+      "step": 160810
+    },
+    {
+      "epoch": 7.987483858150393,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00016100526472633358,
+      "loss": 0.4677,
+      "step": 160820
+    },
+    {
+      "epoch": 7.987980530446012,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00016096553094268403,
+      "loss": 0.4925,
+      "step": 160830
+    },
+    {
+      "epoch": 7.988477202741631,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001609257971590345,
+      "loss": 0.489,
+      "step": 160840
+    },
+    {
+      "epoch": 7.98897387503725,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00016088606337538491,
+      "loss": 0.4792,
+      "step": 160850
+    },
+    {
+      "epoch": 7.9894705473328695,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00016084632959173539,
+      "loss": 0.5002,
+      "step": 160860
+    },
+    {
+      "epoch": 7.989967219628489,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00016080659580808583,
+      "loss": 0.4733,
+      "step": 160870
+    },
+    {
+      "epoch": 7.990463891924108,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001607668620244363,
+      "loss": 0.4713,
+      "step": 160880
+    },
+    {
+      "epoch": 7.990960564219728,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00016072712824078672,
+      "loss": 0.5031,
+      "step": 160890
+    },
+    {
+      "epoch": 7.991457236515347,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001606873944571372,
+      "loss": 0.4917,
+      "step": 160900
+    },
+    {
+      "epoch": 7.991953908810967,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00016064766067348763,
+      "loss": 0.4995,
+      "step": 160910
+    },
+    {
+      "epoch": 7.992450581106586,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001606079268898381,
+      "loss": 0.5008,
+      "step": 160920
+    },
+    {
+      "epoch": 7.992947253402205,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00016056819310618855,
+      "loss": 0.4829,
+      "step": 160930
+    },
+    {
+      "epoch": 7.993443925697824,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.000160528459322539,
+      "loss": 0.5061,
+      "step": 160940
+    },
+    {
+      "epoch": 7.9939405979934435,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00016048872553888944,
+      "loss": 0.4793,
+      "step": 160950
+    },
+    {
+      "epoch": 7.994437270289064,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001604489917552399,
+      "loss": 0.4777,
+      "step": 160960
+    },
+    {
+      "epoch": 7.994933942584683,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016040925797159035,
+      "loss": 0.4421,
+      "step": 160970
+    },
+    {
+      "epoch": 7.995430614880302,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001603695241879408,
+      "loss": 0.4834,
+      "step": 160980
+    },
+    {
+      "epoch": 7.995927287175921,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00016032979040429127,
+      "loss": 0.5118,
+      "step": 160990
+    },
+    {
+      "epoch": 7.9964239594715405,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001602900566206417,
+      "loss": 0.4914,
+      "step": 161000
+    },
+    {
+      "epoch": 7.99692063176716,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00016025032283699216,
+      "loss": 0.4723,
+      "step": 161010
+    },
+    {
+      "epoch": 7.997417304062779,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001602105890533426,
+      "loss": 0.5193,
+      "step": 161020
+    },
+    {
+      "epoch": 7.997913976358399,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00016017085526969307,
+      "loss": 0.5055,
+      "step": 161030
+    },
+    {
+      "epoch": 7.998410648654018,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00016013112148604352,
+      "loss": 0.454,
+      "step": 161040
+    },
+    {
+      "epoch": 7.9989073209496375,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00016009138770239396,
+      "loss": 0.4973,
+      "step": 161050
+    },
+    {
+      "epoch": 7.999403993245257,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001600516539187444,
+      "loss": 0.4694,
+      "step": 161060
+    },
+    {
+      "epoch": 7.999900665540876,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00016001192013509488,
+      "loss": 0.5157,
+      "step": 161070
+    },
+    {
+      "epoch": 8.000397337836496,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00015997218635144532,
+      "loss": 0.5245,
+      "step": 161080
+    },
+    {
+      "epoch": 8.000894010132114,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015993245256779577,
+      "loss": 0.4829,
+      "step": 161090
+    },
+    {
+      "epoch": 8.001390682427735,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001598927187841462,
+      "loss": 0.4883,
+      "step": 161100
+    },
+    {
+      "epoch": 8.001887354723353,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00015985298500049668,
+      "loss": 0.478,
+      "step": 161110
+    },
+    {
+      "epoch": 8.002384027018973,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00015981325121684713,
+      "loss": 0.508,
+      "step": 161120
+    },
+    {
+      "epoch": 8.002880699314593,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0001597735174331976,
+      "loss": 0.4908,
+      "step": 161130
+    },
+    {
+      "epoch": 8.003377371610211,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00015973378364954804,
+      "loss": 0.4701,
+      "step": 161140
+    },
+    {
+      "epoch": 8.003874043905832,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00015969404986589849,
+      "loss": 0.459,
+      "step": 161150
+    },
+    {
+      "epoch": 8.00437071620145,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00015965431608224896,
+      "loss": 0.4498,
+      "step": 161160
+    },
+    {
+      "epoch": 8.00486738849707,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001596145822985994,
+      "loss": 0.4693,
+      "step": 161170
+    },
+    {
+      "epoch": 8.005364060792688,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00015957484851494985,
+      "loss": 0.4906,
+      "step": 161180
+    },
+    {
+      "epoch": 8.005860733088308,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001595351147313003,
+      "loss": 0.478,
+      "step": 161190
+    },
+    {
+      "epoch": 8.006357405383927,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00015949538094765076,
+      "loss": 0.4693,
+      "step": 161200
+    },
+    {
+      "epoch": 8.006854077679547,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001594556471640012,
+      "loss": 0.4263,
+      "step": 161210
+    },
+    {
+      "epoch": 8.007350749975167,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00015941591338035165,
+      "loss": 0.4807,
+      "step": 161220
+    },
+    {
+      "epoch": 8.007847422270785,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001593761795967021,
+      "loss": 0.541,
+      "step": 161230
+    },
+    {
+      "epoch": 8.008344094566406,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00015933644581305256,
+      "loss": 0.4734,
+      "step": 161240
+    },
+    {
+      "epoch": 8.008840766862024,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.000159296712029403,
+      "loss": 0.4799,
+      "step": 161250
+    },
+    {
+      "epoch": 8.009337439157644,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00015925697824575345,
+      "loss": 0.4768,
+      "step": 161260
+    },
+    {
+      "epoch": 8.009834111453262,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001592172444621039,
+      "loss": 0.5115,
+      "step": 161270
+    },
+    {
+      "epoch": 8.010330783748882,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00015917751067845437,
+      "loss": 0.4876,
+      "step": 161280
+    },
+    {
+      "epoch": 8.010827456044503,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001591377768948048,
+      "loss": 0.4816,
+      "step": 161290
+    },
+    {
+      "epoch": 8.01132412834012,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00015909804311115526,
+      "loss": 0.4597,
+      "step": 161300
+    },
+    {
+      "epoch": 8.011820800635741,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00015905830932750573,
+      "loss": 0.4957,
+      "step": 161310
+    },
+    {
+      "epoch": 8.01231747293136,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00015901857554385617,
+      "loss": 0.4762,
+      "step": 161320
+    },
+    {
+      "epoch": 8.01281414522698,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00015897884176020664,
+      "loss": 0.4954,
+      "step": 161330
+    },
+    {
+      "epoch": 8.013310817522598,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00015893910797655706,
+      "loss": 0.4877,
+      "step": 161340
+    },
+    {
+      "epoch": 8.013807489818218,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00015889937419290753,
+      "loss": 0.4988,
+      "step": 161350
+    },
+    {
+      "epoch": 8.014304162113838,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00015885964040925798,
+      "loss": 0.4862,
+      "step": 161360
+    },
+    {
+      "epoch": 8.014800834409456,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00015881990662560845,
+      "loss": 0.4652,
+      "step": 161370
+    },
+    {
+      "epoch": 8.015297506705076,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00015878017284195887,
+      "loss": 0.4944,
+      "step": 161380
+    },
+    {
+      "epoch": 8.015794179000695,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015874043905830934,
+      "loss": 0.4868,
+      "step": 161390
+    },
+    {
+      "epoch": 8.016290851296315,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00015870070527465978,
+      "loss": 0.527,
+      "step": 161400
+    },
+    {
+      "epoch": 8.016787523591933,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00015866097149101025,
+      "loss": 0.4963,
+      "step": 161410
+    },
+    {
+      "epoch": 8.017284195887553,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00015862123770736067,
+      "loss": 0.4732,
+      "step": 161420
+    },
+    {
+      "epoch": 8.017780868183173,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00015858150392371114,
+      "loss": 0.4751,
+      "step": 161430
+    },
+    {
+      "epoch": 8.018277540478792,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00015854177014006159,
+      "loss": 0.45,
+      "step": 161440
+    },
+    {
+      "epoch": 8.018774212774412,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00015850203635641206,
+      "loss": 0.495,
+      "step": 161450
+    },
+    {
+      "epoch": 8.01927088507003,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001584623025727625,
+      "loss": 0.4826,
+      "step": 161460
+    },
+    {
+      "epoch": 8.01976755736565,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00015842256878911295,
+      "loss": 0.5109,
+      "step": 161470
+    },
+    {
+      "epoch": 8.020264229661269,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00015838283500546342,
+      "loss": 0.4613,
+      "step": 161480
+    },
+    {
+      "epoch": 8.020760901956889,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00015834310122181386,
+      "loss": 0.4482,
+      "step": 161490
+    },
+    {
+      "epoch": 8.021257574252509,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001583033674381643,
+      "loss": 0.4715,
+      "step": 161500
+    },
+    {
+      "epoch": 8.021754246548127,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00015826363365451475,
+      "loss": 0.5069,
+      "step": 161510
+    },
+    {
+      "epoch": 8.022250918843747,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00015822389987086522,
+      "loss": 0.4852,
+      "step": 161520
+    },
+    {
+      "epoch": 8.022747591139366,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00015818416608721566,
+      "loss": 0.4645,
+      "step": 161530
+    },
+    {
+      "epoch": 8.023244263434986,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00015814443230356614,
+      "loss": 0.4513,
+      "step": 161540
+    },
+    {
+      "epoch": 8.023740935730604,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00015810469851991655,
+      "loss": 0.4544,
+      "step": 161550
+    },
+    {
+      "epoch": 8.024237608026224,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00015806496473626702,
+      "loss": 0.4841,
+      "step": 161560
+    },
+    {
+      "epoch": 8.024734280321844,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00015802523095261747,
+      "loss": 0.5216,
+      "step": 161570
+    },
+    {
+      "epoch": 8.025230952617463,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00015798549716896794,
+      "loss": 0.5066,
+      "step": 161580
+    },
+    {
+      "epoch": 8.025727624913083,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00015794576338531836,
+      "loss": 0.4658,
+      "step": 161590
+    },
+    {
+      "epoch": 8.026224297208701,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00015790602960166883,
+      "loss": 0.4869,
+      "step": 161600
+    },
+    {
+      "epoch": 8.026720969504321,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00015786629581801927,
+      "loss": 0.4999,
+      "step": 161610
+    },
+    {
+      "epoch": 8.02721764179994,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00015782656203436974,
+      "loss": 0.483,
+      "step": 161620
+    },
+    {
+      "epoch": 8.02771431409556,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001577868282507202,
+      "loss": 0.491,
+      "step": 161630
+    },
+    {
+      "epoch": 8.02821098639118,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00015774709446707063,
+      "loss": 0.4676,
+      "step": 161640
+    },
+    {
+      "epoch": 8.028707658686798,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00015770736068342108,
+      "loss": 0.466,
+      "step": 161650
+    },
+    {
+      "epoch": 8.029204330982418,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00015766762689977155,
+      "loss": 0.5024,
+      "step": 161660
+    },
+    {
+      "epoch": 8.029701003278037,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.000157627893116122,
+      "loss": 0.4902,
+      "step": 161670
+    },
+    {
+      "epoch": 8.030197675573657,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00015758815933247244,
+      "loss": 0.4683,
+      "step": 161680
+    },
+    {
+      "epoch": 8.030694347869275,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001575484255488229,
+      "loss": 0.484,
+      "step": 161690
+    },
+    {
+      "epoch": 8.031191020164895,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00015750869176517335,
+      "loss": 0.5135,
+      "step": 161700
+    },
+    {
+      "epoch": 8.031687692460515,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0001574689579815238,
+      "loss": 0.4741,
+      "step": 161710
+    },
+    {
+      "epoch": 8.032184364756134,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00015742922419787424,
+      "loss": 0.4908,
+      "step": 161720
+    },
+    {
+      "epoch": 8.032681037051754,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001573894904142247,
+      "loss": 0.4432,
+      "step": 161730
+    },
+    {
+      "epoch": 8.033177709347372,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00015734975663057516,
+      "loss": 0.4724,
+      "step": 161740
+    },
+    {
+      "epoch": 8.033674381642992,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001573100228469256,
+      "loss": 0.4892,
+      "step": 161750
+    },
+    {
+      "epoch": 8.03417105393861,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015727028906327605,
+      "loss": 0.4908,
+      "step": 161760
+    },
+    {
+      "epoch": 8.03466772623423,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00015723055527962652,
+      "loss": 0.493,
+      "step": 161770
+    },
+    {
+      "epoch": 8.03516439852985,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00015719082149597696,
+      "loss": 0.4561,
+      "step": 161780
+    },
+    {
+      "epoch": 8.03566107082547,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001571510877123274,
+      "loss": 0.5212,
+      "step": 161790
+    },
+    {
+      "epoch": 8.03615774312109,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00015711135392867785,
+      "loss": 0.4831,
+      "step": 161800
+    },
+    {
+      "epoch": 8.036654415416708,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00015707162014502832,
+      "loss": 0.4874,
+      "step": 161810
+    },
+    {
+      "epoch": 8.037151087712328,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00015703188636137877,
+      "loss": 0.4625,
+      "step": 161820
+    },
+    {
+      "epoch": 8.037647760007946,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001569921525777292,
+      "loss": 0.4806,
+      "step": 161830
+    },
+    {
+      "epoch": 8.038144432303566,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00015695241879407968,
+      "loss": 0.4679,
+      "step": 161840
+    },
+    {
+      "epoch": 8.038641104599186,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00015691268501043012,
+      "loss": 0.4581,
+      "step": 161850
+    },
+    {
+      "epoch": 8.039137776894805,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0001568729512267806,
+      "loss": 0.4933,
+      "step": 161860
+    },
+    {
+      "epoch": 8.039634449190425,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.000156833217443131,
+      "loss": 0.4792,
+      "step": 161870
+    },
+    {
+      "epoch": 8.040131121486043,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00015679348365948148,
+      "loss": 0.4653,
+      "step": 161880
+    },
+    {
+      "epoch": 8.040627793781663,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00015675374987583193,
+      "loss": 0.4661,
+      "step": 161890
+    },
+    {
+      "epoch": 8.041124466077282,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001567140160921824,
+      "loss": 0.4943,
+      "step": 161900
+    },
+    {
+      "epoch": 8.041621138372902,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00015667428230853282,
+      "loss": 0.5065,
+      "step": 161910
+    },
+    {
+      "epoch": 8.04211781066852,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001566345485248833,
+      "loss": 0.4755,
+      "step": 161920
+    },
+    {
+      "epoch": 8.04261448296414,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00015659481474123373,
+      "loss": 0.4787,
+      "step": 161930
+    },
+    {
+      "epoch": 8.04311115525976,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001565550809575842,
+      "loss": 0.4992,
+      "step": 161940
+    },
+    {
+      "epoch": 8.043607827555379,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00015651534717393462,
+      "loss": 0.4444,
+      "step": 161950
+    },
+    {
+      "epoch": 8.044104499850999,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001564756133902851,
+      "loss": 0.4849,
+      "step": 161960
+    },
+    {
+      "epoch": 8.044601172146617,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00015643587960663554,
+      "loss": 0.5151,
+      "step": 161970
+    },
+    {
+      "epoch": 8.045097844442237,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.000156396145822986,
+      "loss": 0.4721,
+      "step": 161980
+    },
+    {
+      "epoch": 8.045594516737856,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00015635641203933645,
+      "loss": 0.4897,
+      "step": 161990
+    },
+    {
+      "epoch": 8.046091189033476,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0001563166782556869,
+      "loss": 0.4776,
+      "step": 162000
+    },
+    {
+      "epoch": 8.046587861329096,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00015627694447203737,
+      "loss": 0.4532,
+      "step": 162010
+    },
+    {
+      "epoch": 8.047084533624714,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001562372106883878,
+      "loss": 0.4664,
+      "step": 162020
+    },
+    {
+      "epoch": 8.047581205920334,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00015619747690473828,
+      "loss": 0.4775,
+      "step": 162030
+    },
+    {
+      "epoch": 8.048077878215953,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001561577431210887,
+      "loss": 0.4718,
+      "step": 162040
+    },
+    {
+      "epoch": 8.048574550511573,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00015611800933743917,
+      "loss": 0.47,
+      "step": 162050
+    },
+    {
+      "epoch": 8.049071222807191,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00015607827555378962,
+      "loss": 0.5064,
+      "step": 162060
+    },
+    {
+      "epoch": 8.049567895102811,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001560385417701401,
+      "loss": 0.4787,
+      "step": 162070
+    },
+    {
+      "epoch": 8.050064567398431,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0001559988079864905,
+      "loss": 0.4881,
+      "step": 162080
+    },
+    {
+      "epoch": 8.05056123969405,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00015595907420284098,
+      "loss": 0.4707,
+      "step": 162090
+    },
+    {
+      "epoch": 8.05105791198967,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00015591934041919142,
+      "loss": 0.4533,
+      "step": 162100
+    },
+    {
+      "epoch": 8.051554584285288,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001558796066355419,
+      "loss": 0.4785,
+      "step": 162110
+    },
+    {
+      "epoch": 8.052051256580908,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001558398728518923,
+      "loss": 0.4745,
+      "step": 162120
+    },
+    {
+      "epoch": 8.052547928876526,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00015580013906824278,
+      "loss": 0.4967,
+      "step": 162130
+    },
+    {
+      "epoch": 8.053044601172147,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00015576040528459323,
+      "loss": 0.4866,
+      "step": 162140
+    },
+    {
+      "epoch": 8.053541273467767,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001557206715009437,
+      "loss": 0.4914,
+      "step": 162150
+    },
+    {
+      "epoch": 8.054037945763385,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00015568093771729414,
+      "loss": 0.4741,
+      "step": 162160
+    },
+    {
+      "epoch": 8.054534618059005,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00015564120393364458,
+      "loss": 0.4544,
+      "step": 162170
+    },
+    {
+      "epoch": 8.055031290354624,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00015560147014999506,
+      "loss": 0.5047,
+      "step": 162180
+    },
+    {
+      "epoch": 8.055527962650244,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0001555617363663455,
+      "loss": 0.4983,
+      "step": 162190
+    },
+    {
+      "epoch": 8.056024634945862,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00015552200258269594,
+      "loss": 0.4744,
+      "step": 162200
+    },
+    {
+      "epoch": 8.056521307241482,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001554822687990464,
+      "loss": 0.5062,
+      "step": 162210
+    },
+    {
+      "epoch": 8.057017979537102,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00015544253501539686,
+      "loss": 0.4831,
+      "step": 162220
+    },
+    {
+      "epoch": 8.05751465183272,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001554028012317473,
+      "loss": 0.4767,
+      "step": 162230
+    },
+    {
+      "epoch": 8.05801132412834,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00015536306744809775,
+      "loss": 0.457,
+      "step": 162240
+    },
+    {
+      "epoch": 8.058507996423959,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001553233336644482,
+      "loss": 0.4757,
+      "step": 162250
+    },
+    {
+      "epoch": 8.059004668719579,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00015528359988079866,
+      "loss": 0.486,
+      "step": 162260
+    },
+    {
+      "epoch": 8.059501341015197,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001552438660971491,
+      "loss": 0.4749,
+      "step": 162270
+    },
+    {
+      "epoch": 8.059998013310818,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00015520413231349955,
+      "loss": 0.4829,
+      "step": 162280
+    },
+    {
+      "epoch": 8.060494685606438,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00015516439852985,
+      "loss": 0.4866,
+      "step": 162290
+    },
+    {
+      "epoch": 8.060991357902056,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00015512466474620047,
+      "loss": 0.4831,
+      "step": 162300
+    },
+    {
+      "epoch": 8.061488030197676,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001550849309625509,
+      "loss": 0.4949,
+      "step": 162310
+    },
+    {
+      "epoch": 8.061984702493294,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00015504519717890136,
+      "loss": 0.4942,
+      "step": 162320
+    },
+    {
+      "epoch": 8.062481374788915,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00015500546339525183,
+      "loss": 0.4613,
+      "step": 162330
+    },
+    {
+      "epoch": 8.062978047084533,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00015496572961160227,
+      "loss": 0.4588,
+      "step": 162340
+    },
+    {
+      "epoch": 8.063474719380153,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00015492599582795272,
+      "loss": 0.4524,
+      "step": 162350
+    },
+    {
+      "epoch": 8.063971391675773,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00015488626204430316,
+      "loss": 0.4873,
+      "step": 162360
+    },
+    {
+      "epoch": 8.064468063971391,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00015484652826065363,
+      "loss": 0.523,
+      "step": 162370
+    },
+    {
+      "epoch": 8.064964736267012,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00015480679447700408,
+      "loss": 0.4733,
+      "step": 162380
+    },
+    {
+      "epoch": 8.06546140856263,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00015476706069335455,
+      "loss": 0.4955,
+      "step": 162390
+    },
+    {
+      "epoch": 8.06595808085825,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.000154727326909705,
+      "loss": 0.5039,
+      "step": 162400
+    },
+    {
+      "epoch": 8.066454753153868,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00015468759312605544,
+      "loss": 0.489,
+      "step": 162410
+    },
+    {
+      "epoch": 8.066951425449489,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00015464785934240588,
+      "loss": 0.479,
+      "step": 162420
+    },
+    {
+      "epoch": 8.067448097745109,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00015460812555875635,
+      "loss": 0.4693,
+      "step": 162430
+    },
+    {
+      "epoch": 8.067944770040727,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001545683917751068,
+      "loss": 0.4643,
+      "step": 162440
+    },
+    {
+      "epoch": 8.068441442336347,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00015452865799145724,
+      "loss": 0.4647,
+      "step": 162450
+    },
+    {
+      "epoch": 8.068938114631965,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00015448892420780769,
+      "loss": 0.482,
+      "step": 162460
+    },
+    {
+      "epoch": 8.069434786927586,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00015444919042415816,
+      "loss": 0.4824,
+      "step": 162470
+    },
+    {
+      "epoch": 8.069931459223204,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001544094566405086,
+      "loss": 0.4639,
+      "step": 162480
+    },
+    {
+      "epoch": 8.070428131518824,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00015436972285685904,
+      "loss": 0.4406,
+      "step": 162490
+    },
+    {
+      "epoch": 8.070924803814444,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001543299890732095,
+      "loss": 0.478,
+      "step": 162500
+    },
+    {
+      "epoch": 8.071421476110062,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00015429025528955996,
+      "loss": 0.4737,
+      "step": 162510
+    },
+    {
+      "epoch": 8.071918148405683,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0001542505215059104,
+      "loss": 0.4339,
+      "step": 162520
+    },
+    {
+      "epoch": 8.072414820701301,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00015421078772226085,
+      "loss": 0.498,
+      "step": 162530
+    },
+    {
+      "epoch": 8.072911492996921,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00015417105393861132,
+      "loss": 0.4988,
+      "step": 162540
+    },
+    {
+      "epoch": 8.07340816529254,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00015413132015496176,
+      "loss": 0.4765,
+      "step": 162550
+    },
+    {
+      "epoch": 8.07390483758816,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00015409158637131224,
+      "loss": 0.4536,
+      "step": 162560
+    },
+    {
+      "epoch": 8.074401509883778,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00015405185258766265,
+      "loss": 0.5135,
+      "step": 162570
+    },
+    {
+      "epoch": 8.074898182179398,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00015401211880401312,
+      "loss": 0.4666,
+      "step": 162580
+    },
+    {
+      "epoch": 8.075394854475018,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00015397238502036357,
+      "loss": 0.4725,
+      "step": 162590
+    },
+    {
+      "epoch": 8.075891526770636,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00015393265123671404,
+      "loss": 0.4554,
+      "step": 162600
+    },
+    {
+      "epoch": 8.076388199066256,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00015389291745306446,
+      "loss": 0.4599,
+      "step": 162610
+    },
+    {
+      "epoch": 8.076884871361875,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00015385318366941493,
+      "loss": 0.4925,
+      "step": 162620
+    },
+    {
+      "epoch": 8.077381543657495,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00015381344988576537,
+      "loss": 0.4482,
+      "step": 162630
+    },
+    {
+      "epoch": 8.077878215953113,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00015377371610211584,
+      "loss": 0.4978,
+      "step": 162640
+    },
+    {
+      "epoch": 8.078374888248733,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00015373398231846626,
+      "loss": 0.469,
+      "step": 162650
+    },
+    {
+      "epoch": 8.078871560544354,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00015369424853481673,
+      "loss": 0.4678,
+      "step": 162660
+    },
+    {
+      "epoch": 8.079368232839972,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00015365451475116718,
+      "loss": 0.4798,
+      "step": 162670
+    },
+    {
+      "epoch": 8.079864905135592,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00015361478096751765,
+      "loss": 0.4581,
+      "step": 162680
+    },
+    {
+      "epoch": 8.08036157743121,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001535750471838681,
+      "loss": 0.4758,
+      "step": 162690
+    },
+    {
+      "epoch": 8.08085824972683,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00015353531340021854,
+      "loss": 0.5031,
+      "step": 162700
+    },
+    {
+      "epoch": 8.081354922022449,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.000153495579616569,
+      "loss": 0.4785,
+      "step": 162710
+    },
+    {
+      "epoch": 8.081851594318069,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00015345584583291945,
+      "loss": 0.4417,
+      "step": 162720
+    },
+    {
+      "epoch": 8.082348266613689,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001534161120492699,
+      "loss": 0.5085,
+      "step": 162730
+    },
+    {
+      "epoch": 8.082844938909307,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00015337637826562034,
+      "loss": 0.4757,
+      "step": 162740
+    },
+    {
+      "epoch": 8.083341611204927,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001533366444819708,
+      "loss": 0.4616,
+      "step": 162750
+    },
+    {
+      "epoch": 8.083838283500546,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00015329691069832126,
+      "loss": 0.4805,
+      "step": 162760
+    },
+    {
+      "epoch": 8.084334955796166,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001532571769146717,
+      "loss": 0.5201,
+      "step": 162770
+    },
+    {
+      "epoch": 8.084831628091784,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.00015321744313102214,
+      "loss": 0.4898,
+      "step": 162780
+    },
+    {
+      "epoch": 8.085328300387404,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00015317770934737262,
+      "loss": 0.5156,
+      "step": 162790
+    },
+    {
+      "epoch": 8.085824972683024,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00015313797556372306,
+      "loss": 0.5232,
+      "step": 162800
+    },
+    {
+      "epoch": 8.086321644978643,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00015309824178007353,
+      "loss": 0.5013,
+      "step": 162810
+    },
+    {
+      "epoch": 8.086818317274263,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00015305850799642395,
+      "loss": 0.5049,
+      "step": 162820
+    },
+    {
+      "epoch": 8.087314989569881,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00015301877421277442,
+      "loss": 0.4755,
+      "step": 162830
+    },
+    {
+      "epoch": 8.087811661865501,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00015297904042912486,
+      "loss": 0.4835,
+      "step": 162840
+    },
+    {
+      "epoch": 8.08830833416112,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00015293930664547534,
+      "loss": 0.5212,
+      "step": 162850
+    },
+    {
+      "epoch": 8.08880500645674,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00015289957286182578,
+      "loss": 0.488,
+      "step": 162860
+    },
+    {
+      "epoch": 8.08930167875236,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00015285983907817622,
+      "loss": 0.5045,
+      "step": 162870
+    },
+    {
+      "epoch": 8.089798351047978,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001528201052945267,
+      "loss": 0.4862,
+      "step": 162880
+    },
+    {
+      "epoch": 8.090295023343598,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00015278037151087714,
+      "loss": 0.4815,
+      "step": 162890
+    },
+    {
+      "epoch": 8.090791695639217,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00015274063772722758,
+      "loss": 0.5271,
+      "step": 162900
+    },
+    {
+      "epoch": 8.091288367934837,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015270090394357803,
+      "loss": 0.5063,
+      "step": 162910
+    },
+    {
+      "epoch": 8.091785040230455,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001526611701599285,
+      "loss": 0.4708,
+      "step": 162920
+    },
+    {
+      "epoch": 8.092281712526075,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00015262143637627894,
+      "loss": 0.4968,
+      "step": 162930
+    },
+    {
+      "epoch": 8.092778384821695,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001525817025926294,
+      "loss": 0.5011,
+      "step": 162940
+    },
+    {
+      "epoch": 8.093275057117314,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00015254196880897983,
+      "loss": 0.488,
+      "step": 162950
+    },
+    {
+      "epoch": 8.093771729412934,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001525022350253303,
+      "loss": 0.4811,
+      "step": 162960
+    },
+    {
+      "epoch": 8.094268401708552,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00015246250124168075,
+      "loss": 0.4724,
+      "step": 162970
+    },
+    {
+      "epoch": 8.094765074004172,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0001524227674580312,
+      "loss": 0.5173,
+      "step": 162980
+    },
+    {
+      "epoch": 8.09526174629979,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00015238303367438164,
+      "loss": 0.4809,
+      "step": 162990
+    },
+    {
+      "epoch": 8.09575841859541,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001523432998907321,
+      "loss": 0.5081,
+      "step": 163000
+    },
+    {
+      "epoch": 8.096255090891031,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00015230356610708255,
+      "loss": 0.4998,
+      "step": 163010
+    },
+    {
+      "epoch": 8.09675176318665,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.000152263832323433,
+      "loss": 0.4874,
+      "step": 163020
+    },
+    {
+      "epoch": 8.09724843548227,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00015222409853978347,
+      "loss": 0.4991,
+      "step": 163030
+    },
+    {
+      "epoch": 8.097745107777888,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001521843647561339,
+      "loss": 0.4939,
+      "step": 163040
+    },
+    {
+      "epoch": 8.098241780073508,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00015214463097248436,
+      "loss": 0.4714,
+      "step": 163050
+    },
+    {
+      "epoch": 8.098738452369126,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001521048971888348,
+      "loss": 0.4791,
+      "step": 163060
+    },
+    {
+      "epoch": 8.099235124664746,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00015206516340518527,
+      "loss": 0.4741,
+      "step": 163070
+    },
+    {
+      "epoch": 8.099731796960366,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00015202542962153572,
+      "loss": 0.5234,
+      "step": 163080
+    },
+    {
+      "epoch": 8.100228469255985,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001519856958378862,
+      "loss": 0.5092,
+      "step": 163090
+    },
+    {
+      "epoch": 8.100725141551605,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001519459620542366,
+      "loss": 0.4809,
+      "step": 163100
+    },
+    {
+      "epoch": 8.101221813847223,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00015190622827058708,
+      "loss": 0.4739,
+      "step": 163110
+    },
+    {
+      "epoch": 8.101718486142843,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00015186649448693752,
+      "loss": 0.5185,
+      "step": 163120
+    },
+    {
+      "epoch": 8.102215158438462,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.000151826760703288,
+      "loss": 0.4776,
+      "step": 163130
+    },
+    {
+      "epoch": 8.102711830734082,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0001517870269196384,
+      "loss": 0.4871,
+      "step": 163140
+    },
+    {
+      "epoch": 8.103208503029702,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00015174729313598888,
+      "loss": 0.4446,
+      "step": 163150
+    },
+    {
+      "epoch": 8.10370517532532,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00015170755935233932,
+      "loss": 0.4787,
+      "step": 163160
+    },
+    {
+      "epoch": 8.10420184762094,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001516678255686898,
+      "loss": 0.4722,
+      "step": 163170
+    },
+    {
+      "epoch": 8.104698519916559,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00015162809178504024,
+      "loss": 0.5063,
+      "step": 163180
+    },
+    {
+      "epoch": 8.105195192212179,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00015158835800139068,
+      "loss": 0.5127,
+      "step": 163190
+    },
+    {
+      "epoch": 8.105691864507797,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00015154862421774113,
+      "loss": 0.4697,
+      "step": 163200
+    },
+    {
+      "epoch": 8.106188536803417,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001515088904340916,
+      "loss": 0.5029,
+      "step": 163210
+    },
+    {
+      "epoch": 8.106685209099037,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00015146915665044204,
+      "loss": 0.4469,
+      "step": 163220
+    },
+    {
+      "epoch": 8.107181881394656,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001514294228667925,
+      "loss": 0.4893,
+      "step": 163230
+    },
+    {
+      "epoch": 8.107678553690276,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00015138968908314296,
+      "loss": 0.5028,
+      "step": 163240
+    },
+    {
+      "epoch": 8.108175225985894,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001513499552994934,
+      "loss": 0.5145,
+      "step": 163250
+    },
+    {
+      "epoch": 8.108671898281514,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00015131022151584388,
+      "loss": 0.4838,
+      "step": 163260
+    },
+    {
+      "epoch": 8.109168570577133,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001512704877321943,
+      "loss": 0.4886,
+      "step": 163270
+    },
+    {
+      "epoch": 8.109665242872753,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00015123075394854476,
+      "loss": 0.4737,
+      "step": 163280
+    },
+    {
+      "epoch": 8.110161915168371,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001511910201648952,
+      "loss": 0.5064,
+      "step": 163290
+    },
+    {
+      "epoch": 8.110658587463991,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00015115128638124568,
+      "loss": 0.4509,
+      "step": 163300
+    },
+    {
+      "epoch": 8.111155259759611,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001511115525975961,
+      "loss": 0.4886,
+      "step": 163310
+    },
+    {
+      "epoch": 8.11165193205523,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00015107181881394657,
+      "loss": 0.4764,
+      "step": 163320
+    },
+    {
+      "epoch": 8.11214860435085,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.000151032085030297,
+      "loss": 0.4824,
+      "step": 163330
+    },
+    {
+      "epoch": 8.112645276646468,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00015099235124664748,
+      "loss": 0.4774,
+      "step": 163340
+    },
+    {
+      "epoch": 8.113141948942088,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0001509526174629979,
+      "loss": 0.4948,
+      "step": 163350
+    },
+    {
+      "epoch": 8.113638621237707,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00015091288367934837,
+      "loss": 0.4568,
+      "step": 163360
+    },
+    {
+      "epoch": 8.114135293533327,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00015087314989569882,
+      "loss": 0.4688,
+      "step": 163370
+    },
+    {
+      "epoch": 8.114631965828947,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001508334161120493,
+      "loss": 0.4855,
+      "step": 163380
+    },
+    {
+      "epoch": 8.115128638124565,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00015079368232839973,
+      "loss": 0.4662,
+      "step": 163390
+    },
+    {
+      "epoch": 8.115625310420185,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00015075394854475018,
+      "loss": 0.491,
+      "step": 163400
+    },
+    {
+      "epoch": 8.116121982715804,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00015071421476110065,
+      "loss": 0.4785,
+      "step": 163410
+    },
+    {
+      "epoch": 8.116618655011424,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001506744809774511,
+      "loss": 0.4968,
+      "step": 163420
+    },
+    {
+      "epoch": 8.117115327307042,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00015063474719380154,
+      "loss": 0.4812,
+      "step": 163430
+    },
+    {
+      "epoch": 8.117611999602662,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00015059501341015198,
+      "loss": 0.5179,
+      "step": 163440
+    },
+    {
+      "epoch": 8.118108671898282,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00015055527962650245,
+      "loss": 0.5029,
+      "step": 163450
+    },
+    {
+      "epoch": 8.1186053441939,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001505155458428529,
+      "loss": 0.5008,
+      "step": 163460
+    },
+    {
+      "epoch": 8.11910201648952,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00015047581205920334,
+      "loss": 0.5056,
+      "step": 163470
+    },
+    {
+      "epoch": 8.119598688785139,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00015043607827555378,
+      "loss": 0.4666,
+      "step": 163480
+    },
+    {
+      "epoch": 8.12009536108076,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00015039634449190426,
+      "loss": 0.5242,
+      "step": 163490
+    },
+    {
+      "epoch": 8.120592033376377,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001503566107082547,
+      "loss": 0.4416,
+      "step": 163500
+    },
+    {
+      "epoch": 8.121088705671998,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00015031687692460514,
+      "loss": 0.5143,
+      "step": 163510
+    },
+    {
+      "epoch": 8.121585377967618,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001502771431409556,
+      "loss": 0.4992,
+      "step": 163520
+    },
+    {
+      "epoch": 8.122082050263236,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00015023740935730606,
+      "loss": 0.4799,
+      "step": 163530
+    },
+    {
+      "epoch": 8.122578722558856,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001501976755736565,
+      "loss": 0.4654,
+      "step": 163540
+    },
+    {
+      "epoch": 8.123075394854474,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00015015794179000695,
+      "loss": 0.5201,
+      "step": 163550
+    },
+    {
+      "epoch": 8.123572067150095,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00015011820800635742,
+      "loss": 0.4963,
+      "step": 163560
+    },
+    {
+      "epoch": 8.124068739445713,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00015007847422270786,
+      "loss": 0.477,
+      "step": 163570
+    },
+    {
+      "epoch": 8.124565411741333,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00015003874043905834,
+      "loss": 0.4793,
+      "step": 163580
+    },
+    {
+      "epoch": 8.125062084036953,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00014999900665540875,
+      "loss": 0.4878,
+      "step": 163590
+    },
+    {
+      "epoch": 8.125558756332572,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00014995927287175922,
+      "loss": 0.4903,
+      "step": 163600
+    },
+    {
+      "epoch": 8.126055428628192,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00014991953908810967,
+      "loss": 0.4818,
+      "step": 163610
+    },
+    {
+      "epoch": 8.12655210092381,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00014987980530446014,
+      "loss": 0.48,
+      "step": 163620
+    },
+    {
+      "epoch": 8.12704877321943,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00014984007152081056,
+      "loss": 0.4938,
+      "step": 163630
+    },
+    {
+      "epoch": 8.127545445515048,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00014980033773716103,
+      "loss": 0.484,
+      "step": 163640
+    },
+    {
+      "epoch": 8.128042117810669,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00014976060395351147,
+      "loss": 0.4536,
+      "step": 163650
+    },
+    {
+      "epoch": 8.128538790106289,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00014972087016986194,
+      "loss": 0.4506,
+      "step": 163660
+    },
+    {
+      "epoch": 8.129035462401907,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001496811363862124,
+      "loss": 0.4585,
+      "step": 163670
+    },
+    {
+      "epoch": 8.129532134697527,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014964140260256283,
+      "loss": 0.4976,
+      "step": 163680
+    },
+    {
+      "epoch": 8.130028806993145,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00014960166881891328,
+      "loss": 0.4749,
+      "step": 163690
+    },
+    {
+      "epoch": 8.130525479288766,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00014956193503526375,
+      "loss": 0.4705,
+      "step": 163700
+    },
+    {
+      "epoch": 8.131022151584384,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0001495222012516142,
+      "loss": 0.4701,
+      "step": 163710
+    },
+    {
+      "epoch": 8.131518823880004,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014948246746796464,
+      "loss": 0.5111,
+      "step": 163720
+    },
+    {
+      "epoch": 8.132015496175624,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001494427336843151,
+      "loss": 0.4588,
+      "step": 163730
+    },
+    {
+      "epoch": 8.132512168471242,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00014940299990066555,
+      "loss": 0.4914,
+      "step": 163740
+    },
+    {
+      "epoch": 8.133008840766863,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.000149363266117016,
+      "loss": 0.5111,
+      "step": 163750
+    },
+    {
+      "epoch": 8.133505513062481,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00014932353233336644,
+      "loss": 0.529,
+      "step": 163760
+    },
+    {
+      "epoch": 8.134002185358101,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0001492837985497169,
+      "loss": 0.4805,
+      "step": 163770
+    },
+    {
+      "epoch": 8.13449885765372,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00014924406476606736,
+      "loss": 0.4843,
+      "step": 163780
+    },
+    {
+      "epoch": 8.13499552994934,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00014920433098241783,
+      "loss": 0.4816,
+      "step": 163790
+    },
+    {
+      "epoch": 8.13549220224496,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00014916459719876824,
+      "loss": 0.5106,
+      "step": 163800
+    },
+    {
+      "epoch": 8.135988874540578,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00014912486341511872,
+      "loss": 0.4676,
+      "step": 163810
+    },
+    {
+      "epoch": 8.136485546836198,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00014908512963146916,
+      "loss": 0.4662,
+      "step": 163820
+    },
+    {
+      "epoch": 8.136982219131816,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00014904539584781963,
+      "loss": 0.4769,
+      "step": 163830
+    },
+    {
+      "epoch": 8.137478891427437,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00014900566206417005,
+      "loss": 0.4501,
+      "step": 163840
+    },
+    {
+      "epoch": 8.137975563723055,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00014896592828052052,
+      "loss": 0.4867,
+      "step": 163850
+    },
+    {
+      "epoch": 8.138472236018675,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014892619449687096,
+      "loss": 0.4927,
+      "step": 163860
+    },
+    {
+      "epoch": 8.138968908314295,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014888646071322144,
+      "loss": 0.4846,
+      "step": 163870
+    },
+    {
+      "epoch": 8.139465580609913,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00014884672692957188,
+      "loss": 0.4848,
+      "step": 163880
+    },
+    {
+      "epoch": 8.139962252905534,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00014880699314592232,
+      "loss": 0.5082,
+      "step": 163890
+    },
+    {
+      "epoch": 8.140458925201152,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00014876725936227277,
+      "loss": 0.4832,
+      "step": 163900
+    },
+    {
+      "epoch": 8.140955597496772,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00014872752557862324,
+      "loss": 0.4672,
+      "step": 163910
+    },
+    {
+      "epoch": 8.14145226979239,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00014868779179497368,
+      "loss": 0.4833,
+      "step": 163920
+    },
+    {
+      "epoch": 8.14194894208801,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00014864805801132413,
+      "loss": 0.484,
+      "step": 163930
+    },
+    {
+      "epoch": 8.142445614383629,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001486083242276746,
+      "loss": 0.498,
+      "step": 163940
+    },
+    {
+      "epoch": 8.142942286679249,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00014856859044402504,
+      "loss": 0.4554,
+      "step": 163950
+    },
+    {
+      "epoch": 8.143438958974869,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001485288566603755,
+      "loss": 0.4805,
+      "step": 163960
+    },
+    {
+      "epoch": 8.143935631270487,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00014848912287672593,
+      "loss": 0.446,
+      "step": 163970
+    },
+    {
+      "epoch": 8.144432303566107,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001484493890930764,
+      "loss": 0.4562,
+      "step": 163980
+    },
+    {
+      "epoch": 8.144928975861726,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00014840965530942685,
+      "loss": 0.4906,
+      "step": 163990
+    },
+    {
+      "epoch": 8.145425648157346,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001483699215257773,
+      "loss": 0.4687,
+      "step": 164000
+    },
+    {
+      "epoch": 8.145922320452964,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014833018774212774,
+      "loss": 0.4506,
+      "step": 164010
+    },
+    {
+      "epoch": 8.146418992748584,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001482904539584782,
+      "loss": 0.4859,
+      "step": 164020
+    },
+    {
+      "epoch": 8.146915665044205,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00014825072017482865,
+      "loss": 0.493,
+      "step": 164030
+    },
+    {
+      "epoch": 8.147412337339823,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001482109863911791,
+      "loss": 0.4749,
+      "step": 164040
+    },
+    {
+      "epoch": 8.147909009635443,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00014817125260752954,
+      "loss": 0.4579,
+      "step": 164050
+    },
+    {
+      "epoch": 8.148405681931061,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00014813151882388,
+      "loss": 0.5064,
+      "step": 164060
+    },
+    {
+      "epoch": 8.148902354226681,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014809178504023046,
+      "loss": 0.4737,
+      "step": 164070
+    },
+    {
+      "epoch": 8.1493990265223,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00014805205125658093,
+      "loss": 0.5041,
+      "step": 164080
+    },
+    {
+      "epoch": 8.14989569881792,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00014801231747293137,
+      "loss": 0.49,
+      "step": 164090
+    },
+    {
+      "epoch": 8.15039237111354,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00014797258368928182,
+      "loss": 0.4585,
+      "step": 164100
+    },
+    {
+      "epoch": 8.150889043409158,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0001479328499056323,
+      "loss": 0.4613,
+      "step": 164110
+    },
+    {
+      "epoch": 8.151385715704778,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00014789311612198273,
+      "loss": 0.4807,
+      "step": 164120
+    },
+    {
+      "epoch": 8.151882388000397,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00014785338233833318,
+      "loss": 0.4522,
+      "step": 164130
+    },
+    {
+      "epoch": 8.152379060296017,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014781364855468362,
+      "loss": 0.4512,
+      "step": 164140
+    },
+    {
+      "epoch": 8.152875732591635,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001477739147710341,
+      "loss": 0.4485,
+      "step": 164150
+    },
+    {
+      "epoch": 8.153372404887255,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014773418098738454,
+      "loss": 0.5078,
+      "step": 164160
+    },
+    {
+      "epoch": 8.153869077182875,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00014769444720373498,
+      "loss": 0.4984,
+      "step": 164170
+    },
+    {
+      "epoch": 8.154365749478494,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.00014765471342008542,
+      "loss": 0.5128,
+      "step": 164180
+    },
+    {
+      "epoch": 8.154862421774114,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001476149796364359,
+      "loss": 0.4821,
+      "step": 164190
+    },
+    {
+      "epoch": 8.155359094069732,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014757524585278634,
+      "loss": 0.4764,
+      "step": 164200
+    },
+    {
+      "epoch": 8.155855766365352,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00014753551206913678,
+      "loss": 0.4622,
+      "step": 164210
+    },
+    {
+      "epoch": 8.15635243866097,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00014749577828548723,
+      "loss": 0.4858,
+      "step": 164220
+    },
+    {
+      "epoch": 8.15684911095659,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001474560445018377,
+      "loss": 0.4987,
+      "step": 164230
+    },
+    {
+      "epoch": 8.157345783252211,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00014741631071818814,
+      "loss": 0.4995,
+      "step": 164240
+    },
+    {
+      "epoch": 8.15784245554783,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001473765769345386,
+      "loss": 0.4922,
+      "step": 164250
+    },
+    {
+      "epoch": 8.15833912784345,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014733684315088906,
+      "loss": 0.5072,
+      "step": 164260
+    },
+    {
+      "epoch": 8.158835800139068,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001472971093672395,
+      "loss": 0.4662,
+      "step": 164270
+    },
+    {
+      "epoch": 8.159332472434688,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00014725737558358998,
+      "loss": 0.4912,
+      "step": 164280
+    },
+    {
+      "epoch": 8.159829144730306,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001472176417999404,
+      "loss": 0.4743,
+      "step": 164290
+    },
+    {
+      "epoch": 8.160325817025926,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00014717790801629086,
+      "loss": 0.5222,
+      "step": 164300
+    },
+    {
+      "epoch": 8.160822489321546,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001471381742326413,
+      "loss": 0.4899,
+      "step": 164310
+    },
+    {
+      "epoch": 8.161319161617165,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00014709844044899178,
+      "loss": 0.5119,
+      "step": 164320
+    },
+    {
+      "epoch": 8.161815833912785,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001470587066653422,
+      "loss": 0.4994,
+      "step": 164330
+    },
+    {
+      "epoch": 8.162312506208403,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00014701897288169267,
+      "loss": 0.4725,
+      "step": 164340
+    },
+    {
+      "epoch": 8.162809178504023,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001469792390980431,
+      "loss": 0.4845,
+      "step": 164350
+    },
+    {
+      "epoch": 8.163305850799642,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00014693950531439358,
+      "loss": 0.5011,
+      "step": 164360
+    },
+    {
+      "epoch": 8.163802523095262,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.000146899771530744,
+      "loss": 0.4875,
+      "step": 164370
+    },
+    {
+      "epoch": 8.164299195390882,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00014686003774709447,
+      "loss": 0.4793,
+      "step": 164380
+    },
+    {
+      "epoch": 8.1647958676865,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00014682030396344492,
+      "loss": 0.4925,
+      "step": 164390
+    },
+    {
+      "epoch": 8.16529253998212,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001467805701797954,
+      "loss": 0.5141,
+      "step": 164400
+    },
+    {
+      "epoch": 8.165789212277739,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00014674083639614583,
+      "loss": 0.4963,
+      "step": 164410
+    },
+    {
+      "epoch": 8.166285884573359,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00014670110261249628,
+      "loss": 0.4683,
+      "step": 164420
+    },
+    {
+      "epoch": 8.166782556868977,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014666136882884675,
+      "loss": 0.4896,
+      "step": 164430
+    },
+    {
+      "epoch": 8.167279229164597,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001466216350451972,
+      "loss": 0.4534,
+      "step": 164440
+    },
+    {
+      "epoch": 8.167775901460217,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00014658190126154764,
+      "loss": 0.4682,
+      "step": 164450
+    },
+    {
+      "epoch": 8.168272573755836,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014654216747789808,
+      "loss": 0.4751,
+      "step": 164460
+    },
+    {
+      "epoch": 8.168769246051456,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00014650243369424855,
+      "loss": 0.4642,
+      "step": 164470
+    },
+    {
+      "epoch": 8.169265918347074,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.000146462699910599,
+      "loss": 0.4879,
+      "step": 164480
+    },
+    {
+      "epoch": 8.169762590642694,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014642296612694947,
+      "loss": 0.497,
+      "step": 164490
+    },
+    {
+      "epoch": 8.170259262938313,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00014638323234329988,
+      "loss": 0.4932,
+      "step": 164500
+    },
+    {
+      "epoch": 8.170755935233933,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00014634349855965036,
+      "loss": 0.5398,
+      "step": 164510
+    },
+    {
+      "epoch": 8.171252607529553,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001463037647760008,
+      "loss": 0.4897,
+      "step": 164520
+    },
+    {
+      "epoch": 8.171749279825171,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00014626403099235127,
+      "loss": 0.4901,
+      "step": 164530
+    },
+    {
+      "epoch": 8.172245952120791,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001462242972087017,
+      "loss": 0.446,
+      "step": 164540
+    },
+    {
+      "epoch": 8.17274262441641,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00014618456342505216,
+      "loss": 0.4869,
+      "step": 164550
+    },
+    {
+      "epoch": 8.17323929671203,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001461448296414026,
+      "loss": 0.4803,
+      "step": 164560
+    },
+    {
+      "epoch": 8.173735969007648,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00014610509585775308,
+      "loss": 0.4773,
+      "step": 164570
+    },
+    {
+      "epoch": 8.174232641303268,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00014606536207410352,
+      "loss": 0.4384,
+      "step": 164580
+    },
+    {
+      "epoch": 8.174729313598888,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00014602562829045396,
+      "loss": 0.4849,
+      "step": 164590
+    },
+    {
+      "epoch": 8.175225985894507,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001459858945068044,
+      "loss": 0.4766,
+      "step": 164600
+    },
+    {
+      "epoch": 8.175722658190127,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00014594616072315488,
+      "loss": 0.4943,
+      "step": 164610
+    },
+    {
+      "epoch": 8.176219330485745,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00014590642693950532,
+      "loss": 0.4399,
+      "step": 164620
+    },
+    {
+      "epoch": 8.176716002781365,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00014586669315585577,
+      "loss": 0.4539,
+      "step": 164630
+    },
+    {
+      "epoch": 8.177212675076984,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00014582695937220624,
+      "loss": 0.5025,
+      "step": 164640
+    },
+    {
+      "epoch": 8.177709347372604,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00014578722558855668,
+      "loss": 0.5087,
+      "step": 164650
+    },
+    {
+      "epoch": 8.178206019668224,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014574749180490713,
+      "loss": 0.5102,
+      "step": 164660
+    },
+    {
+      "epoch": 8.178702691963842,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00014570775802125757,
+      "loss": 0.4862,
+      "step": 164670
+    },
+    {
+      "epoch": 8.179199364259462,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00014566802423760804,
+      "loss": 0.4795,
+      "step": 164680
+    },
+    {
+      "epoch": 8.17969603655508,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001456282904539585,
+      "loss": 0.5009,
+      "step": 164690
+    },
+    {
+      "epoch": 8.1801927088507,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014558855667030893,
+      "loss": 0.4772,
+      "step": 164700
+    },
+    {
+      "epoch": 8.180689381146319,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00014554882288665938,
+      "loss": 0.5138,
+      "step": 164710
+    },
+    {
+      "epoch": 8.18118605344194,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00014550908910300985,
+      "loss": 0.4987,
+      "step": 164720
+    },
+    {
+      "epoch": 8.181682725737558,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001454693553193603,
+      "loss": 0.4644,
+      "step": 164730
+    },
+    {
+      "epoch": 8.182179398033178,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00014542962153571074,
+      "loss": 0.481,
+      "step": 164740
+    },
+    {
+      "epoch": 8.182676070328798,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00014538988775206118,
+      "loss": 0.4189,
+      "step": 164750
+    },
+    {
+      "epoch": 8.183172742624416,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00014535015396841165,
+      "loss": 0.4587,
+      "step": 164760
+    },
+    {
+      "epoch": 8.183669414920036,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001453104201847621,
+      "loss": 0.4845,
+      "step": 164770
+    },
+    {
+      "epoch": 8.184166087215655,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00014527068640111254,
+      "loss": 0.4905,
+      "step": 164780
+    },
+    {
+      "epoch": 8.184662759511275,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.000145230952617463,
+      "loss": 0.4633,
+      "step": 164790
+    },
+    {
+      "epoch": 8.185159431806893,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00014519121883381346,
+      "loss": 0.4864,
+      "step": 164800
+    },
+    {
+      "epoch": 8.185656104102513,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00014515148505016393,
+      "loss": 0.501,
+      "step": 164810
+    },
+    {
+      "epoch": 8.186152776398133,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00014511175126651434,
+      "loss": 0.4567,
+      "step": 164820
+    },
+    {
+      "epoch": 8.186649448693752,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00014507201748286482,
+      "loss": 0.4716,
+      "step": 164830
+    },
+    {
+      "epoch": 8.187146120989372,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00014503228369921526,
+      "loss": 0.4943,
+      "step": 164840
+    },
+    {
+      "epoch": 8.18764279328499,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00014499254991556573,
+      "loss": 0.4988,
+      "step": 164850
+    },
+    {
+      "epoch": 8.18813946558061,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00014495281613191615,
+      "loss": 0.4557,
+      "step": 164860
+    },
+    {
+      "epoch": 8.188636137876228,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00014491308234826662,
+      "loss": 0.4808,
+      "step": 164870
+    },
+    {
+      "epoch": 8.189132810171849,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00014487334856461706,
+      "loss": 0.4897,
+      "step": 164880
+    },
+    {
+      "epoch": 8.189629482467469,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00014483361478096754,
+      "loss": 0.4593,
+      "step": 164890
+    },
+    {
+      "epoch": 8.190126154763087,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00014479388099731795,
+      "loss": 0.5139,
+      "step": 164900
+    },
+    {
+      "epoch": 8.190622827058707,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00014475414721366842,
+      "loss": 0.4925,
+      "step": 164910
+    },
+    {
+      "epoch": 8.191119499354325,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00014471441343001887,
+      "loss": 0.4642,
+      "step": 164920
+    },
+    {
+      "epoch": 8.191616171649946,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00014467467964636934,
+      "loss": 0.4647,
+      "step": 164930
+    },
+    {
+      "epoch": 8.192112843945564,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00014463494586271978,
+      "loss": 0.4486,
+      "step": 164940
+    },
+    {
+      "epoch": 8.192609516241184,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00014459521207907023,
+      "loss": 0.5145,
+      "step": 164950
+    },
+    {
+      "epoch": 8.193106188536804,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001445554782954207,
+      "loss": 0.4644,
+      "step": 164960
+    },
+    {
+      "epoch": 8.193602860832423,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00014451574451177114,
+      "loss": 0.4735,
+      "step": 164970
+    },
+    {
+      "epoch": 8.194099533128043,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00014447601072812161,
+      "loss": 0.4492,
+      "step": 164980
+    },
+    {
+      "epoch": 8.194596205423661,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00014443627694447203,
+      "loss": 0.4675,
+      "step": 164990
+    },
+    {
+      "epoch": 8.195092877719281,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001443965431608225,
+      "loss": 0.4816,
+      "step": 165000
+    },
+    {
+      "epoch": 8.1955895500149,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00014435680937717295,
+      "loss": 0.4675,
+      "step": 165010
+    },
+    {
+      "epoch": 8.19608622231052,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00014431707559352342,
+      "loss": 0.4893,
+      "step": 165020
+    },
+    {
+      "epoch": 8.19658289460614,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00014427734180987384,
+      "loss": 0.4923,
+      "step": 165030
+    },
+    {
+      "epoch": 8.197079566901758,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001442376080262243,
+      "loss": 0.5033,
+      "step": 165040
+    },
+    {
+      "epoch": 8.197576239197378,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00014419787424257475,
+      "loss": 0.454,
+      "step": 165050
+    },
+    {
+      "epoch": 8.198072911492996,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014415814045892522,
+      "loss": 0.4845,
+      "step": 165060
+    },
+    {
+      "epoch": 8.198569583788617,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00014411840667527564,
+      "loss": 0.4828,
+      "step": 165070
+    },
+    {
+      "epoch": 8.199066256084235,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001440786728916261,
+      "loss": 0.4931,
+      "step": 165080
+    },
+    {
+      "epoch": 8.199562928379855,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014403893910797656,
+      "loss": 0.5006,
+      "step": 165090
+    },
+    {
+      "epoch": 8.200059600675475,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014399920532432703,
+      "loss": 0.4702,
+      "step": 165100
+    },
+    {
+      "epoch": 8.200556272971093,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00014395947154067747,
+      "loss": 0.4977,
+      "step": 165110
+    },
+    {
+      "epoch": 8.201052945266714,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00014391973775702792,
+      "loss": 0.4561,
+      "step": 165120
+    },
+    {
+      "epoch": 8.201549617562332,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001438800039733784,
+      "loss": 0.5116,
+      "step": 165130
+    },
+    {
+      "epoch": 8.202046289857952,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00014384027018972883,
+      "loss": 0.4462,
+      "step": 165140
+    },
+    {
+      "epoch": 8.20254296215357,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00014380053640607928,
+      "loss": 0.4528,
+      "step": 165150
+    },
+    {
+      "epoch": 8.20303963444919,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00014376080262242972,
+      "loss": 0.4725,
+      "step": 165160
+    },
+    {
+      "epoch": 8.20353630674481,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001437210688387802,
+      "loss": 0.483,
+      "step": 165170
+    },
+    {
+      "epoch": 8.204032979040429,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00014368133505513064,
+      "loss": 0.4928,
+      "step": 165180
+    },
+    {
+      "epoch": 8.204529651336049,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00014364160127148108,
+      "loss": 0.4657,
+      "step": 165190
+    },
+    {
+      "epoch": 8.205026323631667,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00014360186748783152,
+      "loss": 0.4717,
+      "step": 165200
+    },
+    {
+      "epoch": 8.205522995927288,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.000143562133704182,
+      "loss": 0.457,
+      "step": 165210
+    },
+    {
+      "epoch": 8.206019668222906,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00014352239992053244,
+      "loss": 0.5057,
+      "step": 165220
+    },
+    {
+      "epoch": 8.206516340518526,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00014348266613688288,
+      "loss": 0.4999,
+      "step": 165230
+    },
+    {
+      "epoch": 8.207013012814146,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00014344293235323333,
+      "loss": 0.4944,
+      "step": 165240
+    },
+    {
+      "epoch": 8.207509685109764,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001434031985695838,
+      "loss": 0.4968,
+      "step": 165250
+    },
+    {
+      "epoch": 8.208006357405385,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00014336346478593424,
+      "loss": 0.4919,
+      "step": 165260
+    },
+    {
+      "epoch": 8.208503029701003,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0001433237310022847,
+      "loss": 0.5149,
+      "step": 165270
+    },
+    {
+      "epoch": 8.208999701996623,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00014328399721863516,
+      "loss": 0.4786,
+      "step": 165280
+    },
+    {
+      "epoch": 8.209496374292241,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001432442634349856,
+      "loss": 0.5123,
+      "step": 165290
+    },
+    {
+      "epoch": 8.209993046587861,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014320452965133605,
+      "loss": 0.4706,
+      "step": 165300
+    },
+    {
+      "epoch": 8.21048971888348,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001431647958676865,
+      "loss": 0.5235,
+      "step": 165310
+    },
+    {
+      "epoch": 8.2109863911791,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00014312506208403696,
+      "loss": 0.4814,
+      "step": 165320
+    },
+    {
+      "epoch": 8.21148306347472,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001430853283003874,
+      "loss": 0.464,
+      "step": 165330
+    },
+    {
+      "epoch": 8.211979735770338,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00014304559451673788,
+      "loss": 0.5068,
+      "step": 165340
+    },
+    {
+      "epoch": 8.212476408065958,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00014300586073308832,
+      "loss": 0.5117,
+      "step": 165350
+    },
+    {
+      "epoch": 8.212973080361577,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00014296612694943877,
+      "loss": 0.4758,
+      "step": 165360
+    },
+    {
+      "epoch": 8.213469752657197,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001429263931657892,
+      "loss": 0.4912,
+      "step": 165370
+    },
+    {
+      "epoch": 8.213966424952815,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00014288665938213968,
+      "loss": 0.476,
+      "step": 165380
+    },
+    {
+      "epoch": 8.214463097248435,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00014284692559849013,
+      "loss": 0.5403,
+      "step": 165390
+    },
+    {
+      "epoch": 8.214959769544055,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00014280719181484057,
+      "loss": 0.4758,
+      "step": 165400
+    },
+    {
+      "epoch": 8.215456441839674,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00014276745803119102,
+      "loss": 0.4988,
+      "step": 165410
+    },
+    {
+      "epoch": 8.215953114135294,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001427277242475415,
+      "loss": 0.4871,
+      "step": 165420
+    },
+    {
+      "epoch": 8.216449786430912,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00014268799046389193,
+      "loss": 0.4574,
+      "step": 165430
+    },
+    {
+      "epoch": 8.216946458726532,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00014264825668024238,
+      "loss": 0.4665,
+      "step": 165440
+    },
+    {
+      "epoch": 8.21744313102215,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00014260852289659282,
+      "loss": 0.4975,
+      "step": 165450
+    },
+    {
+      "epoch": 8.21793980331777,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001425687891129433,
+      "loss": 0.479,
+      "step": 165460
+    },
+    {
+      "epoch": 8.218436475613391,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00014252905532929374,
+      "loss": 0.4817,
+      "step": 165470
+    },
+    {
+      "epoch": 8.21893314790901,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00014248932154564418,
+      "loss": 0.4735,
+      "step": 165480
+    },
+    {
+      "epoch": 8.21942982020463,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00014244958776199465,
+      "loss": 0.4687,
+      "step": 165490
+    },
+    {
+      "epoch": 8.219926492500248,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0001424098539783451,
+      "loss": 0.4996,
+      "step": 165500
+    },
+    {
+      "epoch": 8.220423164795868,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00014237012019469557,
+      "loss": 0.4872,
+      "step": 165510
+    },
+    {
+      "epoch": 8.220919837091486,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00014233038641104598,
+      "loss": 0.5011,
+      "step": 165520
+    },
+    {
+      "epoch": 8.221416509387106,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00014229065262739646,
+      "loss": 0.4914,
+      "step": 165530
+    },
+    {
+      "epoch": 8.221913181682726,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001422509188437469,
+      "loss": 0.4813,
+      "step": 165540
+    },
+    {
+      "epoch": 8.222409853978345,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00014221118506009737,
+      "loss": 0.4656,
+      "step": 165550
+    },
+    {
+      "epoch": 8.222906526273965,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001421714512764478,
+      "loss": 0.4844,
+      "step": 165560
+    },
+    {
+      "epoch": 8.223403198569583,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.00014213171749279826,
+      "loss": 0.5205,
+      "step": 165570
+    },
+    {
+      "epoch": 8.223899870865203,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001420919837091487,
+      "loss": 0.4786,
+      "step": 165580
+    },
+    {
+      "epoch": 8.224396543160822,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014205224992549917,
+      "loss": 0.5083,
+      "step": 165590
+    },
+    {
+      "epoch": 8.224893215456442,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001420125161418496,
+      "loss": 0.4834,
+      "step": 165600
+    },
+    {
+      "epoch": 8.225389887752062,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00014197278235820006,
+      "loss": 0.4658,
+      "step": 165610
+    },
+    {
+      "epoch": 8.22588656004768,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001419330485745505,
+      "loss": 0.5276,
+      "step": 165620
+    },
+    {
+      "epoch": 8.2263832323433,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00014189331479090098,
+      "loss": 0.4855,
+      "step": 165630
+    },
+    {
+      "epoch": 8.226879904638919,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00014185358100725142,
+      "loss": 0.4836,
+      "step": 165640
+    },
+    {
+      "epoch": 8.227376576934539,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014181384722360187,
+      "loss": 0.4809,
+      "step": 165650
+    },
+    {
+      "epoch": 8.227873249230157,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00014177411343995234,
+      "loss": 0.498,
+      "step": 165660
+    },
+    {
+      "epoch": 8.228369921525777,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00014173437965630278,
+      "loss": 0.5017,
+      "step": 165670
+    },
+    {
+      "epoch": 8.228866593821397,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00014169464587265323,
+      "loss": 0.4874,
+      "step": 165680
+    },
+    {
+      "epoch": 8.229363266117016,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014165491208900367,
+      "loss": 0.4943,
+      "step": 165690
+    },
+    {
+      "epoch": 8.229859938412636,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00014161517830535414,
+      "loss": 0.4968,
+      "step": 165700
+    },
+    {
+      "epoch": 8.230356610708254,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001415754445217046,
+      "loss": 0.4797,
+      "step": 165710
+    },
+    {
+      "epoch": 8.230853283003874,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00014153571073805503,
+      "loss": 0.4862,
+      "step": 165720
+    },
+    {
+      "epoch": 8.231349955299493,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00014149597695440548,
+      "loss": 0.4464,
+      "step": 165730
+    },
+    {
+      "epoch": 8.231846627595113,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00014145624317075595,
+      "loss": 0.4743,
+      "step": 165740
+    },
+    {
+      "epoch": 8.232343299890733,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001414165093871064,
+      "loss": 0.5065,
+      "step": 165750
+    },
+    {
+      "epoch": 8.232839972186351,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00014137677560345686,
+      "loss": 0.4617,
+      "step": 165760
+    },
+    {
+      "epoch": 8.233336644481971,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00014133704181980728,
+      "loss": 0.494,
+      "step": 165770
+    },
+    {
+      "epoch": 8.23383331677759,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00014129730803615775,
+      "loss": 0.4749,
+      "step": 165780
+    },
+    {
+      "epoch": 8.23432998907321,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001412575742525082,
+      "loss": 0.467,
+      "step": 165790
+    },
+    {
+      "epoch": 8.234826661368828,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00014121784046885867,
+      "loss": 0.4945,
+      "step": 165800
+    },
+    {
+      "epoch": 8.235323333664448,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001411781066852091,
+      "loss": 0.5129,
+      "step": 165810
+    },
+    {
+      "epoch": 8.235820005960068,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.00014113837290155956,
+      "loss": 0.5185,
+      "step": 165820
+    },
+    {
+      "epoch": 8.236316678255687,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00014109863911791003,
+      "loss": 0.5067,
+      "step": 165830
+    },
+    {
+      "epoch": 8.236813350551307,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00014105890533426047,
+      "loss": 0.4612,
+      "step": 165840
+    },
+    {
+      "epoch": 8.237310022846925,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014101917155061092,
+      "loss": 0.5043,
+      "step": 165850
+    },
+    {
+      "epoch": 8.237806695142545,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00014097943776696136,
+      "loss": 0.4934,
+      "step": 165860
+    },
+    {
+      "epoch": 8.238303367438164,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014093970398331183,
+      "loss": 0.495,
+      "step": 165870
+    },
+    {
+      "epoch": 8.238800039733784,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00014089997019966228,
+      "loss": 0.4677,
+      "step": 165880
+    },
+    {
+      "epoch": 8.239296712029404,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00014086023641601272,
+      "loss": 0.4895,
+      "step": 165890
+    },
+    {
+      "epoch": 8.239793384325022,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00014082050263236316,
+      "loss": 0.4675,
+      "step": 165900
+    },
+    {
+      "epoch": 8.240290056620642,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014078076884871363,
+      "loss": 0.4758,
+      "step": 165910
+    },
+    {
+      "epoch": 8.24078672891626,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00014074103506506408,
+      "loss": 0.4697,
+      "step": 165920
+    },
+    {
+      "epoch": 8.24128340121188,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014070130128141452,
+      "loss": 0.4968,
+      "step": 165930
+    },
+    {
+      "epoch": 8.241780073507499,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00014066156749776497,
+      "loss": 0.477,
+      "step": 165940
+    },
+    {
+      "epoch": 8.24227674580312,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00014062183371411544,
+      "loss": 0.4564,
+      "step": 165950
+    },
+    {
+      "epoch": 8.24277341809874,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014058209993046588,
+      "loss": 0.4658,
+      "step": 165960
+    },
+    {
+      "epoch": 8.243270090394358,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00014054236614681633,
+      "loss": 0.4952,
+      "step": 165970
+    },
+    {
+      "epoch": 8.243766762689978,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0001405026323631668,
+      "loss": 0.4651,
+      "step": 165980
+    },
+    {
+      "epoch": 8.244263434985596,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.00014046289857951724,
+      "loss": 0.4815,
+      "step": 165990
+    },
+    {
+      "epoch": 8.244760107281216,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00014042316479586771,
+      "loss": 0.5055,
+      "step": 166000
+    },
+    {
+      "epoch": 8.245256779576835,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00014038343101221813,
+      "loss": 0.5135,
+      "step": 166010
+    },
+    {
+      "epoch": 8.245753451872455,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001403436972285686,
+      "loss": 0.52,
+      "step": 166020
+    },
+    {
+      "epoch": 8.246250124168075,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00014030396344491905,
+      "loss": 0.4857,
+      "step": 166030
+    },
+    {
+      "epoch": 8.246746796463693,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00014026422966126952,
+      "loss": 0.5122,
+      "step": 166040
+    },
+    {
+      "epoch": 8.247243468759313,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00014022449587761994,
+      "loss": 0.489,
+      "step": 166050
+    },
+    {
+      "epoch": 8.247740141054932,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001401847620939704,
+      "loss": 0.4951,
+      "step": 166060
+    },
+    {
+      "epoch": 8.248236813350552,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00014014502831032085,
+      "loss": 0.5143,
+      "step": 166070
+    },
+    {
+      "epoch": 8.24873348564617,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00014010529452667132,
+      "loss": 0.4649,
+      "step": 166080
+    },
+    {
+      "epoch": 8.24923015794179,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00014006556074302174,
+      "loss": 0.4765,
+      "step": 166090
+    },
+    {
+      "epoch": 8.249726830237408,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001400258269593722,
+      "loss": 0.4906,
+      "step": 166100
+    },
+    {
+      "epoch": 8.250223502533029,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00013998609317572266,
+      "loss": 0.461,
+      "step": 166110
+    },
+    {
+      "epoch": 8.250720174828649,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00013994635939207313,
+      "loss": 0.4916,
+      "step": 166120
+    },
+    {
+      "epoch": 8.251216847124267,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00013990662560842357,
+      "loss": 0.478,
+      "step": 166130
+    },
+    {
+      "epoch": 8.251713519419887,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00013986689182477402,
+      "loss": 0.5184,
+      "step": 166140
+    },
+    {
+      "epoch": 8.252210191715506,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00013982715804112446,
+      "loss": 0.4736,
+      "step": 166150
+    },
+    {
+      "epoch": 8.252706864011126,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00013978742425747493,
+      "loss": 0.4896,
+      "step": 166160
+    },
+    {
+      "epoch": 8.253203536306744,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00013974769047382538,
+      "loss": 0.4874,
+      "step": 166170
+    },
+    {
+      "epoch": 8.253700208602364,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00013970795669017582,
+      "loss": 0.46,
+      "step": 166180
+    },
+    {
+      "epoch": 8.254196880897984,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001396682229065263,
+      "loss": 0.486,
+      "step": 166190
+    },
+    {
+      "epoch": 8.254693553193603,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00013962848912287673,
+      "loss": 0.4872,
+      "step": 166200
+    },
+    {
+      "epoch": 8.255190225489223,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0001395887553392272,
+      "loss": 0.5024,
+      "step": 166210
+    },
+    {
+      "epoch": 8.255686897784841,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00013954902155557762,
+      "loss": 0.4826,
+      "step": 166220
+    },
+    {
+      "epoch": 8.256183570080461,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001395092877719281,
+      "loss": 0.4846,
+      "step": 166230
+    },
+    {
+      "epoch": 8.25668024237608,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00013946955398827854,
+      "loss": 0.4903,
+      "step": 166240
+    },
+    {
+      "epoch": 8.2571769146717,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.000139429820204629,
+      "loss": 0.5027,
+      "step": 166250
+    },
+    {
+      "epoch": 8.25767358696732,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00013939008642097943,
+      "loss": 0.4871,
+      "step": 166260
+    },
+    {
+      "epoch": 8.258170259262938,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001393503526373299,
+      "loss": 0.481,
+      "step": 166270
+    },
+    {
+      "epoch": 8.258666931558558,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013931061885368034,
+      "loss": 0.4921,
+      "step": 166280
+    },
+    {
+      "epoch": 8.259163603854176,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013927088507003081,
+      "loss": 0.491,
+      "step": 166290
+    },
+    {
+      "epoch": 8.259660276149797,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00013923115128638123,
+      "loss": 0.4899,
+      "step": 166300
+    },
+    {
+      "epoch": 8.260156948445415,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001391914175027317,
+      "loss": 0.4881,
+      "step": 166310
+    },
+    {
+      "epoch": 8.260653620741035,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00013915168371908215,
+      "loss": 0.4838,
+      "step": 166320
+    },
+    {
+      "epoch": 8.261150293036655,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00013911194993543262,
+      "loss": 0.4698,
+      "step": 166330
+    },
+    {
+      "epoch": 8.261646965332273,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00013907221615178306,
+      "loss": 0.4901,
+      "step": 166340
+    },
+    {
+      "epoch": 8.262143637627894,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001390324823681335,
+      "loss": 0.46,
+      "step": 166350
+    },
+    {
+      "epoch": 8.262640309923512,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00013899274858448398,
+      "loss": 0.5303,
+      "step": 166360
+    },
+    {
+      "epoch": 8.263136982219132,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00013895301480083442,
+      "loss": 0.5084,
+      "step": 166370
+    },
+    {
+      "epoch": 8.26363365451475,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00013891328101718487,
+      "loss": 0.5177,
+      "step": 166380
+    },
+    {
+      "epoch": 8.26413032681037,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001388735472335353,
+      "loss": 0.4443,
+      "step": 166390
+    },
+    {
+      "epoch": 8.26462699910599,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013883381344988578,
+      "loss": 0.4598,
+      "step": 166400
+    },
+    {
+      "epoch": 8.265123671401609,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00013879407966623623,
+      "loss": 0.482,
+      "step": 166410
+    },
+    {
+      "epoch": 8.265620343697229,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00013875434588258667,
+      "loss": 0.5358,
+      "step": 166420
+    },
+    {
+      "epoch": 8.266117015992847,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00013871461209893712,
+      "loss": 0.4901,
+      "step": 166430
+    },
+    {
+      "epoch": 8.266613688288468,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001386748783152876,
+      "loss": 0.5012,
+      "step": 166440
+    },
+    {
+      "epoch": 8.267110360584086,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00013863514453163803,
+      "loss": 0.5218,
+      "step": 166450
+    },
+    {
+      "epoch": 8.267607032879706,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00013859541074798848,
+      "loss": 0.4715,
+      "step": 166460
+    },
+    {
+      "epoch": 8.268103705175326,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00013855567696433892,
+      "loss": 0.4567,
+      "step": 166470
+    },
+    {
+      "epoch": 8.268600377470944,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001385159431806894,
+      "loss": 0.4753,
+      "step": 166480
+    },
+    {
+      "epoch": 8.269097049766565,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013847620939703984,
+      "loss": 0.4756,
+      "step": 166490
+    },
+    {
+      "epoch": 8.269593722062183,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00013843647561339028,
+      "loss": 0.4921,
+      "step": 166500
+    },
+    {
+      "epoch": 8.270090394357803,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00013839674182974075,
+      "loss": 0.4789,
+      "step": 166510
+    },
+    {
+      "epoch": 8.270587066653421,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001383570080460912,
+      "loss": 0.4963,
+      "step": 166520
+    },
+    {
+      "epoch": 8.271083738949041,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00013831727426244167,
+      "loss": 0.485,
+      "step": 166530
+    },
+    {
+      "epoch": 8.271580411244662,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00013827754047879208,
+      "loss": 0.4849,
+      "step": 166540
+    },
+    {
+      "epoch": 8.27207708354028,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00013823780669514255,
+      "loss": 0.4716,
+      "step": 166550
+    },
+    {
+      "epoch": 8.2725737558359,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.000138198072911493,
+      "loss": 0.5257,
+      "step": 166560
+    },
+    {
+      "epoch": 8.273070428131518,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00013815833912784347,
+      "loss": 0.5066,
+      "step": 166570
+    },
+    {
+      "epoch": 8.273567100427138,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001381186053441939,
+      "loss": 0.4887,
+      "step": 166580
+    },
+    {
+      "epoch": 8.274063772722757,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00013807887156054436,
+      "loss": 0.4754,
+      "step": 166590
+    },
+    {
+      "epoch": 8.274560445018377,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001380391377768948,
+      "loss": 0.4981,
+      "step": 166600
+    },
+    {
+      "epoch": 8.275057117313997,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00013799940399324527,
+      "loss": 0.4992,
+      "step": 166610
+    },
+    {
+      "epoch": 8.275553789609615,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00013795967020959572,
+      "loss": 0.4977,
+      "step": 166620
+    },
+    {
+      "epoch": 8.276050461905236,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00013791993642594616,
+      "loss": 0.464,
+      "step": 166630
+    },
+    {
+      "epoch": 8.276547134200854,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001378802026422966,
+      "loss": 0.475,
+      "step": 166640
+    },
+    {
+      "epoch": 8.277043806496474,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00013784046885864708,
+      "loss": 0.5013,
+      "step": 166650
+    },
+    {
+      "epoch": 8.277540478792092,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013780073507499752,
+      "loss": 0.4751,
+      "step": 166660
+    },
+    {
+      "epoch": 8.278037151087712,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00013776100129134797,
+      "loss": 0.4773,
+      "step": 166670
+    },
+    {
+      "epoch": 8.27853382338333,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00013772126750769844,
+      "loss": 0.4894,
+      "step": 166680
+    },
+    {
+      "epoch": 8.27903049567895,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00013768153372404888,
+      "loss": 0.4941,
+      "step": 166690
+    },
+    {
+      "epoch": 8.279527167974571,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00013764179994039935,
+      "loss": 0.5007,
+      "step": 166700
+    },
+    {
+      "epoch": 8.28002384027019,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00013760206615674977,
+      "loss": 0.4881,
+      "step": 166710
+    },
+    {
+      "epoch": 8.28052051256581,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00013756233237310024,
+      "loss": 0.4957,
+      "step": 166720
+    },
+    {
+      "epoch": 8.281017184861428,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001375225985894507,
+      "loss": 0.4816,
+      "step": 166730
+    },
+    {
+      "epoch": 8.281513857157048,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00013748286480580116,
+      "loss": 0.4573,
+      "step": 166740
+    },
+    {
+      "epoch": 8.282010529452666,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00013744313102215158,
+      "loss": 0.4782,
+      "step": 166750
+    },
+    {
+      "epoch": 8.282507201748286,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013740339723850205,
+      "loss": 0.508,
+      "step": 166760
+    },
+    {
+      "epoch": 8.283003874043906,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001373636634548525,
+      "loss": 0.486,
+      "step": 166770
+    },
+    {
+      "epoch": 8.283500546339525,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00013732392967120296,
+      "loss": 0.4679,
+      "step": 166780
+    },
+    {
+      "epoch": 8.283997218635145,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013728419588755338,
+      "loss": 0.4881,
+      "step": 166790
+    },
+    {
+      "epoch": 8.284493890930763,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00013724446210390385,
+      "loss": 0.4946,
+      "step": 166800
+    },
+    {
+      "epoch": 8.284990563226383,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001372047283202543,
+      "loss": 0.5021,
+      "step": 166810
+    },
+    {
+      "epoch": 8.285487235522002,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00013716499453660477,
+      "loss": 0.4723,
+      "step": 166820
+    },
+    {
+      "epoch": 8.285983907817622,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001371252607529552,
+      "loss": 0.4937,
+      "step": 166830
+    },
+    {
+      "epoch": 8.286480580113242,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00013708552696930565,
+      "loss": 0.4536,
+      "step": 166840
+    },
+    {
+      "epoch": 8.28697725240886,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00013704579318565613,
+      "loss": 0.5034,
+      "step": 166850
+    },
+    {
+      "epoch": 8.28747392470448,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00013700605940200657,
+      "loss": 0.475,
+      "step": 166860
+    },
+    {
+      "epoch": 8.287970597000099,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00013696632561835701,
+      "loss": 0.4466,
+      "step": 166870
+    },
+    {
+      "epoch": 8.288467269295719,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00013692659183470746,
+      "loss": 0.4629,
+      "step": 166880
+    },
+    {
+      "epoch": 8.288963941591337,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00013688685805105793,
+      "loss": 0.4828,
+      "step": 166890
+    },
+    {
+      "epoch": 8.289460613886957,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00013684712426740837,
+      "loss": 0.4674,
+      "step": 166900
+    },
+    {
+      "epoch": 8.289957286182577,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00013680739048375882,
+      "loss": 0.4985,
+      "step": 166910
+    },
+    {
+      "epoch": 8.290453958478196,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00013676765670010926,
+      "loss": 0.462,
+      "step": 166920
+    },
+    {
+      "epoch": 8.290950630773816,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00013672792291645973,
+      "loss": 0.4734,
+      "step": 166930
+    },
+    {
+      "epoch": 8.291447303069434,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00013668818913281018,
+      "loss": 0.4651,
+      "step": 166940
+    },
+    {
+      "epoch": 8.291943975365054,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00013664845534916062,
+      "loss": 0.473,
+      "step": 166950
+    },
+    {
+      "epoch": 8.292440647660673,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00013660872156551107,
+      "loss": 0.5003,
+      "step": 166960
+    },
+    {
+      "epoch": 8.292937319956293,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00013656898778186154,
+      "loss": 0.4831,
+      "step": 166970
+    },
+    {
+      "epoch": 8.293433992251913,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00013652925399821198,
+      "loss": 0.4685,
+      "step": 166980
+    },
+    {
+      "epoch": 8.293930664547531,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00013648952021456243,
+      "loss": 0.4819,
+      "step": 166990
+    },
+    {
+      "epoch": 8.294427336843151,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00013644978643091287,
+      "loss": 0.4569,
+      "step": 167000
+    },
+    {
+      "epoch": 8.29492400913877,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00013641005264726334,
+      "loss": 0.4882,
+      "step": 167010
+    },
+    {
+      "epoch": 8.29542068143439,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001363703188636138,
+      "loss": 0.4702,
+      "step": 167020
+    },
+    {
+      "epoch": 8.295917353730008,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00013633058507996426,
+      "loss": 0.4986,
+      "step": 167030
+    },
+    {
+      "epoch": 8.296414026025628,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001362908512963147,
+      "loss": 0.4749,
+      "step": 167040
+    },
+    {
+      "epoch": 8.296910698321248,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00013625111751266515,
+      "loss": 0.4875,
+      "step": 167050
+    },
+    {
+      "epoch": 8.297407370616867,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00013621138372901562,
+      "loss": 0.4534,
+      "step": 167060
+    },
+    {
+      "epoch": 8.297904042912487,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00013617164994536606,
+      "loss": 0.5031,
+      "step": 167070
+    },
+    {
+      "epoch": 8.298400715208105,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001361319161617165,
+      "loss": 0.4927,
+      "step": 167080
+    },
+    {
+      "epoch": 8.298897387503725,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013609218237806695,
+      "loss": 0.4837,
+      "step": 167090
+    },
+    {
+      "epoch": 8.299394059799344,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00013605244859441742,
+      "loss": 0.4649,
+      "step": 167100
+    },
+    {
+      "epoch": 8.299890732094964,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013601271481076787,
+      "loss": 0.4737,
+      "step": 167110
+    },
+    {
+      "epoch": 8.300387404390584,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001359729810271183,
+      "loss": 0.4946,
+      "step": 167120
+    },
+    {
+      "epoch": 8.300884076686202,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00013593324724346876,
+      "loss": 0.4786,
+      "step": 167130
+    },
+    {
+      "epoch": 8.301380748981822,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00013589351345981923,
+      "loss": 0.4839,
+      "step": 167140
+    },
+    {
+      "epoch": 8.30187742127744,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00013585377967616967,
+      "loss": 0.4802,
+      "step": 167150
+    },
+    {
+      "epoch": 8.30237409357306,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00013581404589252011,
+      "loss": 0.5023,
+      "step": 167160
+    },
+    {
+      "epoch": 8.30287076586868,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00013577431210887056,
+      "loss": 0.497,
+      "step": 167170
+    },
+    {
+      "epoch": 8.3033674381643,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00013573457832522103,
+      "loss": 0.4474,
+      "step": 167180
+    },
+    {
+      "epoch": 8.30386411045992,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00013569484454157147,
+      "loss": 0.5103,
+      "step": 167190
+    },
+    {
+      "epoch": 8.304360782755538,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00013565511075792192,
+      "loss": 0.503,
+      "step": 167200
+    },
+    {
+      "epoch": 8.304857455051158,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001356153769742724,
+      "loss": 0.5081,
+      "step": 167210
+    },
+    {
+      "epoch": 8.305354127346776,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00013557564319062283,
+      "loss": 0.494,
+      "step": 167220
+    },
+    {
+      "epoch": 8.305850799642396,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001355359094069733,
+      "loss": 0.4958,
+      "step": 167230
+    },
+    {
+      "epoch": 8.306347471938015,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00013549617562332372,
+      "loss": 0.4953,
+      "step": 167240
+    },
+    {
+      "epoch": 8.306844144233635,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001354564418396742,
+      "loss": 0.5154,
+      "step": 167250
+    },
+    {
+      "epoch": 8.307340816529255,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013541670805602464,
+      "loss": 0.4744,
+      "step": 167260
+    },
+    {
+      "epoch": 8.307837488824873,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001353769742723751,
+      "loss": 0.4729,
+      "step": 167270
+    },
+    {
+      "epoch": 8.308334161120493,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00013533724048872553,
+      "loss": 0.5144,
+      "step": 167280
+    },
+    {
+      "epoch": 8.308830833416112,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.000135297506705076,
+      "loss": 0.4924,
+      "step": 167290
+    },
+    {
+      "epoch": 8.309327505711732,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013525777292142644,
+      "loss": 0.4912,
+      "step": 167300
+    },
+    {
+      "epoch": 8.30982417800735,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00013521803913777691,
+      "loss": 0.4705,
+      "step": 167310
+    },
+    {
+      "epoch": 8.31032085030297,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00013517830535412733,
+      "loss": 0.5122,
+      "step": 167320
+    },
+    {
+      "epoch": 8.31081752259859,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001351385715704778,
+      "loss": 0.4926,
+      "step": 167330
+    },
+    {
+      "epoch": 8.311314194894209,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00013509883778682825,
+      "loss": 0.4612,
+      "step": 167340
+    },
+    {
+      "epoch": 8.311810867189829,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00013505910400317872,
+      "loss": 0.4623,
+      "step": 167350
+    },
+    {
+      "epoch": 8.312307539485447,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00013501937021952916,
+      "loss": 0.4782,
+      "step": 167360
+    },
+    {
+      "epoch": 8.312804211781067,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001349796364358796,
+      "loss": 0.4786,
+      "step": 167370
+    },
+    {
+      "epoch": 8.313300884076686,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00013493990265223008,
+      "loss": 0.4938,
+      "step": 167380
+    },
+    {
+      "epoch": 8.313797556372306,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00013490016886858052,
+      "loss": 0.4889,
+      "step": 167390
+    },
+    {
+      "epoch": 8.314294228667926,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00013486043508493097,
+      "loss": 0.528,
+      "step": 167400
+    },
+    {
+      "epoch": 8.314790900963544,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001348207013012814,
+      "loss": 0.4834,
+      "step": 167410
+    },
+    {
+      "epoch": 8.315287573259164,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00013478096751763188,
+      "loss": 0.4942,
+      "step": 167420
+    },
+    {
+      "epoch": 8.315784245554783,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00013474123373398233,
+      "loss": 0.4927,
+      "step": 167430
+    },
+    {
+      "epoch": 8.316280917850403,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001347014999503328,
+      "loss": 0.482,
+      "step": 167440
+    },
+    {
+      "epoch": 8.316777590146021,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00013466176616668321,
+      "loss": 0.4717,
+      "step": 167450
+    },
+    {
+      "epoch": 8.317274262441641,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013462203238303369,
+      "loss": 0.4825,
+      "step": 167460
+    },
+    {
+      "epoch": 8.317770934737261,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00013458229859938413,
+      "loss": 0.4978,
+      "step": 167470
+    },
+    {
+      "epoch": 8.31826760703288,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0001345425648157346,
+      "loss": 0.5176,
+      "step": 167480
+    },
+    {
+      "epoch": 8.3187642793285,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013450283103208502,
+      "loss": 0.4729,
+      "step": 167490
+    },
+    {
+      "epoch": 8.319260951624118,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001344630972484355,
+      "loss": 0.5029,
+      "step": 167500
+    },
+    {
+      "epoch": 8.319757623919738,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013442336346478593,
+      "loss": 0.4453,
+      "step": 167510
+    },
+    {
+      "epoch": 8.320254296215357,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001343836296811364,
+      "loss": 0.5059,
+      "step": 167520
+    },
+    {
+      "epoch": 8.320750968510977,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00013434389589748685,
+      "loss": 0.4966,
+      "step": 167530
+    },
+    {
+      "epoch": 8.321247640806595,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001343041621138373,
+      "loss": 0.4385,
+      "step": 167540
+    },
+    {
+      "epoch": 8.321744313102215,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00013426442833018777,
+      "loss": 0.4618,
+      "step": 167550
+    },
+    {
+      "epoch": 8.322240985397835,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001342246945465382,
+      "loss": 0.4689,
+      "step": 167560
+    },
+    {
+      "epoch": 8.322737657693454,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00013418496076288865,
+      "loss": 0.4645,
+      "step": 167570
+    },
+    {
+      "epoch": 8.323234329989074,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001341452269792391,
+      "loss": 0.4748,
+      "step": 167580
+    },
+    {
+      "epoch": 8.323731002284692,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00013410549319558957,
+      "loss": 0.4818,
+      "step": 167590
+    },
+    {
+      "epoch": 8.324227674580312,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00013406575941194001,
+      "loss": 0.4575,
+      "step": 167600
+    },
+    {
+      "epoch": 8.32472434687593,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00013402602562829046,
+      "loss": 0.4976,
+      "step": 167610
+    },
+    {
+      "epoch": 8.32522101917155,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001339862918446409,
+      "loss": 0.4791,
+      "step": 167620
+    },
+    {
+      "epoch": 8.32571769146717,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00013394655806099137,
+      "loss": 0.4636,
+      "step": 167630
+    },
+    {
+      "epoch": 8.326214363762789,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00013390682427734182,
+      "loss": 0.4777,
+      "step": 167640
+    },
+    {
+      "epoch": 8.32671103605841,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00013386709049369226,
+      "loss": 0.4811,
+      "step": 167650
+    },
+    {
+      "epoch": 8.327207708354027,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001338273567100427,
+      "loss": 0.4824,
+      "step": 167660
+    },
+    {
+      "epoch": 8.327704380649648,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00013378762292639318,
+      "loss": 0.464,
+      "step": 167670
+    },
+    {
+      "epoch": 8.328201052945266,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00013374788914274362,
+      "loss": 0.4885,
+      "step": 167680
+    },
+    {
+      "epoch": 8.328697725240886,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00013370815535909407,
+      "loss": 0.4901,
+      "step": 167690
+    },
+    {
+      "epoch": 8.329194397536506,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00013366842157544454,
+      "loss": 0.4973,
+      "step": 167700
+    },
+    {
+      "epoch": 8.329691069832124,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00013362868779179498,
+      "loss": 0.4965,
+      "step": 167710
+    },
+    {
+      "epoch": 8.330187742127745,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00013358895400814543,
+      "loss": 0.4656,
+      "step": 167720
+    },
+    {
+      "epoch": 8.330684414423363,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00013354922022449587,
+      "loss": 0.4857,
+      "step": 167730
+    },
+    {
+      "epoch": 8.331181086718983,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00013350948644084634,
+      "loss": 0.494,
+      "step": 167740
+    },
+    {
+      "epoch": 8.331677759014601,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001334697526571968,
+      "loss": 0.4538,
+      "step": 167750
+    },
+    {
+      "epoch": 8.332174431310222,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00013343001887354726,
+      "loss": 0.48,
+      "step": 167760
+    },
+    {
+      "epoch": 8.332671103605842,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00013339028508989767,
+      "loss": 0.4763,
+      "step": 167770
+    },
+    {
+      "epoch": 8.33316777590146,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00013335055130624815,
+      "loss": 0.4868,
+      "step": 167780
+    },
+    {
+      "epoch": 8.33366444819708,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001333108175225986,
+      "loss": 0.4968,
+      "step": 167790
+    },
+    {
+      "epoch": 8.334161120492698,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00013327108373894906,
+      "loss": 0.4734,
+      "step": 167800
+    },
+    {
+      "epoch": 8.334657792788319,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00013323134995529948,
+      "loss": 0.5021,
+      "step": 167810
+    },
+    {
+      "epoch": 8.335154465083937,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00013319161617164995,
+      "loss": 0.4863,
+      "step": 167820
+    },
+    {
+      "epoch": 8.335651137379557,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001331518823880004,
+      "loss": 0.5015,
+      "step": 167830
+    },
+    {
+      "epoch": 8.336147809675177,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00013311214860435087,
+      "loss": 0.5054,
+      "step": 167840
+    },
+    {
+      "epoch": 8.336644481970795,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00013307241482070128,
+      "loss": 0.4841,
+      "step": 167850
+    },
+    {
+      "epoch": 8.337141154266416,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00013303268103705175,
+      "loss": 0.4526,
+      "step": 167860
+    },
+    {
+      "epoch": 8.337637826562034,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001329929472534022,
+      "loss": 0.508,
+      "step": 167870
+    },
+    {
+      "epoch": 8.338134498857654,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00013295321346975267,
+      "loss": 0.5082,
+      "step": 167880
+    },
+    {
+      "epoch": 8.338631171153272,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00013291347968610311,
+      "loss": 0.5039,
+      "step": 167890
+    },
+    {
+      "epoch": 8.339127843448892,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00013287374590245356,
+      "loss": 0.4792,
+      "step": 167900
+    },
+    {
+      "epoch": 8.339624515744513,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00013283401211880403,
+      "loss": 0.4819,
+      "step": 167910
+    },
+    {
+      "epoch": 8.340121188040131,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00013279427833515447,
+      "loss": 0.5087,
+      "step": 167920
+    },
+    {
+      "epoch": 8.340617860335751,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00013275454455150495,
+      "loss": 0.4872,
+      "step": 167930
+    },
+    {
+      "epoch": 8.34111453263137,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00013271481076785536,
+      "loss": 0.4954,
+      "step": 167940
+    },
+    {
+      "epoch": 8.34161120492699,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013267507698420583,
+      "loss": 0.4834,
+      "step": 167950
+    },
+    {
+      "epoch": 8.342107877222608,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00013263534320055628,
+      "loss": 0.4951,
+      "step": 167960
+    },
+    {
+      "epoch": 8.342604549518228,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00013259560941690675,
+      "loss": 0.464,
+      "step": 167970
+    },
+    {
+      "epoch": 8.343101221813848,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00013255587563325717,
+      "loss": 0.4987,
+      "step": 167980
+    },
+    {
+      "epoch": 8.343597894109466,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00013251614184960764,
+      "loss": 0.5149,
+      "step": 167990
+    },
+    {
+      "epoch": 8.344094566405087,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00013247640806595808,
+      "loss": 0.4995,
+      "step": 168000
+    },
+    {
+      "epoch": 8.344591238700705,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00013243667428230855,
+      "loss": 0.4931,
+      "step": 168010
+    },
+    {
+      "epoch": 8.345087910996325,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00013239694049865897,
+      "loss": 0.5091,
+      "step": 168020
+    },
+    {
+      "epoch": 8.345584583291943,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013235720671500944,
+      "loss": 0.4792,
+      "step": 168030
+    },
+    {
+      "epoch": 8.346081255587563,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001323174729313599,
+      "loss": 0.4816,
+      "step": 168040
+    },
+    {
+      "epoch": 8.346577927883184,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00013227773914771036,
+      "loss": 0.483,
+      "step": 168050
+    },
+    {
+      "epoch": 8.347074600178802,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001322380053640608,
+      "loss": 0.4836,
+      "step": 168060
+    },
+    {
+      "epoch": 8.347571272474422,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00013219827158041125,
+      "loss": 0.4806,
+      "step": 168070
+    },
+    {
+      "epoch": 8.34806794477004,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00013215853779676172,
+      "loss": 0.4986,
+      "step": 168080
+    },
+    {
+      "epoch": 8.34856461706566,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013211880401311216,
+      "loss": 0.49,
+      "step": 168090
+    },
+    {
+      "epoch": 8.349061289361279,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001320790702294626,
+      "loss": 0.4752,
+      "step": 168100
+    },
+    {
+      "epoch": 8.349557961656899,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00013203933644581305,
+      "loss": 0.4681,
+      "step": 168110
+    },
+    {
+      "epoch": 8.350054633952517,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00013199960266216352,
+      "loss": 0.4882,
+      "step": 168120
+    },
+    {
+      "epoch": 8.350551306248137,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00013195986887851397,
+      "loss": 0.4639,
+      "step": 168130
+    },
+    {
+      "epoch": 8.351047978543757,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001319201350948644,
+      "loss": 0.5005,
+      "step": 168140
+    },
+    {
+      "epoch": 8.351544650839376,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00013188040131121485,
+      "loss": 0.4698,
+      "step": 168150
+    },
+    {
+      "epoch": 8.352041323134996,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00013184066752756533,
+      "loss": 0.4761,
+      "step": 168160
+    },
+    {
+      "epoch": 8.352537995430614,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00013180093374391577,
+      "loss": 0.4914,
+      "step": 168170
+    },
+    {
+      "epoch": 8.353034667726234,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00013176119996026621,
+      "loss": 0.4655,
+      "step": 168180
+    },
+    {
+      "epoch": 8.353531340021853,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00013172146617661666,
+      "loss": 0.4935,
+      "step": 168190
+    },
+    {
+      "epoch": 8.354028012317473,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00013168173239296713,
+      "loss": 0.4744,
+      "step": 168200
+    },
+    {
+      "epoch": 8.354524684613093,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00013164199860931757,
+      "loss": 0.4751,
+      "step": 168210
+    },
+    {
+      "epoch": 8.355021356908711,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013160226482566802,
+      "loss": 0.4717,
+      "step": 168220
+    },
+    {
+      "epoch": 8.355518029204331,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001315625310420185,
+      "loss": 0.4954,
+      "step": 168230
+    },
+    {
+      "epoch": 8.35601470149995,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00013152279725836893,
+      "loss": 0.5185,
+      "step": 168240
+    },
+    {
+      "epoch": 8.35651137379557,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001314830634747194,
+      "loss": 0.4877,
+      "step": 168250
+    },
+    {
+      "epoch": 8.357008046091188,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00013144332969106982,
+      "loss": 0.4824,
+      "step": 168260
+    },
+    {
+      "epoch": 8.357504718386808,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001314035959074203,
+      "loss": 0.5182,
+      "step": 168270
+    },
+    {
+      "epoch": 8.358001390682428,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00013136386212377074,
+      "loss": 0.4792,
+      "step": 168280
+    },
+    {
+      "epoch": 8.358498062978047,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001313241283401212,
+      "loss": 0.479,
+      "step": 168290
+    },
+    {
+      "epoch": 8.358994735273667,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00013128439455647165,
+      "loss": 0.513,
+      "step": 168300
+    },
+    {
+      "epoch": 8.359491407569285,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001312446607728221,
+      "loss": 0.481,
+      "step": 168310
+    },
+    {
+      "epoch": 8.359988079864905,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00013120492698917254,
+      "loss": 0.462,
+      "step": 168320
+    },
+    {
+      "epoch": 8.360484752160524,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013116519320552301,
+      "loss": 0.5018,
+      "step": 168330
+    },
+    {
+      "epoch": 8.360981424456144,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00013112545942187346,
+      "loss": 0.4758,
+      "step": 168340
+    },
+    {
+      "epoch": 8.361478096751764,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001310857256382239,
+      "loss": 0.4626,
+      "step": 168350
+    },
+    {
+      "epoch": 8.361974769047382,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00013104599185457435,
+      "loss": 0.4703,
+      "step": 168360
+    },
+    {
+      "epoch": 8.362471441343002,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00013100625807092482,
+      "loss": 0.478,
+      "step": 168370
+    },
+    {
+      "epoch": 8.36296811363862,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00013096652428727526,
+      "loss": 0.4709,
+      "step": 168380
+    },
+    {
+      "epoch": 8.36346478593424,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001309267905036257,
+      "loss": 0.4842,
+      "step": 168390
+    },
+    {
+      "epoch": 8.36396145822986,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00013088705671997618,
+      "loss": 0.5029,
+      "step": 168400
+    },
+    {
+      "epoch": 8.36445813052548,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00013084732293632662,
+      "loss": 0.4755,
+      "step": 168410
+    },
+    {
+      "epoch": 8.3649548028211,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00013080758915267707,
+      "loss": 0.4617,
+      "step": 168420
+    },
+    {
+      "epoch": 8.365451475116718,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001307678553690275,
+      "loss": 0.4698,
+      "step": 168430
+    },
+    {
+      "epoch": 8.365948147412338,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00013072812158537798,
+      "loss": 0.4857,
+      "step": 168440
+    },
+    {
+      "epoch": 8.366444819707956,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00013068838780172843,
+      "loss": 0.4522,
+      "step": 168450
+    },
+    {
+      "epoch": 8.366941492003576,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001306486540180789,
+      "loss": 0.5127,
+      "step": 168460
+    },
+    {
+      "epoch": 8.367438164299195,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00013060892023442931,
+      "loss": 0.4891,
+      "step": 168470
+    },
+    {
+      "epoch": 8.367934836594815,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00013056918645077979,
+      "loss": 0.4853,
+      "step": 168480
+    },
+    {
+      "epoch": 8.368431508890435,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00013052945266713023,
+      "loss": 0.4682,
+      "step": 168490
+    },
+    {
+      "epoch": 8.368928181186053,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001304897188834807,
+      "loss": 0.4787,
+      "step": 168500
+    },
+    {
+      "epoch": 8.369424853481673,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00013044998509983112,
+      "loss": 0.4798,
+      "step": 168510
+    },
+    {
+      "epoch": 8.369921525777292,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0001304102513161816,
+      "loss": 0.4655,
+      "step": 168520
+    },
+    {
+      "epoch": 8.370418198072912,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00013037051753253203,
+      "loss": 0.4892,
+      "step": 168530
+    },
+    {
+      "epoch": 8.37091487036853,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001303307837488825,
+      "loss": 0.5185,
+      "step": 168540
+    },
+    {
+      "epoch": 8.37141154266415,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00013029104996523295,
+      "loss": 0.4781,
+      "step": 168550
+    },
+    {
+      "epoch": 8.37190821495977,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001302513161815834,
+      "loss": 0.4573,
+      "step": 168560
+    },
+    {
+      "epoch": 8.372404887255389,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00013021158239793384,
+      "loss": 0.4944,
+      "step": 168570
+    },
+    {
+      "epoch": 8.372901559551009,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0001301718486142843,
+      "loss": 0.4696,
+      "step": 168580
+    },
+    {
+      "epoch": 8.373398231846627,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00013013211483063475,
+      "loss": 0.4993,
+      "step": 168590
+    },
+    {
+      "epoch": 8.373894904142247,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001300923810469852,
+      "loss": 0.4979,
+      "step": 168600
+    },
+    {
+      "epoch": 8.374391576437866,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00013005264726333567,
+      "loss": 0.4934,
+      "step": 168610
+    },
+    {
+      "epoch": 8.374888248733486,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013001291347968611,
+      "loss": 0.4886,
+      "step": 168620
+    },
+    {
+      "epoch": 8.375384921029106,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00012997317969603656,
+      "loss": 0.4948,
+      "step": 168630
+    },
+    {
+      "epoch": 8.375881593324724,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.000129933445912387,
+      "loss": 0.4783,
+      "step": 168640
+    },
+    {
+      "epoch": 8.376378265620344,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00012989371212873747,
+      "loss": 0.4894,
+      "step": 168650
+    },
+    {
+      "epoch": 8.376874937915963,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00012985397834508792,
+      "loss": 0.4993,
+      "step": 168660
+    },
+    {
+      "epoch": 8.377371610211583,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00012981424456143836,
+      "loss": 0.4976,
+      "step": 168670
+    },
+    {
+      "epoch": 8.377868282507201,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001297745107777888,
+      "loss": 0.4744,
+      "step": 168680
+    },
+    {
+      "epoch": 8.378364954802821,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012973477699413928,
+      "loss": 0.4681,
+      "step": 168690
+    },
+    {
+      "epoch": 8.378861627098441,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00012969504321048972,
+      "loss": 0.5058,
+      "step": 168700
+    },
+    {
+      "epoch": 8.37935829939406,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001296553094268402,
+      "loss": 0.5096,
+      "step": 168710
+    },
+    {
+      "epoch": 8.37985497168968,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001296155756431906,
+      "loss": 0.4912,
+      "step": 168720
+    },
+    {
+      "epoch": 8.380351643985298,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00012957584185954108,
+      "loss": 0.5029,
+      "step": 168730
+    },
+    {
+      "epoch": 8.380848316280918,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00012953610807589153,
+      "loss": 0.5057,
+      "step": 168740
+    },
+    {
+      "epoch": 8.381344988576537,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.000129496374292242,
+      "loss": 0.4688,
+      "step": 168750
+    },
+    {
+      "epoch": 8.381841660872157,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00012945664050859244,
+      "loss": 0.4389,
+      "step": 168760
+    },
+    {
+      "epoch": 8.382338333167777,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00012941690672494289,
+      "loss": 0.4644,
+      "step": 168770
+    },
+    {
+      "epoch": 8.382835005463395,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00012937717294129336,
+      "loss": 0.5107,
+      "step": 168780
+    },
+    {
+      "epoch": 8.383331677759015,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0001293374391576438,
+      "loss": 0.4563,
+      "step": 168790
+    },
+    {
+      "epoch": 8.383828350054634,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00012929770537399425,
+      "loss": 0.5044,
+      "step": 168800
+    },
+    {
+      "epoch": 8.384325022350254,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001292579715903447,
+      "loss": 0.4981,
+      "step": 168810
+    },
+    {
+      "epoch": 8.384821694645872,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00012921823780669516,
+      "loss": 0.4397,
+      "step": 168820
+    },
+    {
+      "epoch": 8.385318366941492,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001291785040230456,
+      "loss": 0.4958,
+      "step": 168830
+    },
+    {
+      "epoch": 8.385815039237112,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00012913877023939605,
+      "loss": 0.4877,
+      "step": 168840
+    },
+    {
+      "epoch": 8.38631171153273,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001290990364557465,
+      "loss": 0.5225,
+      "step": 168850
+    },
+    {
+      "epoch": 8.38680838382835,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00012905930267209697,
+      "loss": 0.4748,
+      "step": 168860
+    },
+    {
+      "epoch": 8.387305056123969,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001290195688884474,
+      "loss": 0.4867,
+      "step": 168870
+    },
+    {
+      "epoch": 8.38780172841959,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012897983510479785,
+      "loss": 0.4634,
+      "step": 168880
+    },
+    {
+      "epoch": 8.388298400715207,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001289401013211483,
+      "loss": 0.4952,
+      "step": 168890
+    },
+    {
+      "epoch": 8.388795073010828,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00012890036753749877,
+      "loss": 0.5218,
+      "step": 168900
+    },
+    {
+      "epoch": 8.389291745306446,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012886063375384921,
+      "loss": 0.4845,
+      "step": 168910
+    },
+    {
+      "epoch": 8.389788417602066,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00012882089997019966,
+      "loss": 0.488,
+      "step": 168920
+    },
+    {
+      "epoch": 8.390285089897686,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012878116618655013,
+      "loss": 0.539,
+      "step": 168930
+    },
+    {
+      "epoch": 8.390781762193305,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00012874143240290057,
+      "loss": 0.4901,
+      "step": 168940
+    },
+    {
+      "epoch": 8.391278434488925,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012870169861925105,
+      "loss": 0.4447,
+      "step": 168950
+    },
+    {
+      "epoch": 8.391775106784543,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00012866196483560146,
+      "loss": 0.5132,
+      "step": 168960
+    },
+    {
+      "epoch": 8.392271779080163,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00012862223105195193,
+      "loss": 0.4866,
+      "step": 168970
+    },
+    {
+      "epoch": 8.392768451375781,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00012858249726830238,
+      "loss": 0.4917,
+      "step": 168980
+    },
+    {
+      "epoch": 8.393265123671402,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00012854276348465285,
+      "loss": 0.4911,
+      "step": 168990
+    },
+    {
+      "epoch": 8.393761795967022,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00012850302970100327,
+      "loss": 0.4666,
+      "step": 169000
+    },
+    {
+      "epoch": 8.39425846826264,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00012846329591735374,
+      "loss": 0.4556,
+      "step": 169010
+    },
+    {
+      "epoch": 8.39475514055826,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00012842356213370418,
+      "loss": 0.5018,
+      "step": 169020
+    },
+    {
+      "epoch": 8.395251812853878,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012838382835005465,
+      "loss": 0.4684,
+      "step": 169030
+    },
+    {
+      "epoch": 8.395748485149499,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00012834409456640507,
+      "loss": 0.4969,
+      "step": 169040
+    },
+    {
+      "epoch": 8.396245157445117,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00012830436078275554,
+      "loss": 0.4415,
+      "step": 169050
+    },
+    {
+      "epoch": 8.396741829740737,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00012826462699910599,
+      "loss": 0.486,
+      "step": 169060
+    },
+    {
+      "epoch": 8.397238502036357,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00012822489321545646,
+      "loss": 0.472,
+      "step": 169070
+    },
+    {
+      "epoch": 8.397735174331975,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001281851594318069,
+      "loss": 0.5017,
+      "step": 169080
+    },
+    {
+      "epoch": 8.398231846627596,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00012814542564815735,
+      "loss": 0.4763,
+      "step": 169090
+    },
+    {
+      "epoch": 8.398728518923214,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00012810569186450782,
+      "loss": 0.498,
+      "step": 169100
+    },
+    {
+      "epoch": 8.399225191218834,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00012806595808085826,
+      "loss": 0.5043,
+      "step": 169110
+    },
+    {
+      "epoch": 8.399721863514452,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001280262242972087,
+      "loss": 0.4787,
+      "step": 169120
+    },
+    {
+      "epoch": 8.400218535810072,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00012798649051355915,
+      "loss": 0.4647,
+      "step": 169130
+    },
+    {
+      "epoch": 8.400715208105693,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00012794675672990962,
+      "loss": 0.4939,
+      "step": 169140
+    },
+    {
+      "epoch": 8.401211880401311,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012790702294626007,
+      "loss": 0.473,
+      "step": 169150
+    },
+    {
+      "epoch": 8.401708552696931,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00012786728916261054,
+      "loss": 0.4581,
+      "step": 169160
+    },
+    {
+      "epoch": 8.40220522499255,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00012782755537896095,
+      "loss": 0.4621,
+      "step": 169170
+    },
+    {
+      "epoch": 8.40270189728817,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00012778782159531143,
+      "loss": 0.4803,
+      "step": 169180
+    },
+    {
+      "epoch": 8.403198569583788,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012774808781166187,
+      "loss": 0.4912,
+      "step": 169190
+    },
+    {
+      "epoch": 8.403695241879408,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012770835402801234,
+      "loss": 0.4673,
+      "step": 169200
+    },
+    {
+      "epoch": 8.404191914175028,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00012766862024436276,
+      "loss": 0.5012,
+      "step": 169210
+    },
+    {
+      "epoch": 8.404688586470646,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00012762888646071323,
+      "loss": 0.4878,
+      "step": 169220
+    },
+    {
+      "epoch": 8.405185258766267,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00012758915267706367,
+      "loss": 0.5079,
+      "step": 169230
+    },
+    {
+      "epoch": 8.405681931061885,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00012754941889341415,
+      "loss": 0.4728,
+      "step": 169240
+    },
+    {
+      "epoch": 8.406178603357505,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001275096851097646,
+      "loss": 0.4736,
+      "step": 169250
+    },
+    {
+      "epoch": 8.406675275653123,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00012746995132611503,
+      "loss": 0.4654,
+      "step": 169260
+    },
+    {
+      "epoch": 8.407171947948743,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00012743021754246548,
+      "loss": 0.4865,
+      "step": 169270
+    },
+    {
+      "epoch": 8.407668620244364,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00012739048375881595,
+      "loss": 0.4786,
+      "step": 169280
+    },
+    {
+      "epoch": 8.408165292539982,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001273507499751664,
+      "loss": 0.4763,
+      "step": 169290
+    },
+    {
+      "epoch": 8.408661964835602,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00012731101619151684,
+      "loss": 0.4851,
+      "step": 169300
+    },
+    {
+      "epoch": 8.40915863713122,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001272712824078673,
+      "loss": 0.4838,
+      "step": 169310
+    },
+    {
+      "epoch": 8.40965530942684,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00012723154862421775,
+      "loss": 0.4796,
+      "step": 169320
+    },
+    {
+      "epoch": 8.410151981722459,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001271918148405682,
+      "loss": 0.5066,
+      "step": 169330
+    },
+    {
+      "epoch": 8.410648654018079,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00012715208105691864,
+      "loss": 0.5034,
+      "step": 169340
+    },
+    {
+      "epoch": 8.411145326313699,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001271123472732691,
+      "loss": 0.4853,
+      "step": 169350
+    },
+    {
+      "epoch": 8.411641998609317,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00012707261348961956,
+      "loss": 0.4663,
+      "step": 169360
+    },
+    {
+      "epoch": 8.412138670904937,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00012703287970597,
+      "loss": 0.4583,
+      "step": 169370
+    },
+    {
+      "epoch": 8.412635343200556,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00012699314592232045,
+      "loss": 0.5136,
+      "step": 169380
+    },
+    {
+      "epoch": 8.413132015496176,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00012695341213867092,
+      "loss": 0.5158,
+      "step": 169390
+    },
+    {
+      "epoch": 8.413628687791794,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00012691367835502136,
+      "loss": 0.4755,
+      "step": 169400
+    },
+    {
+      "epoch": 8.414125360087414,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001268739445713718,
+      "loss": 0.4769,
+      "step": 169410
+    },
+    {
+      "epoch": 8.414622032383035,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00012683421078772225,
+      "loss": 0.4683,
+      "step": 169420
+    },
+    {
+      "epoch": 8.415118704678653,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00012679447700407272,
+      "loss": 0.4729,
+      "step": 169430
+    },
+    {
+      "epoch": 8.415615376974273,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00012675474322042317,
+      "loss": 0.4856,
+      "step": 169440
+    },
+    {
+      "epoch": 8.416112049269891,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0001267150094367736,
+      "loss": 0.5044,
+      "step": 169450
+    },
+    {
+      "epoch": 8.416608721565511,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00012667527565312408,
+      "loss": 0.4702,
+      "step": 169460
+    },
+    {
+      "epoch": 8.41710539386113,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00012663554186947453,
+      "loss": 0.4685,
+      "step": 169470
+    },
+    {
+      "epoch": 8.41760206615675,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.000126595808085825,
+      "loss": 0.4905,
+      "step": 169480
+    },
+    {
+      "epoch": 8.418098738452368,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00012655607430217541,
+      "loss": 0.4802,
+      "step": 169490
+    },
+    {
+      "epoch": 8.418595410747988,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012651634051852589,
+      "loss": 0.4804,
+      "step": 169500
+    },
+    {
+      "epoch": 8.419092083043608,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00012647660673487633,
+      "loss": 0.4988,
+      "step": 169510
+    },
+    {
+      "epoch": 8.419588755339227,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001264368729512268,
+      "loss": 0.5153,
+      "step": 169520
+    },
+    {
+      "epoch": 8.420085427634847,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00012639713916757722,
+      "loss": 0.4659,
+      "step": 169530
+    },
+    {
+      "epoch": 8.420582099930465,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001263574053839277,
+      "loss": 0.4871,
+      "step": 169540
+    },
+    {
+      "epoch": 8.421078772226085,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00012631767160027813,
+      "loss": 0.4974,
+      "step": 169550
+    },
+    {
+      "epoch": 8.421575444521704,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001262779378166286,
+      "loss": 0.4921,
+      "step": 169560
+    },
+    {
+      "epoch": 8.422072116817324,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00012623820403297905,
+      "loss": 0.5091,
+      "step": 169570
+    },
+    {
+      "epoch": 8.422568789112944,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001261984702493295,
+      "loss": 0.472,
+      "step": 169580
+    },
+    {
+      "epoch": 8.423065461408562,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00012615873646567994,
+      "loss": 0.4821,
+      "step": 169590
+    },
+    {
+      "epoch": 8.423562133704182,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0001261190026820304,
+      "loss": 0.5078,
+      "step": 169600
+    },
+    {
+      "epoch": 8.4240588059998,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00012607926889838085,
+      "loss": 0.4889,
+      "step": 169610
+    },
+    {
+      "epoch": 8.42455547829542,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001260395351147313,
+      "loss": 0.4557,
+      "step": 169620
+    },
+    {
+      "epoch": 8.42505215059104,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00012599980133108177,
+      "loss": 0.4891,
+      "step": 169630
+    },
+    {
+      "epoch": 8.42554882288666,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001259600675474322,
+      "loss": 0.4704,
+      "step": 169640
+    },
+    {
+      "epoch": 8.42604549518228,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00012592033376378268,
+      "loss": 0.4891,
+      "step": 169650
+    },
+    {
+      "epoch": 8.426542167477898,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001258805999801331,
+      "loss": 0.5038,
+      "step": 169660
+    },
+    {
+      "epoch": 8.427038839773518,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00012584086619648357,
+      "loss": 0.5051,
+      "step": 169670
+    },
+    {
+      "epoch": 8.427535512069136,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00012580113241283402,
+      "loss": 0.4965,
+      "step": 169680
+    },
+    {
+      "epoch": 8.428032184364756,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0001257613986291845,
+      "loss": 0.4932,
+      "step": 169690
+    },
+    {
+      "epoch": 8.428528856660375,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001257216648455349,
+      "loss": 0.5121,
+      "step": 169700
+    },
+    {
+      "epoch": 8.429025528955995,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012568193106188538,
+      "loss": 0.4726,
+      "step": 169710
+    },
+    {
+      "epoch": 8.429522201251615,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00012564219727823582,
+      "loss": 0.4629,
+      "step": 169720
+    },
+    {
+      "epoch": 8.430018873547233,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001256024634945863,
+      "loss": 0.4679,
+      "step": 169730
+    },
+    {
+      "epoch": 8.430515545842853,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0001255627297109367,
+      "loss": 0.4935,
+      "step": 169740
+    },
+    {
+      "epoch": 8.431012218138472,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00012552299592728718,
+      "loss": 0.4959,
+      "step": 169750
+    },
+    {
+      "epoch": 8.431508890434092,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00012548326214363763,
+      "loss": 0.4851,
+      "step": 169760
+    },
+    {
+      "epoch": 8.43200556272971,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001254435283599881,
+      "loss": 0.4833,
+      "step": 169770
+    },
+    {
+      "epoch": 8.43250223502533,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00012540379457633854,
+      "loss": 0.4903,
+      "step": 169780
+    },
+    {
+      "epoch": 8.43299890732095,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00012536406079268899,
+      "loss": 0.4767,
+      "step": 169790
+    },
+    {
+      "epoch": 8.433495579616569,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00012532432700903946,
+      "loss": 0.4849,
+      "step": 169800
+    },
+    {
+      "epoch": 8.433992251912189,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001252845932253899,
+      "loss": 0.5229,
+      "step": 169810
+    },
+    {
+      "epoch": 8.434488924207807,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00012524485944174035,
+      "loss": 0.4598,
+      "step": 169820
+    },
+    {
+      "epoch": 8.434985596503427,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001252051256580908,
+      "loss": 0.5054,
+      "step": 169830
+    },
+    {
+      "epoch": 8.435482268799046,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00012516539187444126,
+      "loss": 0.459,
+      "step": 169840
+    },
+    {
+      "epoch": 8.435978941094666,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001251256580907917,
+      "loss": 0.491,
+      "step": 169850
+    },
+    {
+      "epoch": 8.436475613390286,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00012508592430714215,
+      "loss": 0.4843,
+      "step": 169860
+    },
+    {
+      "epoch": 8.436972285685904,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001250461905234926,
+      "loss": 0.5034,
+      "step": 169870
+    },
+    {
+      "epoch": 8.437468957981524,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00012500645673984307,
+      "loss": 0.4848,
+      "step": 169880
+    },
+    {
+      "epoch": 8.437965630277143,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0001249667229561935,
+      "loss": 0.5074,
+      "step": 169890
+    },
+    {
+      "epoch": 8.438462302572763,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00012492698917254395,
+      "loss": 0.5188,
+      "step": 169900
+    },
+    {
+      "epoch": 8.438958974868381,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0001248872553888944,
+      "loss": 0.5135,
+      "step": 169910
+    },
+    {
+      "epoch": 8.439455647164001,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00012484752160524487,
+      "loss": 0.457,
+      "step": 169920
+    },
+    {
+      "epoch": 8.439952319459621,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00012480778782159531,
+      "loss": 0.5279,
+      "step": 169930
+    },
+    {
+      "epoch": 8.44044899175524,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00012476805403794576,
+      "loss": 0.4814,
+      "step": 169940
+    },
+    {
+      "epoch": 8.44094566405086,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00012472832025429623,
+      "loss": 0.485,
+      "step": 169950
+    },
+    {
+      "epoch": 8.441442336346478,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00012468858647064667,
+      "loss": 0.5091,
+      "step": 169960
+    },
+    {
+      "epoch": 8.441939008642098,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00012464885268699712,
+      "loss": 0.4557,
+      "step": 169970
+    },
+    {
+      "epoch": 8.442435680937717,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001246091189033476,
+      "loss": 0.4626,
+      "step": 169980
+    },
+    {
+      "epoch": 8.442932353233337,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012456938511969803,
+      "loss": 0.4495,
+      "step": 169990
+    },
+    {
+      "epoch": 8.443429025528957,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00012452965133604848,
+      "loss": 0.4806,
+      "step": 170000
+    },
+    {
+      "epoch": 8.443925697824575,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00012448991755239895,
+      "loss": 0.4817,
+      "step": 170010
+    },
+    {
+      "epoch": 8.444422370120195,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001244501837687494,
+      "loss": 0.4995,
+      "step": 170020
+    },
+    {
+      "epoch": 8.444919042415814,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00012441044998509984,
+      "loss": 0.4927,
+      "step": 170030
+    },
+    {
+      "epoch": 8.445415714711434,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012437071620145028,
+      "loss": 0.5012,
+      "step": 170040
+    },
+    {
+      "epoch": 8.445912387007052,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012433098241780075,
+      "loss": 0.493,
+      "step": 170050
+    },
+    {
+      "epoch": 8.446409059302672,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001242912486341512,
+      "loss": 0.525,
+      "step": 170060
+    },
+    {
+      "epoch": 8.446905731598292,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00012425151485050164,
+      "loss": 0.4894,
+      "step": 170070
+    },
+    {
+      "epoch": 8.44740240389391,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00012421178106685209,
+      "loss": 0.4814,
+      "step": 170080
+    },
+    {
+      "epoch": 8.44789907618953,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00012417204728320256,
+      "loss": 0.4779,
+      "step": 170090
+    },
+    {
+      "epoch": 8.448395748485149,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.000124132313499553,
+      "loss": 0.4791,
+      "step": 170100
+    },
+    {
+      "epoch": 8.44889242078077,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00012409257971590345,
+      "loss": 0.4647,
+      "step": 170110
+    },
+    {
+      "epoch": 8.449389093076388,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001240528459322539,
+      "loss": 0.5275,
+      "step": 170120
+    },
+    {
+      "epoch": 8.449885765372008,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00012401311214860436,
+      "loss": 0.4688,
+      "step": 170130
+    },
+    {
+      "epoch": 8.450382437667628,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001239733783649548,
+      "loss": 0.458,
+      "step": 170140
+    },
+    {
+      "epoch": 8.450879109963246,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.00012393364458130525,
+      "loss": 0.4909,
+      "step": 170150
+    },
+    {
+      "epoch": 8.451375782258866,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00012389391079765572,
+      "loss": 0.5083,
+      "step": 170160
+    },
+    {
+      "epoch": 8.451872454554485,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012385417701400617,
+      "loss": 0.5134,
+      "step": 170170
+    },
+    {
+      "epoch": 8.452369126850105,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00012381444323035664,
+      "loss": 0.4914,
+      "step": 170180
+    },
+    {
+      "epoch": 8.452865799145723,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00012377470944670705,
+      "loss": 0.4646,
+      "step": 170190
+    },
+    {
+      "epoch": 8.453362471441343,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00012373497566305753,
+      "loss": 0.4769,
+      "step": 170200
+    },
+    {
+      "epoch": 8.453859143736963,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00012369524187940797,
+      "loss": 0.4946,
+      "step": 170210
+    },
+    {
+      "epoch": 8.454355816032582,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00012365550809575844,
+      "loss": 0.503,
+      "step": 170220
+    },
+    {
+      "epoch": 8.454852488328202,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00012361577431210886,
+      "loss": 0.5124,
+      "step": 170230
+    },
+    {
+      "epoch": 8.45534916062382,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00012357604052845933,
+      "loss": 0.4763,
+      "step": 170240
+    },
+    {
+      "epoch": 8.45584583291944,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012353630674480977,
+      "loss": 0.4672,
+      "step": 170250
+    },
+    {
+      "epoch": 8.456342505215058,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00012349657296116024,
+      "loss": 0.5051,
+      "step": 170260
+    },
+    {
+      "epoch": 8.456839177510679,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00012345683917751066,
+      "loss": 0.4574,
+      "step": 170270
+    },
+    {
+      "epoch": 8.457335849806299,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00012341710539386113,
+      "loss": 0.4711,
+      "step": 170280
+    },
+    {
+      "epoch": 8.457832522101917,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00012337737161021158,
+      "loss": 0.4805,
+      "step": 170290
+    },
+    {
+      "epoch": 8.458329194397537,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00012333763782656205,
+      "loss": 0.4692,
+      "step": 170300
+    },
+    {
+      "epoch": 8.458825866693156,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001232979040429125,
+      "loss": 0.4767,
+      "step": 170310
+    },
+    {
+      "epoch": 8.459322538988776,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012325817025926294,
+      "loss": 0.4833,
+      "step": 170320
+    },
+    {
+      "epoch": 8.459819211284394,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001232184364756134,
+      "loss": 0.488,
+      "step": 170330
+    },
+    {
+      "epoch": 8.460315883580014,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00012317870269196385,
+      "loss": 0.529,
+      "step": 170340
+    },
+    {
+      "epoch": 8.460812555875632,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001231389689083143,
+      "loss": 0.4887,
+      "step": 170350
+    },
+    {
+      "epoch": 8.461309228171253,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00012309923512466474,
+      "loss": 0.5137,
+      "step": 170360
+    },
+    {
+      "epoch": 8.461805900466873,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001230595013410152,
+      "loss": 0.499,
+      "step": 170370
+    },
+    {
+      "epoch": 8.462302572762491,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00012301976755736566,
+      "loss": 0.5015,
+      "step": 170380
+    },
+    {
+      "epoch": 8.462799245058111,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00012298003377371613,
+      "loss": 0.4844,
+      "step": 170390
+    },
+    {
+      "epoch": 8.46329591735373,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00012294029999006655,
+      "loss": 0.4932,
+      "step": 170400
+    },
+    {
+      "epoch": 8.46379258964935,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00012290056620641702,
+      "loss": 0.4661,
+      "step": 170410
+    },
+    {
+      "epoch": 8.464289261944968,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00012286083242276746,
+      "loss": 0.504,
+      "step": 170420
+    },
+    {
+      "epoch": 8.464785934240588,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00012282109863911793,
+      "loss": 0.4773,
+      "step": 170430
+    },
+    {
+      "epoch": 8.465282606536208,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00012278136485546835,
+      "loss": 0.4944,
+      "step": 170440
+    },
+    {
+      "epoch": 8.465779278831826,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00012274163107181882,
+      "loss": 0.4932,
+      "step": 170450
+    },
+    {
+      "epoch": 8.466275951127447,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012270189728816927,
+      "loss": 0.4765,
+      "step": 170460
+    },
+    {
+      "epoch": 8.466772623423065,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012266216350451974,
+      "loss": 0.487,
+      "step": 170470
+    },
+    {
+      "epoch": 8.467269295718685,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 0.00012262242972087018,
+      "loss": 0.4692,
+      "step": 170480
+    },
+    {
+      "epoch": 8.467765968014303,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00012258269593722063,
+      "loss": 0.4835,
+      "step": 170490
+    },
+    {
+      "epoch": 8.468262640309923,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001225429621535711,
+      "loss": 0.4907,
+      "step": 170500
+    },
+    {
+      "epoch": 8.468759312605544,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00012250322836992154,
+      "loss": 0.4705,
+      "step": 170510
+    },
+    {
+      "epoch": 8.469255984901162,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00012246349458627199,
+      "loss": 0.4721,
+      "step": 170520
+    },
+    {
+      "epoch": 8.469752657196782,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00012242376080262243,
+      "loss": 0.4743,
+      "step": 170530
+    },
+    {
+      "epoch": 8.4702493294924,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001223840270189729,
+      "loss": 0.4876,
+      "step": 170540
+    },
+    {
+      "epoch": 8.47074600178802,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00012234429323532335,
+      "loss": 0.5025,
+      "step": 170550
+    },
+    {
+      "epoch": 8.471242674083639,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001223045594516738,
+      "loss": 0.4845,
+      "step": 170560
+    },
+    {
+      "epoch": 8.471739346379259,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00012226482566802423,
+      "loss": 0.5168,
+      "step": 170570
+    },
+    {
+      "epoch": 8.472236018674879,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001222250918843747,
+      "loss": 0.498,
+      "step": 170580
+    },
+    {
+      "epoch": 8.472732690970497,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012218535810072515,
+      "loss": 0.4583,
+      "step": 170590
+    },
+    {
+      "epoch": 8.473229363266118,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001221456243170756,
+      "loss": 0.4809,
+      "step": 170600
+    },
+    {
+      "epoch": 8.473726035561736,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00012210589053342604,
+      "loss": 0.5027,
+      "step": 170610
+    },
+    {
+      "epoch": 8.474222707857356,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00012206615674977651,
+      "loss": 0.459,
+      "step": 170620
+    },
+    {
+      "epoch": 8.474719380152974,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00012202642296612697,
+      "loss": 0.4806,
+      "step": 170630
+    },
+    {
+      "epoch": 8.475216052448594,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001219866891824774,
+      "loss": 0.4887,
+      "step": 170640
+    },
+    {
+      "epoch": 8.475712724744215,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00012194695539882786,
+      "loss": 0.4781,
+      "step": 170650
+    },
+    {
+      "epoch": 8.476209397039833,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00012190722161517831,
+      "loss": 0.4878,
+      "step": 170660
+    },
+    {
+      "epoch": 8.476706069335453,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00012186748783152877,
+      "loss": 0.5021,
+      "step": 170670
+    },
+    {
+      "epoch": 8.477202741631071,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001218277540478792,
+      "loss": 0.4751,
+      "step": 170680
+    },
+    {
+      "epoch": 8.477699413926691,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00012178802026422966,
+      "loss": 0.4795,
+      "step": 170690
+    },
+    {
+      "epoch": 8.47819608622231,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00012174828648058012,
+      "loss": 0.5012,
+      "step": 170700
+    },
+    {
+      "epoch": 8.47869275851793,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00012170855269693058,
+      "loss": 0.5053,
+      "step": 170710
+    },
+    {
+      "epoch": 8.47918943081355,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00012166881891328102,
+      "loss": 0.4818,
+      "step": 170720
+    },
+    {
+      "epoch": 8.479686103109168,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00012162908512963148,
+      "loss": 0.4805,
+      "step": 170730
+    },
+    {
+      "epoch": 8.480182775404788,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00012158935134598192,
+      "loss": 0.5127,
+      "step": 170740
+    },
+    {
+      "epoch": 8.480679447700407,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00012154961756233238,
+      "loss": 0.5017,
+      "step": 170750
+    },
+    {
+      "epoch": 8.481176119996027,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00012150988377868282,
+      "loss": 0.4945,
+      "step": 170760
+    },
+    {
+      "epoch": 8.481672792291645,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00012147014999503328,
+      "loss": 0.4775,
+      "step": 170770
+    },
+    {
+      "epoch": 8.482169464587265,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00012143041621138374,
+      "loss": 0.4812,
+      "step": 170780
+    },
+    {
+      "epoch": 8.482666136882886,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001213906824277342,
+      "loss": 0.4761,
+      "step": 170790
+    },
+    {
+      "epoch": 8.483162809178504,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00012135094864408463,
+      "loss": 0.4856,
+      "step": 170800
+    },
+    {
+      "epoch": 8.483659481474124,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00012131121486043509,
+      "loss": 0.4723,
+      "step": 170810
+    },
+    {
+      "epoch": 8.484156153769742,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00012127148107678554,
+      "loss": 0.5,
+      "step": 170820
+    },
+    {
+      "epoch": 8.484652826065362,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.000121231747293136,
+      "loss": 0.4597,
+      "step": 170830
+    },
+    {
+      "epoch": 8.48514949836098,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00012119201350948646,
+      "loss": 0.4632,
+      "step": 170840
+    },
+    {
+      "epoch": 8.4856461706566,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00012115227972583689,
+      "loss": 0.4383,
+      "step": 170850
+    },
+    {
+      "epoch": 8.48614284295222,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00012111254594218735,
+      "loss": 0.4637,
+      "step": 170860
+    },
+    {
+      "epoch": 8.48663951524784,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001210728121585378,
+      "loss": 0.485,
+      "step": 170870
+    },
+    {
+      "epoch": 8.48713618754346,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00012103307837488826,
+      "loss": 0.4719,
+      "step": 170880
+    },
+    {
+      "epoch": 8.487632859839078,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001209933445912387,
+      "loss": 0.4889,
+      "step": 170890
+    },
+    {
+      "epoch": 8.488129532134698,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00012095361080758915,
+      "loss": 0.5083,
+      "step": 170900
+    },
+    {
+      "epoch": 8.488626204430316,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00012091387702393961,
+      "loss": 0.4648,
+      "step": 170910
+    },
+    {
+      "epoch": 8.489122876725936,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00012087414324029007,
+      "loss": 0.462,
+      "step": 170920
+    },
+    {
+      "epoch": 8.489619549021555,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00012083440945664051,
+      "loss": 0.4874,
+      "step": 170930
+    },
+    {
+      "epoch": 8.490116221317175,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00012079467567299097,
+      "loss": 0.4802,
+      "step": 170940
+    },
+    {
+      "epoch": 8.490612893612795,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00012075494188934143,
+      "loss": 0.5076,
+      "step": 170950
+    },
+    {
+      "epoch": 8.491109565908413,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00012071520810569188,
+      "loss": 0.4954,
+      "step": 170960
+    },
+    {
+      "epoch": 8.491606238204033,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00012067547432204232,
+      "loss": 0.497,
+      "step": 170970
+    },
+    {
+      "epoch": 8.492102910499652,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012063574053839277,
+      "loss": 0.5424,
+      "step": 170980
+    },
+    {
+      "epoch": 8.492599582795272,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00012059600675474323,
+      "loss": 0.5002,
+      "step": 170990
+    },
+    {
+      "epoch": 8.49309625509089,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00012055627297109369,
+      "loss": 0.4813,
+      "step": 171000
+    },
+    {
+      "epoch": 8.49359292738651,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00012051653918744412,
+      "loss": 0.5072,
+      "step": 171010
+    },
+    {
+      "epoch": 8.49408959968213,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00012047680540379458,
+      "loss": 0.5017,
+      "step": 171020
+    },
+    {
+      "epoch": 8.494586271977749,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012043707162014503,
+      "loss": 0.4959,
+      "step": 171030
+    },
+    {
+      "epoch": 8.495082944273369,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00012039733783649549,
+      "loss": 0.4379,
+      "step": 171040
+    },
+    {
+      "epoch": 8.495579616568987,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00012035760405284592,
+      "loss": 0.4611,
+      "step": 171050
+    },
+    {
+      "epoch": 8.496076288864607,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00012031787026919638,
+      "loss": 0.5006,
+      "step": 171060
+    },
+    {
+      "epoch": 8.496572961160226,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00012027813648554684,
+      "loss": 0.4515,
+      "step": 171070
+    },
+    {
+      "epoch": 8.497069633455846,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001202384027018973,
+      "loss": 0.487,
+      "step": 171080
+    },
+    {
+      "epoch": 8.497566305751466,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00012019866891824774,
+      "loss": 0.484,
+      "step": 171090
+    },
+    {
+      "epoch": 8.498062978047084,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0001201589351345982,
+      "loss": 0.4772,
+      "step": 171100
+    },
+    {
+      "epoch": 8.498559650342704,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00012011920135094866,
+      "loss": 0.4699,
+      "step": 171110
+    },
+    {
+      "epoch": 8.499056322638323,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012007946756729911,
+      "loss": 0.4888,
+      "step": 171120
+    },
+    {
+      "epoch": 8.499552994933943,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00012003973378364955,
+      "loss": 0.461,
+      "step": 171130
+    },
+    {
+      "epoch": 8.500049667229561,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00012,
+      "loss": 0.5063,
+      "step": 171140
+    },
+    {
+      "epoch": 8.500546339525181,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00011996026621635046,
+      "loss": 0.4832,
+      "step": 171150
+    },
+    {
+      "epoch": 8.501043011820801,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011992053243270092,
+      "loss": 0.4892,
+      "step": 171160
+    },
+    {
+      "epoch": 8.50153968411642,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00011988079864905135,
+      "loss": 0.4947,
+      "step": 171170
+    },
+    {
+      "epoch": 8.50203635641204,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00011984106486540181,
+      "loss": 0.4856,
+      "step": 171180
+    },
+    {
+      "epoch": 8.502533028707658,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00011980133108175226,
+      "loss": 0.4755,
+      "step": 171190
+    },
+    {
+      "epoch": 8.503029701003278,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00011976159729810272,
+      "loss": 0.5107,
+      "step": 171200
+    },
+    {
+      "epoch": 8.503526373298897,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00011972186351445315,
+      "loss": 0.5227,
+      "step": 171210
+    },
+    {
+      "epoch": 8.504023045594517,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00011968212973080361,
+      "loss": 0.4864,
+      "step": 171220
+    },
+    {
+      "epoch": 8.504519717890137,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00011964239594715407,
+      "loss": 0.4398,
+      "step": 171230
+    },
+    {
+      "epoch": 8.505016390185755,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00011960266216350453,
+      "loss": 0.4832,
+      "step": 171240
+    },
+    {
+      "epoch": 8.505513062481375,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00011956292837985498,
+      "loss": 0.4942,
+      "step": 171250
+    },
+    {
+      "epoch": 8.506009734776994,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00011952319459620543,
+      "loss": 0.4764,
+      "step": 171260
+    },
+    {
+      "epoch": 8.506506407072614,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00011948346081255589,
+      "loss": 0.4805,
+      "step": 171270
+    },
+    {
+      "epoch": 8.507003079368232,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00011944372702890634,
+      "loss": 0.5,
+      "step": 171280
+    },
+    {
+      "epoch": 8.507499751663852,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00011940399324525679,
+      "loss": 0.4862,
+      "step": 171290
+    },
+    {
+      "epoch": 8.507996423959472,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00011936425946160723,
+      "loss": 0.4842,
+      "step": 171300
+    },
+    {
+      "epoch": 8.50849309625509,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00011932452567795769,
+      "loss": 0.4655,
+      "step": 171310
+    },
+    {
+      "epoch": 8.50898976855071,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00011928479189430815,
+      "loss": 0.4868,
+      "step": 171320
+    },
+    {
+      "epoch": 8.509486440846329,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001192450581106586,
+      "loss": 0.4836,
+      "step": 171330
+    },
+    {
+      "epoch": 8.50998311314195,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00011920532432700904,
+      "loss": 0.4904,
+      "step": 171340
+    },
+    {
+      "epoch": 8.510479785437568,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001191655905433595,
+      "loss": 0.4697,
+      "step": 171350
+    },
+    {
+      "epoch": 8.510976457733188,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00011912585675970995,
+      "loss": 0.5178,
+      "step": 171360
+    },
+    {
+      "epoch": 8.511473130028808,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00011908612297606041,
+      "loss": 0.4777,
+      "step": 171370
+    },
+    {
+      "epoch": 8.511969802324426,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00011904638919241084,
+      "loss": 0.5055,
+      "step": 171380
+    },
+    {
+      "epoch": 8.512466474620046,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001190066554087613,
+      "loss": 0.4759,
+      "step": 171390
+    },
+    {
+      "epoch": 8.512963146915665,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00011896692162511176,
+      "loss": 0.4742,
+      "step": 171400
+    },
+    {
+      "epoch": 8.513459819211285,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00011892718784146221,
+      "loss": 0.4928,
+      "step": 171410
+    },
+    {
+      "epoch": 8.513956491506903,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00011888745405781266,
+      "loss": 0.4833,
+      "step": 171420
+    },
+    {
+      "epoch": 8.514453163802523,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00011884772027416312,
+      "loss": 0.514,
+      "step": 171430
+    },
+    {
+      "epoch": 8.514949836098143,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00011880798649051356,
+      "loss": 0.4703,
+      "step": 171440
+    },
+    {
+      "epoch": 8.515446508393762,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00011876825270686402,
+      "loss": 0.4729,
+      "step": 171450
+    },
+    {
+      "epoch": 8.515943180689382,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011872851892321446,
+      "loss": 0.4881,
+      "step": 171460
+    },
+    {
+      "epoch": 8.516439852985,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00011868878513956492,
+      "loss": 0.4682,
+      "step": 171470
+    },
+    {
+      "epoch": 8.51693652528062,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00011864905135591538,
+      "loss": 0.4637,
+      "step": 171480
+    },
+    {
+      "epoch": 8.517433197576239,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00011860931757226584,
+      "loss": 0.4956,
+      "step": 171490
+    },
+    {
+      "epoch": 8.517929869871859,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00011856958378861627,
+      "loss": 0.4511,
+      "step": 171500
+    },
+    {
+      "epoch": 8.518426542167479,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00011852985000496672,
+      "loss": 0.4889,
+      "step": 171510
+    },
+    {
+      "epoch": 8.518923214463097,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00011849011622131718,
+      "loss": 0.45,
+      "step": 171520
+    },
+    {
+      "epoch": 8.519419886758717,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00011845038243766764,
+      "loss": 0.4617,
+      "step": 171530
+    },
+    {
+      "epoch": 8.519916559054336,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00011841064865401807,
+      "loss": 0.5195,
+      "step": 171540
+    },
+    {
+      "epoch": 8.520413231349956,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00011837091487036853,
+      "loss": 0.4709,
+      "step": 171550
+    },
+    {
+      "epoch": 8.520909903645574,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011833118108671899,
+      "loss": 0.4762,
+      "step": 171560
+    },
+    {
+      "epoch": 8.521406575941194,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011829144730306944,
+      "loss": 0.4603,
+      "step": 171570
+    },
+    {
+      "epoch": 8.521903248236814,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00011825171351941989,
+      "loss": 0.4745,
+      "step": 171580
+    },
+    {
+      "epoch": 8.522399920532433,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00011821197973577033,
+      "loss": 0.501,
+      "step": 171590
+    },
+    {
+      "epoch": 8.522896592828053,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00011817224595212079,
+      "loss": 0.4991,
+      "step": 171600
+    },
+    {
+      "epoch": 8.523393265123671,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00011813251216847125,
+      "loss": 0.4872,
+      "step": 171610
+    },
+    {
+      "epoch": 8.523889937419291,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00011809277838482169,
+      "loss": 0.4998,
+      "step": 171620
+    },
+    {
+      "epoch": 8.52438660971491,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00011805304460117215,
+      "loss": 0.4902,
+      "step": 171630
+    },
+    {
+      "epoch": 8.52488328201053,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00011801331081752261,
+      "loss": 0.4906,
+      "step": 171640
+    },
+    {
+      "epoch": 8.52537995430615,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00011797357703387307,
+      "loss": 0.4801,
+      "step": 171650
+    },
+    {
+      "epoch": 8.525876626601768,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011793384325022352,
+      "loss": 0.4798,
+      "step": 171660
+    },
+    {
+      "epoch": 8.526373298897388,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00011789410946657395,
+      "loss": 0.4599,
+      "step": 171670
+    },
+    {
+      "epoch": 8.526869971193006,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00011785437568292441,
+      "loss": 0.4866,
+      "step": 171680
+    },
+    {
+      "epoch": 8.527366643488627,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00011781464189927487,
+      "loss": 0.5076,
+      "step": 171690
+    },
+    {
+      "epoch": 8.527863315784245,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00011777490811562533,
+      "loss": 0.4867,
+      "step": 171700
+    },
+    {
+      "epoch": 8.528359988079865,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011773517433197576,
+      "loss": 0.4963,
+      "step": 171710
+    },
+    {
+      "epoch": 8.528856660375485,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00011769544054832622,
+      "loss": 0.5126,
+      "step": 171720
+    },
+    {
+      "epoch": 8.529353332671104,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00011765570676467667,
+      "loss": 0.4964,
+      "step": 171730
+    },
+    {
+      "epoch": 8.529850004966724,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00011761597298102713,
+      "loss": 0.4977,
+      "step": 171740
+    },
+    {
+      "epoch": 8.530346677262342,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00011757623919737756,
+      "loss": 0.5222,
+      "step": 171750
+    },
+    {
+      "epoch": 8.530843349557962,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00011753650541372802,
+      "loss": 0.4701,
+      "step": 171760
+    },
+    {
+      "epoch": 8.53134002185358,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00011749677163007848,
+      "loss": 0.4865,
+      "step": 171770
+    },
+    {
+      "epoch": 8.5318366941492,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00011745703784642894,
+      "loss": 0.4622,
+      "step": 171780
+    },
+    {
+      "epoch": 8.532333366444819,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00011741730406277938,
+      "loss": 0.4803,
+      "step": 171790
+    },
+    {
+      "epoch": 8.532830038740439,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011737757027912984,
+      "loss": 0.4632,
+      "step": 171800
+    },
+    {
+      "epoch": 8.533326711036059,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001173378364954803,
+      "loss": 0.4871,
+      "step": 171810
+    },
+    {
+      "epoch": 8.533823383331677,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00011729810271183075,
+      "loss": 0.5087,
+      "step": 171820
+    },
+    {
+      "epoch": 8.534320055627298,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00011725836892818118,
+      "loss": 0.4876,
+      "step": 171830
+    },
+    {
+      "epoch": 8.534816727922916,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00011721863514453164,
+      "loss": 0.4843,
+      "step": 171840
+    },
+    {
+      "epoch": 8.535313400218536,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001171789013608821,
+      "loss": 0.4838,
+      "step": 171850
+    },
+    {
+      "epoch": 8.535810072514154,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00011713916757723256,
+      "loss": 0.463,
+      "step": 171860
+    },
+    {
+      "epoch": 8.536306744809774,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00011709943379358299,
+      "loss": 0.4703,
+      "step": 171870
+    },
+    {
+      "epoch": 8.536803417105395,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011705970000993345,
+      "loss": 0.4724,
+      "step": 171880
+    },
+    {
+      "epoch": 8.537300089401013,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0001170199662262839,
+      "loss": 0.5123,
+      "step": 171890
+    },
+    {
+      "epoch": 8.537796761696633,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00011698023244263436,
+      "loss": 0.4887,
+      "step": 171900
+    },
+    {
+      "epoch": 8.538293433992251,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00011694049865898479,
+      "loss": 0.4584,
+      "step": 171910
+    },
+    {
+      "epoch": 8.538790106287871,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00011690076487533525,
+      "loss": 0.4788,
+      "step": 171920
+    },
+    {
+      "epoch": 8.53928677858349,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00011686103109168571,
+      "loss": 0.5075,
+      "step": 171930
+    },
+    {
+      "epoch": 8.53978345087911,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011682129730803617,
+      "loss": 0.5154,
+      "step": 171940
+    },
+    {
+      "epoch": 8.54028012317473,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00011678156352438661,
+      "loss": 0.4921,
+      "step": 171950
+    },
+    {
+      "epoch": 8.540776795470348,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00011674182974073707,
+      "loss": 0.4844,
+      "step": 171960
+    },
+    {
+      "epoch": 8.541273467765969,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00011670209595708753,
+      "loss": 0.4855,
+      "step": 171970
+    },
+    {
+      "epoch": 8.541770140061587,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00011666236217343798,
+      "loss": 0.4719,
+      "step": 171980
+    },
+    {
+      "epoch": 8.542266812357207,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.00011662262838978841,
+      "loss": 0.4725,
+      "step": 171990
+    },
+    {
+      "epoch": 8.542763484652825,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00011658289460613887,
+      "loss": 0.4709,
+      "step": 172000
+    },
+    {
+      "epoch": 8.543260156948445,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00011654316082248933,
+      "loss": 0.4989,
+      "step": 172010
+    },
+    {
+      "epoch": 8.543756829244066,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00011650342703883979,
+      "loss": 0.4963,
+      "step": 172020
+    },
+    {
+      "epoch": 8.544253501539684,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00011646369325519022,
+      "loss": 0.4863,
+      "step": 172030
+    },
+    {
+      "epoch": 8.544750173835304,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00011642395947154068,
+      "loss": 0.4976,
+      "step": 172040
+    },
+    {
+      "epoch": 8.545246846130922,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00011638422568789113,
+      "loss": 0.4842,
+      "step": 172050
+    },
+    {
+      "epoch": 8.545743518426542,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00011634449190424159,
+      "loss": 0.4863,
+      "step": 172060
+    },
+    {
+      "epoch": 8.54624019072216,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00011630475812059205,
+      "loss": 0.4896,
+      "step": 172070
+    },
+    {
+      "epoch": 8.546736863017781,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00011626502433694248,
+      "loss": 0.495,
+      "step": 172080
+    },
+    {
+      "epoch": 8.547233535313401,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00011622529055329294,
+      "loss": 0.4976,
+      "step": 172090
+    },
+    {
+      "epoch": 8.54773020760902,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001161855567696434,
+      "loss": 0.5086,
+      "step": 172100
+    },
+    {
+      "epoch": 8.54822687990464,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00011614582298599385,
+      "loss": 0.4871,
+      "step": 172110
+    },
+    {
+      "epoch": 8.548723552200258,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001161060892023443,
+      "loss": 0.4809,
+      "step": 172120
+    },
+    {
+      "epoch": 8.549220224495878,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00011606635541869476,
+      "loss": 0.4747,
+      "step": 172130
+    },
+    {
+      "epoch": 8.549716896791496,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001160266216350452,
+      "loss": 0.5212,
+      "step": 172140
+    },
+    {
+      "epoch": 8.550213569087116,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00011598688785139566,
+      "loss": 0.4953,
+      "step": 172150
+    },
+    {
+      "epoch": 8.550710241382736,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001159471540677461,
+      "loss": 0.4812,
+      "step": 172160
+    },
+    {
+      "epoch": 8.551206913678355,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00011590742028409656,
+      "loss": 0.4604,
+      "step": 172170
+    },
+    {
+      "epoch": 8.551703585973975,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00011586768650044702,
+      "loss": 0.4723,
+      "step": 172180
+    },
+    {
+      "epoch": 8.552200258269593,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00011582795271679748,
+      "loss": 0.4686,
+      "step": 172190
+    },
+    {
+      "epoch": 8.552696930565213,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001157882189331479,
+      "loss": 0.4577,
+      "step": 172200
+    },
+    {
+      "epoch": 8.553193602860832,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00011574848514949836,
+      "loss": 0.4925,
+      "step": 172210
+    },
+    {
+      "epoch": 8.553690275156452,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00011570875136584882,
+      "loss": 0.5066,
+      "step": 172220
+    },
+    {
+      "epoch": 8.55418694745207,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00011566901758219928,
+      "loss": 0.4873,
+      "step": 172230
+    },
+    {
+      "epoch": 8.55468361974769,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00011562928379854971,
+      "loss": 0.4874,
+      "step": 172240
+    },
+    {
+      "epoch": 8.55518029204331,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00011558955001490017,
+      "loss": 0.4744,
+      "step": 172250
+    },
+    {
+      "epoch": 8.555676964338929,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011554981623125063,
+      "loss": 0.4727,
+      "step": 172260
+    },
+    {
+      "epoch": 8.556173636634549,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00011551008244760108,
+      "loss": 0.4891,
+      "step": 172270
+    },
+    {
+      "epoch": 8.556670308930167,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00011547034866395153,
+      "loss": 0.4895,
+      "step": 172280
+    },
+    {
+      "epoch": 8.557166981225787,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00011543061488030197,
+      "loss": 0.4906,
+      "step": 172290
+    },
+    {
+      "epoch": 8.557663653521406,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00011539088109665243,
+      "loss": 0.493,
+      "step": 172300
+    },
+    {
+      "epoch": 8.558160325817026,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00011535114731300289,
+      "loss": 0.5026,
+      "step": 172310
+    },
+    {
+      "epoch": 8.558656998112646,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00011531141352935333,
+      "loss": 0.4587,
+      "step": 172320
+    },
+    {
+      "epoch": 8.559153670408264,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00011527167974570379,
+      "loss": 0.4745,
+      "step": 172330
+    },
+    {
+      "epoch": 8.559650342703884,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00011523194596205425,
+      "loss": 0.4961,
+      "step": 172340
+    },
+    {
+      "epoch": 8.560147014999503,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001151922121784047,
+      "loss": 0.5004,
+      "step": 172350
+    },
+    {
+      "epoch": 8.560643687295123,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00011515247839475514,
+      "loss": 0.4838,
+      "step": 172360
+    },
+    {
+      "epoch": 8.561140359590741,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001151127446111056,
+      "loss": 0.5345,
+      "step": 172370
+    },
+    {
+      "epoch": 8.561637031886361,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011507301082745605,
+      "loss": 0.4883,
+      "step": 172380
+    },
+    {
+      "epoch": 8.562133704181981,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00011503327704380651,
+      "loss": 0.4802,
+      "step": 172390
+    },
+    {
+      "epoch": 8.5626303764776,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00011499354326015694,
+      "loss": 0.477,
+      "step": 172400
+    },
+    {
+      "epoch": 8.56312704877322,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001149538094765074,
+      "loss": 0.4884,
+      "step": 172410
+    },
+    {
+      "epoch": 8.563623721068838,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00011491407569285786,
+      "loss": 0.4693,
+      "step": 172420
+    },
+    {
+      "epoch": 8.564120393364458,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00011487434190920831,
+      "loss": 0.4957,
+      "step": 172430
+    },
+    {
+      "epoch": 8.564617065660077,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00011483460812555874,
+      "loss": 0.4737,
+      "step": 172440
+    },
+    {
+      "epoch": 8.565113737955697,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001147948743419092,
+      "loss": 0.5143,
+      "step": 172450
+    },
+    {
+      "epoch": 8.565610410251317,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011475514055825966,
+      "loss": 0.4957,
+      "step": 172460
+    },
+    {
+      "epoch": 8.566107082546935,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00011471540677461012,
+      "loss": 0.475,
+      "step": 172470
+    },
+    {
+      "epoch": 8.566603754842555,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00011467567299096056,
+      "loss": 0.4944,
+      "step": 172480
+    },
+    {
+      "epoch": 8.567100427138174,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00011463593920731102,
+      "loss": 0.4913,
+      "step": 172490
+    },
+    {
+      "epoch": 8.567597099433794,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00011459620542366148,
+      "loss": 0.4978,
+      "step": 172500
+    },
+    {
+      "epoch": 8.568093771729412,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00011455647164001194,
+      "loss": 0.4725,
+      "step": 172510
+    },
+    {
+      "epoch": 8.568590444025032,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0001145167378563624,
+      "loss": 0.5416,
+      "step": 172520
+    },
+    {
+      "epoch": 8.569087116320652,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00011447700407271282,
+      "loss": 0.4676,
+      "step": 172530
+    },
+    {
+      "epoch": 8.56958378861627,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00011443727028906328,
+      "loss": 0.4885,
+      "step": 172540
+    },
+    {
+      "epoch": 8.57008046091189,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00011439753650541374,
+      "loss": 0.4661,
+      "step": 172550
+    },
+    {
+      "epoch": 8.57057713320751,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001143578027217642,
+      "loss": 0.4737,
+      "step": 172560
+    },
+    {
+      "epoch": 8.57107380550313,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00011431806893811463,
+      "loss": 0.481,
+      "step": 172570
+    },
+    {
+      "epoch": 8.571570477798748,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00011427833515446509,
+      "loss": 0.4818,
+      "step": 172580
+    },
+    {
+      "epoch": 8.572067150094368,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00011423860137081554,
+      "loss": 0.4792,
+      "step": 172590
+    },
+    {
+      "epoch": 8.572563822389988,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.000114198867587166,
+      "loss": 0.5084,
+      "step": 172600
+    },
+    {
+      "epoch": 8.573060494685606,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00011415913380351643,
+      "loss": 0.5065,
+      "step": 172610
+    },
+    {
+      "epoch": 8.573557166981226,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00011411940001986689,
+      "loss": 0.5042,
+      "step": 172620
+    },
+    {
+      "epoch": 8.574053839276845,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00011407966623621735,
+      "loss": 0.5009,
+      "step": 172630
+    },
+    {
+      "epoch": 8.574550511572465,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001140399324525678,
+      "loss": 0.4828,
+      "step": 172640
+    },
+    {
+      "epoch": 8.575047183868083,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00011400019866891825,
+      "loss": 0.4809,
+      "step": 172650
+    },
+    {
+      "epoch": 8.575543856163703,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011396046488526871,
+      "loss": 0.4985,
+      "step": 172660
+    },
+    {
+      "epoch": 8.576040528459323,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00011392073110161917,
+      "loss": 0.4727,
+      "step": 172670
+    },
+    {
+      "epoch": 8.576537200754942,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00011388099731796962,
+      "loss": 0.4763,
+      "step": 172680
+    },
+    {
+      "epoch": 8.577033873050562,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00011384126353432005,
+      "loss": 0.4663,
+      "step": 172690
+    },
+    {
+      "epoch": 8.57753054534618,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00011380152975067051,
+      "loss": 0.4856,
+      "step": 172700
+    },
+    {
+      "epoch": 8.5780272176418,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00011376179596702097,
+      "loss": 0.4837,
+      "step": 172710
+    },
+    {
+      "epoch": 8.578523889937419,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00011372206218337143,
+      "loss": 0.49,
+      "step": 172720
+    },
+    {
+      "epoch": 8.579020562233039,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00011368232839972186,
+      "loss": 0.4778,
+      "step": 172730
+    },
+    {
+      "epoch": 8.579517234528659,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00011364259461607232,
+      "loss": 0.5232,
+      "step": 172740
+    },
+    {
+      "epoch": 8.580013906824277,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00011360286083242277,
+      "loss": 0.5044,
+      "step": 172750
+    },
+    {
+      "epoch": 8.580510579119897,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00011356312704877323,
+      "loss": 0.5027,
+      "step": 172760
+    },
+    {
+      "epoch": 8.581007251415516,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00011352339326512366,
+      "loss": 0.473,
+      "step": 172770
+    },
+    {
+      "epoch": 8.581503923711136,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00011348365948147412,
+      "loss": 0.4845,
+      "step": 172780
+    },
+    {
+      "epoch": 8.582000596006754,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00011344392569782458,
+      "loss": 0.4811,
+      "step": 172790
+    },
+    {
+      "epoch": 8.582497268302374,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00011340419191417504,
+      "loss": 0.4984,
+      "step": 172800
+    },
+    {
+      "epoch": 8.582993940597994,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00011336445813052548,
+      "loss": 0.4727,
+      "step": 172810
+    },
+    {
+      "epoch": 8.583490612893613,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00011332472434687594,
+      "loss": 0.5024,
+      "step": 172820
+    },
+    {
+      "epoch": 8.583987285189233,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001132849905632264,
+      "loss": 0.4802,
+      "step": 172830
+    },
+    {
+      "epoch": 8.584483957484851,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00011324525677957685,
+      "loss": 0.4615,
+      "step": 172840
+    },
+    {
+      "epoch": 8.584980629780471,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00011320552299592728,
+      "loss": 0.5001,
+      "step": 172850
+    },
+    {
+      "epoch": 8.58547730207609,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00011316578921227774,
+      "loss": 0.48,
+      "step": 172860
+    },
+    {
+      "epoch": 8.58597397437171,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001131260554286282,
+      "loss": 0.4833,
+      "step": 172870
+    },
+    {
+      "epoch": 8.58647064666733,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.00011308632164497866,
+      "loss": 0.4868,
+      "step": 172880
+    },
+    {
+      "epoch": 8.586967318962948,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00011304658786132909,
+      "loss": 0.4963,
+      "step": 172890
+    },
+    {
+      "epoch": 8.587463991258568,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.00011300685407767955,
+      "loss": 0.4841,
+      "step": 172900
+    },
+    {
+      "epoch": 8.587960663554187,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00011296712029403,
+      "loss": 0.5264,
+      "step": 172910
+    },
+    {
+      "epoch": 8.588457335849807,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011292738651038046,
+      "loss": 0.4982,
+      "step": 172920
+    },
+    {
+      "epoch": 8.588954008145425,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00011288765272673092,
+      "loss": 0.4825,
+      "step": 172930
+    },
+    {
+      "epoch": 8.589450680441045,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00011284791894308135,
+      "loss": 0.5479,
+      "step": 172940
+    },
+    {
+      "epoch": 8.589947352736665,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00011280818515943181,
+      "loss": 0.4673,
+      "step": 172950
+    },
+    {
+      "epoch": 8.590444025032284,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00011276845137578227,
+      "loss": 0.4555,
+      "step": 172960
+    },
+    {
+      "epoch": 8.590940697327904,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00011272871759213272,
+      "loss": 0.501,
+      "step": 172970
+    },
+    {
+      "epoch": 8.591437369623522,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00011268898380848317,
+      "loss": 0.4642,
+      "step": 172980
+    },
+    {
+      "epoch": 8.591934041919142,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00011264925002483361,
+      "loss": 0.4943,
+      "step": 172990
+    },
+    {
+      "epoch": 8.59243071421476,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00011260951624118407,
+      "loss": 0.4543,
+      "step": 173000
+    },
+    {
+      "epoch": 8.59292738651038,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00011256978245753453,
+      "loss": 0.4762,
+      "step": 173010
+    },
+    {
+      "epoch": 8.593424058806,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00011253004867388497,
+      "loss": 0.4986,
+      "step": 173020
+    },
+    {
+      "epoch": 8.593920731101619,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00011249031489023543,
+      "loss": 0.5028,
+      "step": 173030
+    },
+    {
+      "epoch": 8.59441740339724,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00011245058110658589,
+      "loss": 0.5066,
+      "step": 173040
+    },
+    {
+      "epoch": 8.594914075692857,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00011241084732293635,
+      "loss": 0.4852,
+      "step": 173050
+    },
+    {
+      "epoch": 8.595410747988478,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00011237111353928678,
+      "loss": 0.4797,
+      "step": 173060
+    },
+    {
+      "epoch": 8.595907420284096,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00011233137975563723,
+      "loss": 0.4802,
+      "step": 173070
+    },
+    {
+      "epoch": 8.596404092579716,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00011229164597198769,
+      "loss": 0.5015,
+      "step": 173080
+    },
+    {
+      "epoch": 8.596900764875336,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00011225191218833815,
+      "loss": 0.4566,
+      "step": 173090
+    },
+    {
+      "epoch": 8.597397437170955,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00011221217840468858,
+      "loss": 0.4922,
+      "step": 173100
+    },
+    {
+      "epoch": 8.597894109466575,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00011217244462103904,
+      "loss": 0.5129,
+      "step": 173110
+    },
+    {
+      "epoch": 8.598390781762193,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001121327108373895,
+      "loss": 0.5091,
+      "step": 173120
+    },
+    {
+      "epoch": 8.598887454057813,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00011209297705373995,
+      "loss": 0.5011,
+      "step": 173130
+    },
+    {
+      "epoch": 8.599384126353431,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00011205324327009038,
+      "loss": 0.4902,
+      "step": 173140
+    },
+    {
+      "epoch": 8.599880798649052,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00011201350948644084,
+      "loss": 0.4722,
+      "step": 173150
+    },
+    {
+      "epoch": 8.600377470944672,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0001119737757027913,
+      "loss": 0.4924,
+      "step": 173160
+    },
+    {
+      "epoch": 8.60087414324029,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00011193404191914176,
+      "loss": 0.5015,
+      "step": 173170
+    },
+    {
+      "epoch": 8.60137081553591,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001118943081354922,
+      "loss": 0.4771,
+      "step": 173180
+    },
+    {
+      "epoch": 8.601867487831528,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00011185457435184266,
+      "loss": 0.4567,
+      "step": 173190
+    },
+    {
+      "epoch": 8.602364160127149,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00011181484056819312,
+      "loss": 0.5131,
+      "step": 173200
+    },
+    {
+      "epoch": 8.602860832422767,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00011177510678454358,
+      "loss": 0.4573,
+      "step": 173210
+    },
+    {
+      "epoch": 8.603357504718387,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.000111735373000894,
+      "loss": 0.4729,
+      "step": 173220
+    },
+    {
+      "epoch": 8.603854177014005,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00011169563921724446,
+      "loss": 0.5176,
+      "step": 173230
+    },
+    {
+      "epoch": 8.604350849309625,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00011165590543359492,
+      "loss": 0.4943,
+      "step": 173240
+    },
+    {
+      "epoch": 8.604847521605246,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00011161617164994538,
+      "loss": 0.4649,
+      "step": 173250
+    },
+    {
+      "epoch": 8.605344193900864,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00011157643786629581,
+      "loss": 0.5047,
+      "step": 173260
+    },
+    {
+      "epoch": 8.605840866196484,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00011153670408264627,
+      "loss": 0.4832,
+      "step": 173270
+    },
+    {
+      "epoch": 8.606337538492102,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00011149697029899673,
+      "loss": 0.4853,
+      "step": 173280
+    },
+    {
+      "epoch": 8.606834210787722,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00011145723651534718,
+      "loss": 0.4814,
+      "step": 173290
+    },
+    {
+      "epoch": 8.60733088308334,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00011141750273169761,
+      "loss": 0.4646,
+      "step": 173300
+    },
+    {
+      "epoch": 8.607827555378961,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00011137776894804807,
+      "loss": 0.4631,
+      "step": 173310
+    },
+    {
+      "epoch": 8.608324227674581,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00011133803516439853,
+      "loss": 0.4459,
+      "step": 173320
+    },
+    {
+      "epoch": 8.6088208999702,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00011129830138074899,
+      "loss": 0.4805,
+      "step": 173330
+    },
+    {
+      "epoch": 8.60931757226582,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00011125856759709945,
+      "loss": 0.4772,
+      "step": 173340
+    },
+    {
+      "epoch": 8.609814244561438,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00011121883381344989,
+      "loss": 0.4769,
+      "step": 173350
+    },
+    {
+      "epoch": 8.610310916857058,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00011117910002980035,
+      "loss": 0.4761,
+      "step": 173360
+    },
+    {
+      "epoch": 8.610807589152676,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001111393662461508,
+      "loss": 0.4986,
+      "step": 173370
+    },
+    {
+      "epoch": 8.611304261448296,
+      "grad_norm": 0.234375,
+      "learning_rate": 0.00011109963246250126,
+      "loss": 0.4688,
+      "step": 173380
+    },
+    {
+      "epoch": 8.611800933743917,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001110598986788517,
+      "loss": 0.4831,
+      "step": 173390
+    },
+    {
+      "epoch": 8.612297606039535,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00011102016489520215,
+      "loss": 0.5004,
+      "step": 173400
+    },
+    {
+      "epoch": 8.612794278335155,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00011098043111155261,
+      "loss": 0.4872,
+      "step": 173410
+    },
+    {
+      "epoch": 8.613290950630773,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00011094069732790307,
+      "loss": 0.4881,
+      "step": 173420
+    },
+    {
+      "epoch": 8.613787622926393,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001109009635442535,
+      "loss": 0.4759,
+      "step": 173430
+    },
+    {
+      "epoch": 8.614284295222012,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00011086122976060396,
+      "loss": 0.489,
+      "step": 173440
+    },
+    {
+      "epoch": 8.614780967517632,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00011082149597695441,
+      "loss": 0.5264,
+      "step": 173450
+    },
+    {
+      "epoch": 8.615277639813252,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011078176219330487,
+      "loss": 0.4752,
+      "step": 173460
+    },
+    {
+      "epoch": 8.61577431210887,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001107420284096553,
+      "loss": 0.508,
+      "step": 173470
+    },
+    {
+      "epoch": 8.61627098440449,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00011070229462600576,
+      "loss": 0.4833,
+      "step": 173480
+    },
+    {
+      "epoch": 8.616767656700109,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.00011066256084235622,
+      "loss": 0.4774,
+      "step": 173490
+    },
+    {
+      "epoch": 8.617264328995729,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00011062282705870668,
+      "loss": 0.4662,
+      "step": 173500
+    },
+    {
+      "epoch": 8.617761001291347,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00011058309327505712,
+      "loss": 0.5317,
+      "step": 173510
+    },
+    {
+      "epoch": 8.618257673586967,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00011054335949140758,
+      "loss": 0.4799,
+      "step": 173520
+    },
+    {
+      "epoch": 8.618754345882587,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00011050362570775804,
+      "loss": 0.4866,
+      "step": 173530
+    },
+    {
+      "epoch": 8.619251018178206,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001104638919241085,
+      "loss": 0.4851,
+      "step": 173540
+    },
+    {
+      "epoch": 8.619747690473826,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00011042415814045892,
+      "loss": 0.4892,
+      "step": 173550
+    },
+    {
+      "epoch": 8.620244362769444,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00011038442435680938,
+      "loss": 0.5096,
+      "step": 173560
+    },
+    {
+      "epoch": 8.620741035065064,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.00011034469057315984,
+      "loss": 0.457,
+      "step": 173570
+    },
+    {
+      "epoch": 8.621237707360683,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001103049567895103,
+      "loss": 0.5146,
+      "step": 173580
+    },
+    {
+      "epoch": 8.621734379656303,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00011026522300586073,
+      "loss": 0.4439,
+      "step": 173590
+    },
+    {
+      "epoch": 8.622231051951921,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00011022548922221119,
+      "loss": 0.5048,
+      "step": 173600
+    },
+    {
+      "epoch": 8.622727724247541,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00011018575543856164,
+      "loss": 0.4622,
+      "step": 173610
+    },
+    {
+      "epoch": 8.623224396543161,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001101460216549121,
+      "loss": 0.503,
+      "step": 173620
+    },
+    {
+      "epoch": 8.62372106883878,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00011010628787126253,
+      "loss": 0.5009,
+      "step": 173630
+    },
+    {
+      "epoch": 8.6242177411344,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00011006655408761299,
+      "loss": 0.4883,
+      "step": 173640
+    },
+    {
+      "epoch": 8.624714413430018,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011002682030396345,
+      "loss": 0.4946,
+      "step": 173650
+    },
+    {
+      "epoch": 8.625211085725638,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001099870865203139,
+      "loss": 0.5035,
+      "step": 173660
+    },
+    {
+      "epoch": 8.625707758021257,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010994735273666435,
+      "loss": 0.4953,
+      "step": 173670
+    },
+    {
+      "epoch": 8.626204430316877,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010990761895301481,
+      "loss": 0.5011,
+      "step": 173680
+    },
+    {
+      "epoch": 8.626701102612497,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00010986788516936527,
+      "loss": 0.4745,
+      "step": 173690
+    },
+    {
+      "epoch": 8.627197774908115,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00010982815138571571,
+      "loss": 0.4817,
+      "step": 173700
+    },
+    {
+      "epoch": 8.627694447203735,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00010978841760206615,
+      "loss": 0.5043,
+      "step": 173710
+    },
+    {
+      "epoch": 8.628191119499354,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010974868381841661,
+      "loss": 0.4751,
+      "step": 173720
+    },
+    {
+      "epoch": 8.628687791794974,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00010970895003476707,
+      "loss": 0.478,
+      "step": 173730
+    },
+    {
+      "epoch": 8.629184464090592,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00010966921625111753,
+      "loss": 0.4573,
+      "step": 173740
+    },
+    {
+      "epoch": 8.629681136386212,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00010962948246746796,
+      "loss": 0.504,
+      "step": 173750
+    },
+    {
+      "epoch": 8.630177808681832,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00010958974868381842,
+      "loss": 0.4997,
+      "step": 173760
+    },
+    {
+      "epoch": 8.63067448097745,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00010955001490016887,
+      "loss": 0.4746,
+      "step": 173770
+    },
+    {
+      "epoch": 8.63117115327307,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00010951028111651933,
+      "loss": 0.4846,
+      "step": 173780
+    },
+    {
+      "epoch": 8.63166782556869,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00010947054733286979,
+      "loss": 0.4894,
+      "step": 173790
+    },
+    {
+      "epoch": 8.63216449786431,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00010943081354922022,
+      "loss": 0.485,
+      "step": 173800
+    },
+    {
+      "epoch": 8.632661170159928,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010939107976557068,
+      "loss": 0.4527,
+      "step": 173810
+    },
+    {
+      "epoch": 8.633157842455548,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00010935134598192114,
+      "loss": 0.51,
+      "step": 173820
+    },
+    {
+      "epoch": 8.633654514751168,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001093116121982716,
+      "loss": 0.4759,
+      "step": 173830
+    },
+    {
+      "epoch": 8.634151187046786,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00010927187841462202,
+      "loss": 0.4881,
+      "step": 173840
+    },
+    {
+      "epoch": 8.634647859342406,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.00010923214463097248,
+      "loss": 0.5048,
+      "step": 173850
+    },
+    {
+      "epoch": 8.635144531638025,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00010919241084732294,
+      "loss": 0.522,
+      "step": 173860
+    },
+    {
+      "epoch": 8.635641203933645,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001091526770636734,
+      "loss": 0.4871,
+      "step": 173870
+    },
+    {
+      "epoch": 8.636137876229263,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00010911294328002384,
+      "loss": 0.4856,
+      "step": 173880
+    },
+    {
+      "epoch": 8.636634548524883,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001090732094963743,
+      "loss": 0.5151,
+      "step": 173890
+    },
+    {
+      "epoch": 8.637131220820503,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00010903347571272476,
+      "loss": 0.4928,
+      "step": 173900
+    },
+    {
+      "epoch": 8.637627893116122,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00010899374192907522,
+      "loss": 0.4525,
+      "step": 173910
+    },
+    {
+      "epoch": 8.638124565411742,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00010895400814542565,
+      "loss": 0.4856,
+      "step": 173920
+    },
+    {
+      "epoch": 8.63862123770736,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001089142743617761,
+      "loss": 0.4888,
+      "step": 173930
+    },
+    {
+      "epoch": 8.63911791000298,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010887454057812656,
+      "loss": 0.4846,
+      "step": 173940
+    },
+    {
+      "epoch": 8.639614582298599,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00010883480679447702,
+      "loss": 0.4762,
+      "step": 173950
+    },
+    {
+      "epoch": 8.640111254594219,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010879507301082745,
+      "loss": 0.4901,
+      "step": 173960
+    },
+    {
+      "epoch": 8.640607926889839,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00010875533922717791,
+      "loss": 0.4973,
+      "step": 173970
+    },
+    {
+      "epoch": 8.641104599185457,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00010871560544352837,
+      "loss": 0.4961,
+      "step": 173980
+    },
+    {
+      "epoch": 8.641601271481077,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00010867587165987882,
+      "loss": 0.484,
+      "step": 173990
+    },
+    {
+      "epoch": 8.642097943776696,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010863613787622925,
+      "loss": 0.4745,
+      "step": 174000
+    },
+    {
+      "epoch": 8.642594616072316,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00010859640409257971,
+      "loss": 0.4634,
+      "step": 174010
+    },
+    {
+      "epoch": 8.643091288367934,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00010855667030893017,
+      "loss": 0.4951,
+      "step": 174020
+    },
+    {
+      "epoch": 8.643587960663554,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010851693652528063,
+      "loss": 0.4963,
+      "step": 174030
+    },
+    {
+      "epoch": 8.644084632959174,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00010847720274163107,
+      "loss": 0.4981,
+      "step": 174040
+    },
+    {
+      "epoch": 8.644581305254793,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00010843746895798153,
+      "loss": 0.5023,
+      "step": 174050
+    },
+    {
+      "epoch": 8.645077977550413,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00010839773517433199,
+      "loss": 0.509,
+      "step": 174060
+    },
+    {
+      "epoch": 8.645574649846031,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00010835800139068245,
+      "loss": 0.465,
+      "step": 174070
+    },
+    {
+      "epoch": 8.646071322141651,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00010831826760703288,
+      "loss": 0.496,
+      "step": 174080
+    },
+    {
+      "epoch": 8.64656799443727,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00010827853382338333,
+      "loss": 0.4552,
+      "step": 174090
+    },
+    {
+      "epoch": 8.64706466673289,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00010823880003973379,
+      "loss": 0.4964,
+      "step": 174100
+    },
+    {
+      "epoch": 8.64756133902851,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00010819906625608425,
+      "loss": 0.4778,
+      "step": 174110
+    },
+    {
+      "epoch": 8.648058011324128,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.00010815933247243468,
+      "loss": 0.4903,
+      "step": 174120
+    },
+    {
+      "epoch": 8.648554683619748,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00010811959868878514,
+      "loss": 0.4899,
+      "step": 174130
+    },
+    {
+      "epoch": 8.649051355915367,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001080798649051356,
+      "loss": 0.5264,
+      "step": 174140
+    },
+    {
+      "epoch": 8.649548028210987,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00010804013112148605,
+      "loss": 0.4519,
+      "step": 174150
+    },
+    {
+      "epoch": 8.650044700506605,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00010800039733783648,
+      "loss": 0.4579,
+      "step": 174160
+    },
+    {
+      "epoch": 8.650541372802225,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00010796066355418694,
+      "loss": 0.4924,
+      "step": 174170
+    },
+    {
+      "epoch": 8.651038045097845,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001079209297705374,
+      "loss": 0.4829,
+      "step": 174180
+    },
+    {
+      "epoch": 8.651534717393464,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00010788119598688786,
+      "loss": 0.495,
+      "step": 174190
+    },
+    {
+      "epoch": 8.652031389689084,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00010784146220323832,
+      "loss": 0.496,
+      "step": 174200
+    },
+    {
+      "epoch": 8.652528061984702,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00010780172841958876,
+      "loss": 0.4862,
+      "step": 174210
+    },
+    {
+      "epoch": 8.653024734280322,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010776199463593922,
+      "loss": 0.4759,
+      "step": 174220
+    },
+    {
+      "epoch": 8.65352140657594,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010772226085228968,
+      "loss": 0.4728,
+      "step": 174230
+    },
+    {
+      "epoch": 8.65401807887156,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00010768252706864013,
+      "loss": 0.4325,
+      "step": 174240
+    },
+    {
+      "epoch": 8.65451475116718,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00010764279328499056,
+      "loss": 0.5091,
+      "step": 174250
+    },
+    {
+      "epoch": 8.655011423462799,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010760305950134102,
+      "loss": 0.5027,
+      "step": 174260
+    },
+    {
+      "epoch": 8.65550809575842,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00010756332571769148,
+      "loss": 0.4376,
+      "step": 174270
+    },
+    {
+      "epoch": 8.656004768054038,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010752359193404194,
+      "loss": 0.4668,
+      "step": 174280
+    },
+    {
+      "epoch": 8.656501440349658,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010748385815039237,
+      "loss": 0.4832,
+      "step": 174290
+    },
+    {
+      "epoch": 8.656998112645276,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00010744412436674283,
+      "loss": 0.4939,
+      "step": 174300
+    },
+    {
+      "epoch": 8.657494784940896,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00010740439058309328,
+      "loss": 0.4924,
+      "step": 174310
+    },
+    {
+      "epoch": 8.657991457236516,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00010736465679944374,
+      "loss": 0.4954,
+      "step": 174320
+    },
+    {
+      "epoch": 8.658488129532135,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00010732492301579417,
+      "loss": 0.4583,
+      "step": 174330
+    },
+    {
+      "epoch": 8.658984801827755,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00010728518923214463,
+      "loss": 0.4932,
+      "step": 174340
+    },
+    {
+      "epoch": 8.659481474123373,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00010724545544849509,
+      "loss": 0.4873,
+      "step": 174350
+    },
+    {
+      "epoch": 8.659978146418993,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00010720572166484555,
+      "loss": 0.4775,
+      "step": 174360
+    },
+    {
+      "epoch": 8.660474818714611,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00010716598788119599,
+      "loss": 0.4945,
+      "step": 174370
+    },
+    {
+      "epoch": 8.660971491010232,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00010712625409754645,
+      "loss": 0.4903,
+      "step": 174380
+    },
+    {
+      "epoch": 8.661468163305852,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001070865203138969,
+      "loss": 0.4777,
+      "step": 174390
+    },
+    {
+      "epoch": 8.66196483560147,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00010704678653024735,
+      "loss": 0.4987,
+      "step": 174400
+    },
+    {
+      "epoch": 8.66246150789709,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001070070527465978,
+      "loss": 0.4975,
+      "step": 174410
+    },
+    {
+      "epoch": 8.662958180192708,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00010696731896294825,
+      "loss": 0.4891,
+      "step": 174420
+    },
+    {
+      "epoch": 8.663454852488329,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00010692758517929871,
+      "loss": 0.463,
+      "step": 174430
+    },
+    {
+      "epoch": 8.663951524783947,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010688785139564917,
+      "loss": 0.4858,
+      "step": 174440
+    },
+    {
+      "epoch": 8.664448197079567,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001068481176119996,
+      "loss": 0.45,
+      "step": 174450
+    },
+    {
+      "epoch": 8.664944869375187,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00010680838382835006,
+      "loss": 0.4699,
+      "step": 174460
+    },
+    {
+      "epoch": 8.665441541670805,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010676865004470051,
+      "loss": 0.4936,
+      "step": 174470
+    },
+    {
+      "epoch": 8.665938213966426,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00010672891626105097,
+      "loss": 0.4886,
+      "step": 174480
+    },
+    {
+      "epoch": 8.666434886262044,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001066891824774014,
+      "loss": 0.4716,
+      "step": 174490
+    },
+    {
+      "epoch": 8.666931558557664,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00010664944869375186,
+      "loss": 0.4806,
+      "step": 174500
+    },
+    {
+      "epoch": 8.667428230853282,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00010660971491010232,
+      "loss": 0.4962,
+      "step": 174510
+    },
+    {
+      "epoch": 8.667924903148903,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00010656998112645278,
+      "loss": 0.4882,
+      "step": 174520
+    },
+    {
+      "epoch": 8.668421575444523,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00010653024734280322,
+      "loss": 0.5009,
+      "step": 174530
+    },
+    {
+      "epoch": 8.668918247740141,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00010649051355915368,
+      "loss": 0.5219,
+      "step": 174540
+    },
+    {
+      "epoch": 8.669414920035761,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00010645077977550412,
+      "loss": 0.5139,
+      "step": 174550
+    },
+    {
+      "epoch": 8.66991159233138,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00010641104599185458,
+      "loss": 0.4798,
+      "step": 174560
+    },
+    {
+      "epoch": 8.670408264627,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00010637131220820502,
+      "loss": 0.481,
+      "step": 174570
+    },
+    {
+      "epoch": 8.670904936922618,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00010633157842455548,
+      "loss": 0.4926,
+      "step": 174580
+    },
+    {
+      "epoch": 8.671401609218238,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010629184464090594,
+      "loss": 0.4913,
+      "step": 174590
+    },
+    {
+      "epoch": 8.671898281513856,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001062521108572564,
+      "loss": 0.4824,
+      "step": 174600
+    },
+    {
+      "epoch": 8.672394953809476,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00010621237707360685,
+      "loss": 0.4928,
+      "step": 174610
+    },
+    {
+      "epoch": 8.672891626105097,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010617264328995729,
+      "loss": 0.5017,
+      "step": 174620
+    },
+    {
+      "epoch": 8.673388298400715,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00010613290950630774,
+      "loss": 0.5175,
+      "step": 174630
+    },
+    {
+      "epoch": 8.673884970696335,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001060931757226582,
+      "loss": 0.5026,
+      "step": 174640
+    },
+    {
+      "epoch": 8.674381642991953,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010605344193900866,
+      "loss": 0.4927,
+      "step": 174650
+    },
+    {
+      "epoch": 8.674878315287573,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010601370815535909,
+      "loss": 0.4765,
+      "step": 174660
+    },
+    {
+      "epoch": 8.675374987583192,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00010597397437170955,
+      "loss": 0.4778,
+      "step": 174670
+    },
+    {
+      "epoch": 8.675871659878812,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00010593424058806,
+      "loss": 0.4864,
+      "step": 174680
+    },
+    {
+      "epoch": 8.676368332174432,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00010589450680441046,
+      "loss": 0.5022,
+      "step": 174690
+    },
+    {
+      "epoch": 8.67686500447005,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0001058547730207609,
+      "loss": 0.4873,
+      "step": 174700
+    },
+    {
+      "epoch": 8.67736167676567,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00010581503923711135,
+      "loss": 0.4798,
+      "step": 174710
+    },
+    {
+      "epoch": 8.677858349061289,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00010577530545346181,
+      "loss": 0.4645,
+      "step": 174720
+    },
+    {
+      "epoch": 8.678355021356909,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00010573557166981227,
+      "loss": 0.498,
+      "step": 174730
+    },
+    {
+      "epoch": 8.678851693652527,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00010569583788616271,
+      "loss": 0.474,
+      "step": 174740
+    },
+    {
+      "epoch": 8.679348365948147,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00010565610410251317,
+      "loss": 0.4782,
+      "step": 174750
+    },
+    {
+      "epoch": 8.679845038243768,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010561637031886363,
+      "loss": 0.4497,
+      "step": 174760
+    },
+    {
+      "epoch": 8.680341710539386,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00010557663653521408,
+      "loss": 0.4833,
+      "step": 174770
+    },
+    {
+      "epoch": 8.680838382835006,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00010553690275156452,
+      "loss": 0.4923,
+      "step": 174780
+    },
+    {
+      "epoch": 8.681335055130624,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00010549716896791497,
+      "loss": 0.4748,
+      "step": 174790
+    },
+    {
+      "epoch": 8.681831727426244,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00010545743518426543,
+      "loss": 0.4766,
+      "step": 174800
+    },
+    {
+      "epoch": 8.682328399721863,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010541770140061589,
+      "loss": 0.5165,
+      "step": 174810
+    },
+    {
+      "epoch": 8.682825072017483,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00010537796761696632,
+      "loss": 0.4782,
+      "step": 174820
+    },
+    {
+      "epoch": 8.683321744313103,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00010533823383331678,
+      "loss": 0.482,
+      "step": 174830
+    },
+    {
+      "epoch": 8.683818416608721,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00010529850004966724,
+      "loss": 0.508,
+      "step": 174840
+    },
+    {
+      "epoch": 8.684315088904341,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00010525876626601769,
+      "loss": 0.4965,
+      "step": 174850
+    },
+    {
+      "epoch": 8.68481176119996,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00010521903248236812,
+      "loss": 0.4568,
+      "step": 174860
+    },
+    {
+      "epoch": 8.68530843349558,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00010517929869871858,
+      "loss": 0.4496,
+      "step": 174870
+    },
+    {
+      "epoch": 8.685805105791198,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00010513956491506904,
+      "loss": 0.4925,
+      "step": 174880
+    },
+    {
+      "epoch": 8.686301778086818,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001050998311314195,
+      "loss": 0.4353,
+      "step": 174890
+    },
+    {
+      "epoch": 8.686798450382438,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00010506009734776994,
+      "loss": 0.4739,
+      "step": 174900
+    },
+    {
+      "epoch": 8.687295122678057,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001050203635641204,
+      "loss": 0.4828,
+      "step": 174910
+    },
+    {
+      "epoch": 8.687791794973677,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00010498062978047086,
+      "loss": 0.4693,
+      "step": 174920
+    },
+    {
+      "epoch": 8.688288467269295,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00010494089599682131,
+      "loss": 0.5162,
+      "step": 174930
+    },
+    {
+      "epoch": 8.688785139564915,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00010490116221317175,
+      "loss": 0.5078,
+      "step": 174940
+    },
+    {
+      "epoch": 8.689281811860534,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001048614284295222,
+      "loss": 0.5057,
+      "step": 174950
+    },
+    {
+      "epoch": 8.689778484156154,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010482169464587266,
+      "loss": 0.4784,
+      "step": 174960
+    },
+    {
+      "epoch": 8.690275156451774,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00010478196086222312,
+      "loss": 0.5044,
+      "step": 174970
+    },
+    {
+      "epoch": 8.690771828747392,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00010474222707857355,
+      "loss": 0.4781,
+      "step": 174980
+    },
+    {
+      "epoch": 8.691268501043012,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00010470249329492401,
+      "loss": 0.473,
+      "step": 174990
+    },
+    {
+      "epoch": 8.69176517333863,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00010466275951127447,
+      "loss": 0.4808,
+      "step": 175000
+    },
+    {
+      "epoch": 8.69226184563425,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010462302572762492,
+      "loss": 0.4748,
+      "step": 175010
+    },
+    {
+      "epoch": 8.69275851792987,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010458329194397538,
+      "loss": 0.4727,
+      "step": 175020
+    },
+    {
+      "epoch": 8.69325519022549,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00010454355816032581,
+      "loss": 0.4716,
+      "step": 175030
+    },
+    {
+      "epoch": 8.693751862521108,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00010450382437667627,
+      "loss": 0.4809,
+      "step": 175040
+    },
+    {
+      "epoch": 8.694248534816728,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00010446409059302673,
+      "loss": 0.492,
+      "step": 175050
+    },
+    {
+      "epoch": 8.694745207112348,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00010442435680937719,
+      "loss": 0.4875,
+      "step": 175060
+    },
+    {
+      "epoch": 8.695241879407966,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00010438462302572763,
+      "loss": 0.4855,
+      "step": 175070
+    },
+    {
+      "epoch": 8.695738551703586,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00010434488924207809,
+      "loss": 0.4954,
+      "step": 175080
+    },
+    {
+      "epoch": 8.696235223999205,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00010430515545842854,
+      "loss": 0.4874,
+      "step": 175090
+    },
+    {
+      "epoch": 8.696731896294825,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00010426542167477899,
+      "loss": 0.4979,
+      "step": 175100
+    },
+    {
+      "epoch": 8.697228568590443,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00010422568789112943,
+      "loss": 0.5075,
+      "step": 175110
+    },
+    {
+      "epoch": 8.697725240886063,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.00010418595410747989,
+      "loss": 0.4886,
+      "step": 175120
+    },
+    {
+      "epoch": 8.698221913181683,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00010414622032383035,
+      "loss": 0.4603,
+      "step": 175130
+    },
+    {
+      "epoch": 8.698718585477302,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00010410648654018081,
+      "loss": 0.4983,
+      "step": 175140
+    },
+    {
+      "epoch": 8.699215257772922,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00010406675275653124,
+      "loss": 0.4894,
+      "step": 175150
+    },
+    {
+      "epoch": 8.69971193006854,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001040270189728817,
+      "loss": 0.4874,
+      "step": 175160
+    },
+    {
+      "epoch": 8.70020860236416,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00010398728518923215,
+      "loss": 0.4918,
+      "step": 175170
+    },
+    {
+      "epoch": 8.700705274659779,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00010394755140558261,
+      "loss": 0.5007,
+      "step": 175180
+    },
+    {
+      "epoch": 8.701201946955399,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00010390781762193304,
+      "loss": 0.478,
+      "step": 175190
+    },
+    {
+      "epoch": 8.701698619251019,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001038680838382835,
+      "loss": 0.4908,
+      "step": 175200
+    },
+    {
+      "epoch": 8.702195291546637,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00010382835005463396,
+      "loss": 0.4689,
+      "step": 175210
+    },
+    {
+      "epoch": 8.702691963842257,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00010378861627098442,
+      "loss": 0.4518,
+      "step": 175220
+    },
+    {
+      "epoch": 8.703188636137876,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010374888248733486,
+      "loss": 0.4886,
+      "step": 175230
+    },
+    {
+      "epoch": 8.703685308433496,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00010370914870368532,
+      "loss": 0.5028,
+      "step": 175240
+    },
+    {
+      "epoch": 8.704181980729114,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010366941492003576,
+      "loss": 0.4998,
+      "step": 175250
+    },
+    {
+      "epoch": 8.704678653024734,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00010362968113638622,
+      "loss": 0.4962,
+      "step": 175260
+    },
+    {
+      "epoch": 8.705175325320354,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00010358994735273666,
+      "loss": 0.4612,
+      "step": 175270
+    },
+    {
+      "epoch": 8.705671997615973,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00010355021356908712,
+      "loss": 0.4968,
+      "step": 175280
+    },
+    {
+      "epoch": 8.706168669911593,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00010351047978543758,
+      "loss": 0.4972,
+      "step": 175290
+    },
+    {
+      "epoch": 8.706665342207211,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00010347074600178804,
+      "loss": 0.4681,
+      "step": 175300
+    },
+    {
+      "epoch": 8.707162014502831,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00010343101221813847,
+      "loss": 0.459,
+      "step": 175310
+    },
+    {
+      "epoch": 8.70765868679845,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00010339127843448893,
+      "loss": 0.4841,
+      "step": 175320
+    },
+    {
+      "epoch": 8.70815535909407,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00010335154465083938,
+      "loss": 0.4949,
+      "step": 175330
+    },
+    {
+      "epoch": 8.70865203138969,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00010331181086718984,
+      "loss": 0.4655,
+      "step": 175340
+    },
+    {
+      "epoch": 8.709148703685308,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00010327207708354027,
+      "loss": 0.4858,
+      "step": 175350
+    },
+    {
+      "epoch": 8.709645375980928,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00010323234329989073,
+      "loss": 0.4825,
+      "step": 175360
+    },
+    {
+      "epoch": 8.710142048276547,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00010319260951624119,
+      "loss": 0.4749,
+      "step": 175370
+    },
+    {
+      "epoch": 8.710638720572167,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00010315287573259164,
+      "loss": 0.5096,
+      "step": 175380
+    },
+    {
+      "epoch": 8.711135392867785,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00010311314194894208,
+      "loss": 0.5066,
+      "step": 175390
+    },
+    {
+      "epoch": 8.711632065163405,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00010307340816529253,
+      "loss": 0.4953,
+      "step": 175400
+    },
+    {
+      "epoch": 8.712128737459025,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00010303367438164299,
+      "loss": 0.4979,
+      "step": 175410
+    },
+    {
+      "epoch": 8.712625409754644,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00010299394059799345,
+      "loss": 0.5207,
+      "step": 175420
+    },
+    {
+      "epoch": 8.713122082050264,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0001029542068143439,
+      "loss": 0.4455,
+      "step": 175430
+    },
+    {
+      "epoch": 8.713618754345882,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00010291447303069435,
+      "loss": 0.4792,
+      "step": 175440
+    },
+    {
+      "epoch": 8.714115426641502,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00010287473924704481,
+      "loss": 0.4731,
+      "step": 175450
+    },
+    {
+      "epoch": 8.71461209893712,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.00010283500546339527,
+      "loss": 0.4873,
+      "step": 175460
+    },
+    {
+      "epoch": 8.71510877123274,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00010279527167974572,
+      "loss": 0.4752,
+      "step": 175470
+    },
+    {
+      "epoch": 8.71560544352836,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00010275553789609616,
+      "loss": 0.5031,
+      "step": 175480
+    },
+    {
+      "epoch": 8.716102115823979,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00010271580411244661,
+      "loss": 0.458,
+      "step": 175490
+    },
+    {
+      "epoch": 8.7165987881196,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00010267607032879707,
+      "loss": 0.4615,
+      "step": 175500
+    },
+    {
+      "epoch": 8.717095460415218,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00010263633654514753,
+      "loss": 0.4954,
+      "step": 175510
+    },
+    {
+      "epoch": 8.717592132710838,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00010259660276149796,
+      "loss": 0.4803,
+      "step": 175520
+    },
+    {
+      "epoch": 8.718088805006456,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.00010255686897784842,
+      "loss": 0.4735,
+      "step": 175530
+    },
+    {
+      "epoch": 8.718585477302076,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00010251713519419887,
+      "loss": 0.4805,
+      "step": 175540
+    },
+    {
+      "epoch": 8.719082149597696,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00010247740141054933,
+      "loss": 0.47,
+      "step": 175550
+    },
+    {
+      "epoch": 8.719578821893315,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010243766762689976,
+      "loss": 0.4884,
+      "step": 175560
+    },
+    {
+      "epoch": 8.720075494188935,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00010239793384325022,
+      "loss": 0.4689,
+      "step": 175570
+    },
+    {
+      "epoch": 8.720572166484553,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00010235820005960068,
+      "loss": 0.5375,
+      "step": 175580
+    },
+    {
+      "epoch": 8.721068838780173,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00010231846627595114,
+      "loss": 0.4968,
+      "step": 175590
+    },
+    {
+      "epoch": 8.721565511075791,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00010227873249230158,
+      "loss": 0.47,
+      "step": 175600
+    },
+    {
+      "epoch": 8.722062183371412,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00010223899870865204,
+      "loss": 0.5143,
+      "step": 175610
+    },
+    {
+      "epoch": 8.722558855667032,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001021992649250025,
+      "loss": 0.4678,
+      "step": 175620
+    },
+    {
+      "epoch": 8.72305552796265,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00010215953114135295,
+      "loss": 0.5128,
+      "step": 175630
+    },
+    {
+      "epoch": 8.72355220025827,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00010211979735770339,
+      "loss": 0.5179,
+      "step": 175640
+    },
+    {
+      "epoch": 8.724048872553889,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00010208006357405384,
+      "loss": 0.4985,
+      "step": 175650
+    },
+    {
+      "epoch": 8.724545544849509,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001020403297904043,
+      "loss": 0.4765,
+      "step": 175660
+    },
+    {
+      "epoch": 8.725042217145127,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010200059600675476,
+      "loss": 0.4867,
+      "step": 175670
+    },
+    {
+      "epoch": 8.725538889440747,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00010196086222310519,
+      "loss": 0.4989,
+      "step": 175680
+    },
+    {
+      "epoch": 8.726035561736367,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00010192112843945565,
+      "loss": 0.4549,
+      "step": 175690
+    },
+    {
+      "epoch": 8.726532234031986,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001018813946558061,
+      "loss": 0.4934,
+      "step": 175700
+    },
+    {
+      "epoch": 8.727028906327606,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00010184166087215656,
+      "loss": 0.4957,
+      "step": 175710
+    },
+    {
+      "epoch": 8.727525578623224,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.000101801927088507,
+      "loss": 0.4774,
+      "step": 175720
+    },
+    {
+      "epoch": 8.728022250918844,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00010176219330485745,
+      "loss": 0.4926,
+      "step": 175730
+    },
+    {
+      "epoch": 8.728518923214462,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00010172245952120791,
+      "loss": 0.4866,
+      "step": 175740
+    },
+    {
+      "epoch": 8.729015595510083,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00010168272573755837,
+      "loss": 0.4714,
+      "step": 175750
+    },
+    {
+      "epoch": 8.729512267805703,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010164299195390881,
+      "loss": 0.4632,
+      "step": 175760
+    },
+    {
+      "epoch": 8.730008940101321,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00010160325817025927,
+      "loss": 0.5331,
+      "step": 175770
+    },
+    {
+      "epoch": 8.730505612396941,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00010156352438660973,
+      "loss": 0.5019,
+      "step": 175780
+    },
+    {
+      "epoch": 8.73100228469256,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00010152379060296018,
+      "loss": 0.4965,
+      "step": 175790
+    },
+    {
+      "epoch": 8.73149895698818,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010148405681931062,
+      "loss": 0.4733,
+      "step": 175800
+    },
+    {
+      "epoch": 8.731995629283798,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00010144432303566107,
+      "loss": 0.4864,
+      "step": 175810
+    },
+    {
+      "epoch": 8.732492301579418,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00010140458925201153,
+      "loss": 0.5009,
+      "step": 175820
+    },
+    {
+      "epoch": 8.732988973875038,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00010136485546836199,
+      "loss": 0.5191,
+      "step": 175830
+    },
+    {
+      "epoch": 8.733485646170656,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010132512168471242,
+      "loss": 0.4973,
+      "step": 175840
+    },
+    {
+      "epoch": 8.733982318466277,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010128538790106288,
+      "loss": 0.4668,
+      "step": 175850
+    },
+    {
+      "epoch": 8.734478990761895,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00010124565411741333,
+      "loss": 0.4487,
+      "step": 175860
+    },
+    {
+      "epoch": 8.734975663057515,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00010120592033376379,
+      "loss": 0.5054,
+      "step": 175870
+    },
+    {
+      "epoch": 8.735472335353133,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00010116618655011425,
+      "loss": 0.4952,
+      "step": 175880
+    },
+    {
+      "epoch": 8.735969007648754,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00010112645276646468,
+      "loss": 0.4726,
+      "step": 175890
+    },
+    {
+      "epoch": 8.736465679944374,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00010108671898281514,
+      "loss": 0.4965,
+      "step": 175900
+    },
+    {
+      "epoch": 8.736962352239992,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001010469851991656,
+      "loss": 0.4819,
+      "step": 175910
+    },
+    {
+      "epoch": 8.737459024535612,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010100725141551605,
+      "loss": 0.5017,
+      "step": 175920
+    },
+    {
+      "epoch": 8.73795569683123,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001009675176318665,
+      "loss": 0.4949,
+      "step": 175930
+    },
+    {
+      "epoch": 8.73845236912685,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00010092778384821696,
+      "loss": 0.4852,
+      "step": 175940
+    },
+    {
+      "epoch": 8.738949041422469,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001008880500645674,
+      "loss": 0.4902,
+      "step": 175950
+    },
+    {
+      "epoch": 8.739445713718089,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00010084831628091786,
+      "loss": 0.4961,
+      "step": 175960
+    },
+    {
+      "epoch": 8.739942386013709,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001008085824972683,
+      "loss": 0.5117,
+      "step": 175970
+    },
+    {
+      "epoch": 8.740439058309327,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00010076884871361876,
+      "loss": 0.4918,
+      "step": 175980
+    },
+    {
+      "epoch": 8.740935730604948,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00010072911492996922,
+      "loss": 0.4626,
+      "step": 175990
+    },
+    {
+      "epoch": 8.741432402900566,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010068938114631968,
+      "loss": 0.4798,
+      "step": 176000
+    },
+    {
+      "epoch": 8.741929075196186,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00010064964736267011,
+      "loss": 0.4802,
+      "step": 176010
+    },
+    {
+      "epoch": 8.742425747491804,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00010060991357902056,
+      "loss": 0.4873,
+      "step": 176020
+    },
+    {
+      "epoch": 8.742922419787424,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00010057017979537102,
+      "loss": 0.4846,
+      "step": 176030
+    },
+    {
+      "epoch": 8.743419092083043,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010053044601172148,
+      "loss": 0.4633,
+      "step": 176040
+    },
+    {
+      "epoch": 8.743915764378663,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00010049071222807191,
+      "loss": 0.4753,
+      "step": 176050
+    },
+    {
+      "epoch": 8.744412436674283,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00010045097844442237,
+      "loss": 0.49,
+      "step": 176060
+    },
+    {
+      "epoch": 8.744909108969901,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.00010041124466077283,
+      "loss": 0.5021,
+      "step": 176070
+    },
+    {
+      "epoch": 8.745405781265521,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00010037151087712328,
+      "loss": 0.4969,
+      "step": 176080
+    },
+    {
+      "epoch": 8.74590245356114,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010033177709347373,
+      "loss": 0.4906,
+      "step": 176090
+    },
+    {
+      "epoch": 8.74639912585676,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010029204330982417,
+      "loss": 0.4859,
+      "step": 176100
+    },
+    {
+      "epoch": 8.746895798152378,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00010025230952617463,
+      "loss": 0.4801,
+      "step": 176110
+    },
+    {
+      "epoch": 8.747392470447998,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010021257574252509,
+      "loss": 0.4762,
+      "step": 176120
+    },
+    {
+      "epoch": 8.747889142743619,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00010017284195887553,
+      "loss": 0.4778,
+      "step": 176130
+    },
+    {
+      "epoch": 8.748385815039237,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.00010013310817522599,
+      "loss": 0.5307,
+      "step": 176140
+    },
+    {
+      "epoch": 8.748882487334857,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00010009337439157645,
+      "loss": 0.4938,
+      "step": 176150
+    },
+    {
+      "epoch": 8.749379159630475,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001000536406079269,
+      "loss": 0.4661,
+      "step": 176160
+    },
+    {
+      "epoch": 8.749875831926095,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00010001390682427734,
+      "loss": 0.483,
+      "step": 176170
+    },
+    {
+      "epoch": 8.750372504221714,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.99741730406278e-05,
+      "loss": 0.4883,
+      "step": 176180
+    },
+    {
+      "epoch": 8.750869176517334,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 9.993443925697825e-05,
+      "loss": 0.4923,
+      "step": 176190
+    },
+    {
+      "epoch": 8.751365848812954,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.98947054733287e-05,
+      "loss": 0.4881,
+      "step": 176200
+    },
+    {
+      "epoch": 8.751862521108572,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.985497168967915e-05,
+      "loss": 0.4714,
+      "step": 176210
+    },
+    {
+      "epoch": 8.752359193404192,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.98152379060296e-05,
+      "loss": 0.4527,
+      "step": 176220
+    },
+    {
+      "epoch": 8.75285586569981,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.977550412238006e-05,
+      "loss": 0.4658,
+      "step": 176230
+    },
+    {
+      "epoch": 8.753352537995431,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 9.97357703387305e-05,
+      "loss": 0.4686,
+      "step": 176240
+    },
+    {
+      "epoch": 8.75384921029105,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.969603655508096e-05,
+      "loss": 0.4648,
+      "step": 176250
+    },
+    {
+      "epoch": 8.75434588258667,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.96563027714314e-05,
+      "loss": 0.4743,
+      "step": 176260
+    },
+    {
+      "epoch": 8.75484255488229,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.961656898778186e-05,
+      "loss": 0.4739,
+      "step": 176270
+    },
+    {
+      "epoch": 8.755339227177908,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.957683520413232e-05,
+      "loss": 0.5123,
+      "step": 176280
+    },
+    {
+      "epoch": 8.755835899473528,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 9.953710142048278e-05,
+      "loss": 0.4968,
+      "step": 176290
+    },
+    {
+      "epoch": 8.756332571769146,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 9.949736763683322e-05,
+      "loss": 0.4837,
+      "step": 176300
+    },
+    {
+      "epoch": 8.756829244064766,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.945763385318368e-05,
+      "loss": 0.5075,
+      "step": 176310
+    },
+    {
+      "epoch": 8.757325916360385,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 9.941790006953414e-05,
+      "loss": 0.4982,
+      "step": 176320
+    },
+    {
+      "epoch": 8.757822588656005,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.937816628588458e-05,
+      "loss": 0.4884,
+      "step": 176330
+    },
+    {
+      "epoch": 8.758319260951625,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.933843250223504e-05,
+      "loss": 0.495,
+      "step": 176340
+    },
+    {
+      "epoch": 8.758815933247243,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.929869871858548e-05,
+      "loss": 0.4847,
+      "step": 176350
+    },
+    {
+      "epoch": 8.759312605542863,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 9.925896493493594e-05,
+      "loss": 0.4984,
+      "step": 176360
+    },
+    {
+      "epoch": 8.759809277838482,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.921923115128638e-05,
+      "loss": 0.4804,
+      "step": 176370
+    },
+    {
+      "epoch": 8.760305950134102,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.917949736763684e-05,
+      "loss": 0.5331,
+      "step": 176380
+    },
+    {
+      "epoch": 8.76080262242972,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.913976358398729e-05,
+      "loss": 0.5116,
+      "step": 176390
+    },
+    {
+      "epoch": 8.76129929472534,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 9.910002980033774e-05,
+      "loss": 0.4954,
+      "step": 176400
+    },
+    {
+      "epoch": 8.761795967020959,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.906029601668819e-05,
+      "loss": 0.5152,
+      "step": 176410
+    },
+    {
+      "epoch": 8.762292639316579,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.902056223303865e-05,
+      "loss": 0.4826,
+      "step": 176420
+    },
+    {
+      "epoch": 8.762789311612199,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.898082844938909e-05,
+      "loss": 0.4719,
+      "step": 176430
+    },
+    {
+      "epoch": 8.763285983907817,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.894109466573955e-05,
+      "loss": 0.4763,
+      "step": 176440
+    },
+    {
+      "epoch": 8.763782656203437,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.890136088208999e-05,
+      "loss": 0.4884,
+      "step": 176450
+    },
+    {
+      "epoch": 8.764279328499056,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.886162709844045e-05,
+      "loss": 0.4742,
+      "step": 176460
+    },
+    {
+      "epoch": 8.764776000794676,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.882189331479091e-05,
+      "loss": 0.4734,
+      "step": 176470
+    },
+    {
+      "epoch": 8.765272673090294,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.878215953114137e-05,
+      "loss": 0.5066,
+      "step": 176480
+    },
+    {
+      "epoch": 8.765769345385914,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.874242574749181e-05,
+      "loss": 0.4681,
+      "step": 176490
+    },
+    {
+      "epoch": 8.766266017681534,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.870269196384227e-05,
+      "loss": 0.4684,
+      "step": 176500
+    },
+    {
+      "epoch": 8.766762689977153,
+      "grad_norm": 0.1796875,
+      "learning_rate": 9.866295818019271e-05,
+      "loss": 0.4754,
+      "step": 176510
+    },
+    {
+      "epoch": 8.767259362272773,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.862322439654317e-05,
+      "loss": 0.4733,
+      "step": 176520
+    },
+    {
+      "epoch": 8.767756034568391,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.858349061289361e-05,
+      "loss": 0.4888,
+      "step": 176530
+    },
+    {
+      "epoch": 8.768252706864011,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.854375682924407e-05,
+      "loss": 0.5491,
+      "step": 176540
+    },
+    {
+      "epoch": 8.76874937915963,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 9.850402304559452e-05,
+      "loss": 0.48,
+      "step": 176550
+    },
+    {
+      "epoch": 8.76924605145525,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.846428926194497e-05,
+      "loss": 0.4741,
+      "step": 176560
+    },
+    {
+      "epoch": 8.76974272375087,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.842455547829542e-05,
+      "loss": 0.4764,
+      "step": 176570
+    },
+    {
+      "epoch": 8.770239396046488,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.838482169464588e-05,
+      "loss": 0.51,
+      "step": 176580
+    },
+    {
+      "epoch": 8.770736068342108,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.834508791099632e-05,
+      "loss": 0.466,
+      "step": 176590
+    },
+    {
+      "epoch": 8.771232740637727,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 9.830535412734678e-05,
+      "loss": 0.4746,
+      "step": 176600
+    },
+    {
+      "epoch": 8.771729412933347,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 9.826562034369722e-05,
+      "loss": 0.5056,
+      "step": 176610
+    },
+    {
+      "epoch": 8.772226085228965,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.822588656004768e-05,
+      "loss": 0.4754,
+      "step": 176620
+    },
+    {
+      "epoch": 8.772722757524585,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 9.818615277639814e-05,
+      "loss": 0.4696,
+      "step": 176630
+    },
+    {
+      "epoch": 8.773219429820205,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.81464189927486e-05,
+      "loss": 0.5043,
+      "step": 176640
+    },
+    {
+      "epoch": 8.773716102115824,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 9.810668520909904e-05,
+      "loss": 0.4881,
+      "step": 176650
+    },
+    {
+      "epoch": 8.774212774411444,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.80669514254495e-05,
+      "loss": 0.4989,
+      "step": 176660
+    },
+    {
+      "epoch": 8.774709446707062,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.802721764179994e-05,
+      "loss": 0.472,
+      "step": 176670
+    },
+    {
+      "epoch": 8.775206119002682,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.79874838581504e-05,
+      "loss": 0.4874,
+      "step": 176680
+    },
+    {
+      "epoch": 8.7757027912983,
+      "grad_norm": 0.1640625,
+      "learning_rate": 9.794775007450084e-05,
+      "loss": 0.4736,
+      "step": 176690
+    },
+    {
+      "epoch": 8.77619946359392,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.79080162908513e-05,
+      "loss": 0.481,
+      "step": 176700
+    },
+    {
+      "epoch": 8.77669613588954,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.786828250720176e-05,
+      "loss": 0.4719,
+      "step": 176710
+    },
+    {
+      "epoch": 8.77719280818516,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 9.78285487235522e-05,
+      "loss": 0.481,
+      "step": 176720
+    },
+    {
+      "epoch": 8.77768948048078,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.778881493990266e-05,
+      "loss": 0.4933,
+      "step": 176730
+    },
+    {
+      "epoch": 8.778186152776398,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.77490811562531e-05,
+      "loss": 0.4894,
+      "step": 176740
+    },
+    {
+      "epoch": 8.778682825072018,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.770934737260356e-05,
+      "loss": 0.4862,
+      "step": 176750
+    },
+    {
+      "epoch": 8.779179497367636,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.766961358895401e-05,
+      "loss": 0.4877,
+      "step": 176760
+    },
+    {
+      "epoch": 8.779676169663256,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 9.762987980530447e-05,
+      "loss": 0.5012,
+      "step": 176770
+    },
+    {
+      "epoch": 8.780172841958876,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.759014602165491e-05,
+      "loss": 0.4863,
+      "step": 176780
+    },
+    {
+      "epoch": 8.780669514254495,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.755041223800537e-05,
+      "loss": 0.4905,
+      "step": 176790
+    },
+    {
+      "epoch": 8.781166186550115,
+      "grad_norm": 0.1796875,
+      "learning_rate": 9.751067845435581e-05,
+      "loss": 0.5004,
+      "step": 176800
+    },
+    {
+      "epoch": 8.781662858845733,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.747094467070627e-05,
+      "loss": 0.469,
+      "step": 176810
+    },
+    {
+      "epoch": 8.782159531141353,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.743121088705673e-05,
+      "loss": 0.5133,
+      "step": 176820
+    },
+    {
+      "epoch": 8.782656203436972,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.739147710340719e-05,
+      "loss": 0.4958,
+      "step": 176830
+    },
+    {
+      "epoch": 8.783152875732592,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.735174331975763e-05,
+      "loss": 0.4747,
+      "step": 176840
+    },
+    {
+      "epoch": 8.783649548028212,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 9.731200953610809e-05,
+      "loss": 0.4967,
+      "step": 176850
+    },
+    {
+      "epoch": 8.78414622032383,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.727227575245853e-05,
+      "loss": 0.5081,
+      "step": 176860
+    },
+    {
+      "epoch": 8.78464289261945,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.723254196880899e-05,
+      "loss": 0.4796,
+      "step": 176870
+    },
+    {
+      "epoch": 8.785139564915069,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.719280818515943e-05,
+      "loss": 0.4937,
+      "step": 176880
+    },
+    {
+      "epoch": 8.785636237210689,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.715307440150989e-05,
+      "loss": 0.4887,
+      "step": 176890
+    },
+    {
+      "epoch": 8.786132909506307,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 9.711334061786034e-05,
+      "loss": 0.4763,
+      "step": 176900
+    },
+    {
+      "epoch": 8.786629581801927,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.70736068342108e-05,
+      "loss": 0.475,
+      "step": 176910
+    },
+    {
+      "epoch": 8.787126254097547,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.703387305056124e-05,
+      "loss": 0.4647,
+      "step": 176920
+    },
+    {
+      "epoch": 8.787622926393166,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 9.69941392669117e-05,
+      "loss": 0.4669,
+      "step": 176930
+    },
+    {
+      "epoch": 8.788119598688786,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.695440548326214e-05,
+      "loss": 0.5126,
+      "step": 176940
+    },
+    {
+      "epoch": 8.788616270984404,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.69146716996126e-05,
+      "loss": 0.473,
+      "step": 176950
+    },
+    {
+      "epoch": 8.789112943280024,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.687493791596304e-05,
+      "loss": 0.4998,
+      "step": 176960
+    },
+    {
+      "epoch": 8.789609615575642,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.68352041323135e-05,
+      "loss": 0.4981,
+      "step": 176970
+    },
+    {
+      "epoch": 8.790106287871263,
+      "grad_norm": 0.203125,
+      "learning_rate": 9.679547034866396e-05,
+      "loss": 0.4983,
+      "step": 176980
+    },
+    {
+      "epoch": 8.790602960166883,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.675573656501442e-05,
+      "loss": 0.4855,
+      "step": 176990
+    },
+    {
+      "epoch": 8.791099632462501,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.671600278136486e-05,
+      "loss": 0.5087,
+      "step": 177000
+    },
+    {
+      "epoch": 8.791596304758121,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.667626899771532e-05,
+      "loss": 0.4864,
+      "step": 177010
+    },
+    {
+      "epoch": 8.79209297705374,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 9.663653521406576e-05,
+      "loss": 0.5112,
+      "step": 177020
+    },
+    {
+      "epoch": 8.79258964934936,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.659680143041622e-05,
+      "loss": 0.4765,
+      "step": 177030
+    },
+    {
+      "epoch": 8.793086321644978,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.655706764676666e-05,
+      "loss": 0.4844,
+      "step": 177040
+    },
+    {
+      "epoch": 8.793582993940598,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.651733386311712e-05,
+      "loss": 0.4911,
+      "step": 177050
+    },
+    {
+      "epoch": 8.794079666236218,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.647760007946757e-05,
+      "loss": 0.499,
+      "step": 177060
+    },
+    {
+      "epoch": 8.794576338531837,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.643786629581802e-05,
+      "loss": 0.4567,
+      "step": 177070
+    },
+    {
+      "epoch": 8.795073010827457,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.639813251216847e-05,
+      "loss": 0.4706,
+      "step": 177080
+    },
+    {
+      "epoch": 8.795569683123075,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.635839872851893e-05,
+      "loss": 0.5288,
+      "step": 177090
+    },
+    {
+      "epoch": 8.796066355418695,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.631866494486937e-05,
+      "loss": 0.503,
+      "step": 177100
+    },
+    {
+      "epoch": 8.796563027714313,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.627893116121983e-05,
+      "loss": 0.514,
+      "step": 177110
+    },
+    {
+      "epoch": 8.797059700009934,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.623919737757027e-05,
+      "loss": 0.5233,
+      "step": 177120
+    },
+    {
+      "epoch": 8.797556372305554,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.619946359392073e-05,
+      "loss": 0.504,
+      "step": 177130
+    },
+    {
+      "epoch": 8.798053044601172,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.615972981027119e-05,
+      "loss": 0.4765,
+      "step": 177140
+    },
+    {
+      "epoch": 8.798549716896792,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.611999602662163e-05,
+      "loss": 0.4937,
+      "step": 177150
+    },
+    {
+      "epoch": 8.79904638919241,
+      "grad_norm": 0.19140625,
+      "learning_rate": 9.608026224297209e-05,
+      "loss": 0.4746,
+      "step": 177160
+    },
+    {
+      "epoch": 8.79954306148803,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.604052845932255e-05,
+      "loss": 0.4774,
+      "step": 177170
+    },
+    {
+      "epoch": 8.800039733783649,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.6000794675673e-05,
+      "loss": 0.4536,
+      "step": 177180
+    },
+    {
+      "epoch": 8.800536406079269,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.596106089202345e-05,
+      "loss": 0.4693,
+      "step": 177190
+    },
+    {
+      "epoch": 8.80103307837489,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.592132710837391e-05,
+      "loss": 0.4815,
+      "step": 177200
+    },
+    {
+      "epoch": 8.801529750670507,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.588159332472435e-05,
+      "loss": 0.4595,
+      "step": 177210
+    },
+    {
+      "epoch": 8.802026422966128,
+      "grad_norm": 0.203125,
+      "learning_rate": 9.584185954107481e-05,
+      "loss": 0.4572,
+      "step": 177220
+    },
+    {
+      "epoch": 8.802523095261746,
+      "grad_norm": 0.16015625,
+      "learning_rate": 9.580212575742525e-05,
+      "loss": 0.5092,
+      "step": 177230
+    },
+    {
+      "epoch": 8.803019767557366,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.576239197377571e-05,
+      "loss": 0.4826,
+      "step": 177240
+    },
+    {
+      "epoch": 8.803516439852984,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.572265819012616e-05,
+      "loss": 0.4733,
+      "step": 177250
+    },
+    {
+      "epoch": 8.804013112148604,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.568292440647661e-05,
+      "loss": 0.4817,
+      "step": 177260
+    },
+    {
+      "epoch": 8.804509784444225,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.564319062282706e-05,
+      "loss": 0.5005,
+      "step": 177270
+    },
+    {
+      "epoch": 8.805006456739843,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.560345683917752e-05,
+      "loss": 0.4755,
+      "step": 177280
+    },
+    {
+      "epoch": 8.805503129035463,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.556372305552796e-05,
+      "loss": 0.486,
+      "step": 177290
+    },
+    {
+      "epoch": 8.805999801331081,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.552398927187842e-05,
+      "loss": 0.4747,
+      "step": 177300
+    },
+    {
+      "epoch": 8.806496473626702,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.548425548822886e-05,
+      "loss": 0.5076,
+      "step": 177310
+    },
+    {
+      "epoch": 8.80699314592232,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.544452170457932e-05,
+      "loss": 0.5059,
+      "step": 177320
+    },
+    {
+      "epoch": 8.80748981821794,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.540478792092978e-05,
+      "loss": 0.4922,
+      "step": 177330
+    },
+    {
+      "epoch": 8.80798649051356,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 9.536505413728024e-05,
+      "loss": 0.4997,
+      "step": 177340
+    },
+    {
+      "epoch": 8.808483162809178,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.532532035363068e-05,
+      "loss": 0.4665,
+      "step": 177350
+    },
+    {
+      "epoch": 8.808979835104799,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.528558656998114e-05,
+      "loss": 0.4884,
+      "step": 177360
+    },
+    {
+      "epoch": 8.809476507400417,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.524585278633158e-05,
+      "loss": 0.512,
+      "step": 177370
+    },
+    {
+      "epoch": 8.809973179696037,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 9.520611900268204e-05,
+      "loss": 0.5118,
+      "step": 177380
+    },
+    {
+      "epoch": 8.810469851991655,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.516638521903248e-05,
+      "loss": 0.4607,
+      "step": 177390
+    },
+    {
+      "epoch": 8.810966524287275,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.512665143538294e-05,
+      "loss": 0.5227,
+      "step": 177400
+    },
+    {
+      "epoch": 8.811463196582894,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 9.508691765173339e-05,
+      "loss": 0.4998,
+      "step": 177410
+    },
+    {
+      "epoch": 8.811959868878514,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.504718386808384e-05,
+      "loss": 0.4814,
+      "step": 177420
+    },
+    {
+      "epoch": 8.812456541174134,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.500745008443429e-05,
+      "loss": 0.4896,
+      "step": 177430
+    },
+    {
+      "epoch": 8.812953213469752,
+      "grad_norm": 0.1796875,
+      "learning_rate": 9.496771630078475e-05,
+      "loss": 0.4792,
+      "step": 177440
+    },
+    {
+      "epoch": 8.813449885765372,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.492798251713519e-05,
+      "loss": 0.4921,
+      "step": 177450
+    },
+    {
+      "epoch": 8.81394655806099,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.488824873348565e-05,
+      "loss": 0.5142,
+      "step": 177460
+    },
+    {
+      "epoch": 8.814443230356611,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.484851494983609e-05,
+      "loss": 0.4699,
+      "step": 177470
+    },
+    {
+      "epoch": 8.81493990265223,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.480878116618655e-05,
+      "loss": 0.4891,
+      "step": 177480
+    },
+    {
+      "epoch": 8.81543657494785,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.476904738253701e-05,
+      "loss": 0.4795,
+      "step": 177490
+    },
+    {
+      "epoch": 8.81593324724347,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.472931359888745e-05,
+      "loss": 0.4878,
+      "step": 177500
+    },
+    {
+      "epoch": 8.816429919539088,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 9.468957981523791e-05,
+      "loss": 0.4549,
+      "step": 177510
+    },
+    {
+      "epoch": 8.816926591834708,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.464984603158837e-05,
+      "loss": 0.5013,
+      "step": 177520
+    },
+    {
+      "epoch": 8.817423264130326,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 9.461011224793881e-05,
+      "loss": 0.4684,
+      "step": 177530
+    },
+    {
+      "epoch": 8.817919936425946,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.457037846428927e-05,
+      "loss": 0.4496,
+      "step": 177540
+    },
+    {
+      "epoch": 8.818416608721565,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.453064468063973e-05,
+      "loss": 0.4796,
+      "step": 177550
+    },
+    {
+      "epoch": 8.818913281017185,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.449091089699017e-05,
+      "loss": 0.4595,
+      "step": 177560
+    },
+    {
+      "epoch": 8.819409953312805,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.445117711334063e-05,
+      "loss": 0.4822,
+      "step": 177570
+    },
+    {
+      "epoch": 8.819906625608423,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.441144332969107e-05,
+      "loss": 0.4814,
+      "step": 177580
+    },
+    {
+      "epoch": 8.820403297904043,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.437170954604153e-05,
+      "loss": 0.4958,
+      "step": 177590
+    },
+    {
+      "epoch": 8.820899970199662,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 9.433197576239198e-05,
+      "loss": 0.4911,
+      "step": 177600
+    },
+    {
+      "epoch": 8.821396642495282,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.429224197874243e-05,
+      "loss": 0.4948,
+      "step": 177610
+    },
+    {
+      "epoch": 8.8218933147909,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 9.425250819509288e-05,
+      "loss": 0.5156,
+      "step": 177620
+    },
+    {
+      "epoch": 8.82238998708652,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.421277441144334e-05,
+      "loss": 0.4728,
+      "step": 177630
+    },
+    {
+      "epoch": 8.82288665938214,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.417304062779378e-05,
+      "loss": 0.4963,
+      "step": 177640
+    },
+    {
+      "epoch": 8.823383331677759,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.413330684414424e-05,
+      "loss": 0.4614,
+      "step": 177650
+    },
+    {
+      "epoch": 8.823880003973379,
+      "grad_norm": 0.20703125,
+      "learning_rate": 9.409357306049468e-05,
+      "loss": 0.4823,
+      "step": 177660
+    },
+    {
+      "epoch": 8.824376676268997,
+      "grad_norm": 0.1796875,
+      "learning_rate": 9.405383927684514e-05,
+      "loss": 0.4824,
+      "step": 177670
+    },
+    {
+      "epoch": 8.824873348564617,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.40141054931956e-05,
+      "loss": 0.4904,
+      "step": 177680
+    },
+    {
+      "epoch": 8.825370020860236,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.397437170954606e-05,
+      "loss": 0.5075,
+      "step": 177690
+    },
+    {
+      "epoch": 8.825866693155856,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.39346379258965e-05,
+      "loss": 0.5011,
+      "step": 177700
+    },
+    {
+      "epoch": 8.826363365451476,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 9.389490414224696e-05,
+      "loss": 0.5239,
+      "step": 177710
+    },
+    {
+      "epoch": 8.826860037747094,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.38551703585974e-05,
+      "loss": 0.4759,
+      "step": 177720
+    },
+    {
+      "epoch": 8.827356710042714,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.381543657494786e-05,
+      "loss": 0.5076,
+      "step": 177730
+    },
+    {
+      "epoch": 8.827853382338333,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.37757027912983e-05,
+      "loss": 0.44,
+      "step": 177740
+    },
+    {
+      "epoch": 8.828350054633953,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 9.373596900764876e-05,
+      "loss": 0.5002,
+      "step": 177750
+    },
+    {
+      "epoch": 8.828846726929571,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.36962352239992e-05,
+      "loss": 0.4905,
+      "step": 177760
+    },
+    {
+      "epoch": 8.829343399225191,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 9.365650144034966e-05,
+      "loss": 0.4814,
+      "step": 177770
+    },
+    {
+      "epoch": 8.82984007152081,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.361676765670011e-05,
+      "loss": 0.4869,
+      "step": 177780
+    },
+    {
+      "epoch": 8.83033674381643,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.357703387305057e-05,
+      "loss": 0.4872,
+      "step": 177790
+    },
+    {
+      "epoch": 8.83083341611205,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.353730008940101e-05,
+      "loss": 0.4742,
+      "step": 177800
+    },
+    {
+      "epoch": 8.831330088407668,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.349756630575147e-05,
+      "loss": 0.4995,
+      "step": 177810
+    },
+    {
+      "epoch": 8.831826760703288,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.345783252210191e-05,
+      "loss": 0.501,
+      "step": 177820
+    },
+    {
+      "epoch": 8.832323432998907,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.341809873845237e-05,
+      "loss": 0.4939,
+      "step": 177830
+    },
+    {
+      "epoch": 8.832820105294527,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.337836495480283e-05,
+      "loss": 0.514,
+      "step": 177840
+    },
+    {
+      "epoch": 8.833316777590145,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 9.333863117115329e-05,
+      "loss": 0.4908,
+      "step": 177850
+    },
+    {
+      "epoch": 8.833813449885765,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 9.329889738750373e-05,
+      "loss": 0.4792,
+      "step": 177860
+    },
+    {
+      "epoch": 8.834310122181385,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 9.325916360385419e-05,
+      "loss": 0.4731,
+      "step": 177870
+    },
+    {
+      "epoch": 8.834806794477004,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.321942982020463e-05,
+      "loss": 0.4529,
+      "step": 177880
+    },
+    {
+      "epoch": 8.835303466772624,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 9.317969603655509e-05,
+      "loss": 0.4739,
+      "step": 177890
+    },
+    {
+      "epoch": 8.835800139068242,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.313996225290553e-05,
+      "loss": 0.4991,
+      "step": 177900
+    },
+    {
+      "epoch": 8.836296811363862,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.310022846925599e-05,
+      "loss": 0.4774,
+      "step": 177910
+    },
+    {
+      "epoch": 8.83679348365948,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 9.306049468560644e-05,
+      "loss": 0.5063,
+      "step": 177920
+    },
+    {
+      "epoch": 8.8372901559551,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.30207609019569e-05,
+      "loss": 0.5064,
+      "step": 177930
+    },
+    {
+      "epoch": 8.83778682825072,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.298102711830734e-05,
+      "loss": 0.5055,
+      "step": 177940
+    },
+    {
+      "epoch": 8.83828350054634,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.29412933346578e-05,
+      "loss": 0.5141,
+      "step": 177950
+    },
+    {
+      "epoch": 8.83878017284196,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.290155955100824e-05,
+      "loss": 0.4801,
+      "step": 177960
+    },
+    {
+      "epoch": 8.839276845137578,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.28618257673587e-05,
+      "loss": 0.4867,
+      "step": 177970
+    },
+    {
+      "epoch": 8.839773517433198,
+      "grad_norm": 0.19140625,
+      "learning_rate": 9.282209198370916e-05,
+      "loss": 0.502,
+      "step": 177980
+    },
+    {
+      "epoch": 8.840270189728816,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.27823582000596e-05,
+      "loss": 0.4746,
+      "step": 177990
+    },
+    {
+      "epoch": 8.840766862024436,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.274262441641006e-05,
+      "loss": 0.4901,
+      "step": 178000
+    },
+    {
+      "epoch": 8.841263534320056,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.27028906327605e-05,
+      "loss": 0.4745,
+      "step": 178010
+    },
+    {
+      "epoch": 8.841760206615675,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.266315684911096e-05,
+      "loss": 0.4777,
+      "step": 178020
+    },
+    {
+      "epoch": 8.842256878911295,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.262342306546142e-05,
+      "loss": 0.4853,
+      "step": 178030
+    },
+    {
+      "epoch": 8.842753551206913,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.258368928181188e-05,
+      "loss": 0.4787,
+      "step": 178040
+    },
+    {
+      "epoch": 8.843250223502533,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.254395549816232e-05,
+      "loss": 0.4891,
+      "step": 178050
+    },
+    {
+      "epoch": 8.843746895798152,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.250422171451278e-05,
+      "loss": 0.508,
+      "step": 178060
+    },
+    {
+      "epoch": 8.844243568093772,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.246448793086322e-05,
+      "loss": 0.4948,
+      "step": 178070
+    },
+    {
+      "epoch": 8.844740240389392,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.242475414721368e-05,
+      "loss": 0.4857,
+      "step": 178080
+    },
+    {
+      "epoch": 8.84523691268501,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.238502036356412e-05,
+      "loss": 0.4985,
+      "step": 178090
+    },
+    {
+      "epoch": 8.84573358498063,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.234528657991458e-05,
+      "loss": 0.4708,
+      "step": 178100
+    },
+    {
+      "epoch": 8.846230257276249,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.230555279626503e-05,
+      "loss": 0.4864,
+      "step": 178110
+    },
+    {
+      "epoch": 8.846726929571869,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.226581901261548e-05,
+      "loss": 0.4788,
+      "step": 178120
+    },
+    {
+      "epoch": 8.847223601867487,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.222608522896593e-05,
+      "loss": 0.536,
+      "step": 178130
+    },
+    {
+      "epoch": 8.847720274163107,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 9.218635144531639e-05,
+      "loss": 0.487,
+      "step": 178140
+    },
+    {
+      "epoch": 8.848216946458727,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.214661766166683e-05,
+      "loss": 0.5071,
+      "step": 178150
+    },
+    {
+      "epoch": 8.848713618754346,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.210688387801729e-05,
+      "loss": 0.4701,
+      "step": 178160
+    },
+    {
+      "epoch": 8.849210291049966,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 9.206715009436773e-05,
+      "loss": 0.4669,
+      "step": 178170
+    },
+    {
+      "epoch": 8.849706963345584,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.202741631071819e-05,
+      "loss": 0.4869,
+      "step": 178180
+    },
+    {
+      "epoch": 8.850203635641204,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.198768252706865e-05,
+      "loss": 0.4847,
+      "step": 178190
+    },
+    {
+      "epoch": 8.850700307936822,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.19479487434191e-05,
+      "loss": 0.5305,
+      "step": 178200
+    },
+    {
+      "epoch": 8.851196980232443,
+      "grad_norm": 0.24609375,
+      "learning_rate": 9.190821495976955e-05,
+      "loss": 0.4845,
+      "step": 178210
+    },
+    {
+      "epoch": 8.851693652528063,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 9.186848117612001e-05,
+      "loss": 0.4765,
+      "step": 178220
+    },
+    {
+      "epoch": 8.852190324823681,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.182874739247045e-05,
+      "loss": 0.4727,
+      "step": 178230
+    },
+    {
+      "epoch": 8.852686997119301,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.178901360882091e-05,
+      "loss": 0.4844,
+      "step": 178240
+    },
+    {
+      "epoch": 8.85318366941492,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.174927982517135e-05,
+      "loss": 0.4649,
+      "step": 178250
+    },
+    {
+      "epoch": 8.85368034171054,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.170954604152181e-05,
+      "loss": 0.5175,
+      "step": 178260
+    },
+    {
+      "epoch": 8.854177014006158,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.166981225787226e-05,
+      "loss": 0.5003,
+      "step": 178270
+    },
+    {
+      "epoch": 8.854673686301778,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.163007847422271e-05,
+      "loss": 0.4849,
+      "step": 178280
+    },
+    {
+      "epoch": 8.855170358597398,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.159034469057316e-05,
+      "loss": 0.4885,
+      "step": 178290
+    },
+    {
+      "epoch": 8.855667030893017,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.155061090692362e-05,
+      "loss": 0.4961,
+      "step": 178300
+    },
+    {
+      "epoch": 8.856163703188637,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.151087712327406e-05,
+      "loss": 0.4878,
+      "step": 178310
+    },
+    {
+      "epoch": 8.856660375484255,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.147114333962452e-05,
+      "loss": 0.523,
+      "step": 178320
+    },
+    {
+      "epoch": 8.857157047779875,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.143140955597496e-05,
+      "loss": 0.4829,
+      "step": 178330
+    },
+    {
+      "epoch": 8.857653720075493,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.139167577232542e-05,
+      "loss": 0.4897,
+      "step": 178340
+    },
+    {
+      "epoch": 8.858150392371114,
+      "grad_norm": 0.19140625,
+      "learning_rate": 9.135194198867586e-05,
+      "loss": 0.4924,
+      "step": 178350
+    },
+    {
+      "epoch": 8.858647064666734,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 9.131220820502632e-05,
+      "loss": 0.4894,
+      "step": 178360
+    },
+    {
+      "epoch": 8.859143736962352,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.127247442137678e-05,
+      "loss": 0.4551,
+      "step": 178370
+    },
+    {
+      "epoch": 8.859640409257972,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 9.123274063772724e-05,
+      "loss": 0.4811,
+      "step": 178380
+    },
+    {
+      "epoch": 8.86013708155359,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.11930068540777e-05,
+      "loss": 0.4676,
+      "step": 178390
+    },
+    {
+      "epoch": 8.86063375384921,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.115327307042814e-05,
+      "loss": 0.4773,
+      "step": 178400
+    },
+    {
+      "epoch": 8.861130426144829,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.11135392867786e-05,
+      "loss": 0.4545,
+      "step": 178410
+    },
+    {
+      "epoch": 8.861627098440449,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.107380550312904e-05,
+      "loss": 0.4835,
+      "step": 178420
+    },
+    {
+      "epoch": 8.86212377073607,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.10340717194795e-05,
+      "loss": 0.466,
+      "step": 178430
+    },
+    {
+      "epoch": 8.862620443031688,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.099433793582994e-05,
+      "loss": 0.4657,
+      "step": 178440
+    },
+    {
+      "epoch": 8.863117115327308,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.09546041521804e-05,
+      "loss": 0.5077,
+      "step": 178450
+    },
+    {
+      "epoch": 8.863613787622926,
+      "grad_norm": 0.251953125,
+      "learning_rate": 9.091487036853085e-05,
+      "loss": 0.4968,
+      "step": 178460
+    },
+    {
+      "epoch": 8.864110459918546,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.08751365848813e-05,
+      "loss": 0.508,
+      "step": 178470
+    },
+    {
+      "epoch": 8.864607132214164,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.083540280123175e-05,
+      "loss": 0.4876,
+      "step": 178480
+    },
+    {
+      "epoch": 8.865103804509785,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.07956690175822e-05,
+      "loss": 0.4909,
+      "step": 178490
+    },
+    {
+      "epoch": 8.865600476805405,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 9.075593523393265e-05,
+      "loss": 0.4989,
+      "step": 178500
+    },
+    {
+      "epoch": 8.866097149101023,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.071620145028311e-05,
+      "loss": 0.5079,
+      "step": 178510
+    },
+    {
+      "epoch": 8.866593821396643,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.067646766663355e-05,
+      "loss": 0.4881,
+      "step": 178520
+    },
+    {
+      "epoch": 8.867090493692261,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.063673388298401e-05,
+      "loss": 0.5281,
+      "step": 178530
+    },
+    {
+      "epoch": 8.867587165987882,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.059700009933447e-05,
+      "loss": 0.4902,
+      "step": 178540
+    },
+    {
+      "epoch": 8.8680838382835,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.055726631568493e-05,
+      "loss": 0.5051,
+      "step": 178550
+    },
+    {
+      "epoch": 8.86858051057912,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.051753253203537e-05,
+      "loss": 0.5191,
+      "step": 178560
+    },
+    {
+      "epoch": 8.86907718287474,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.047779874838583e-05,
+      "loss": 0.5161,
+      "step": 178570
+    },
+    {
+      "epoch": 8.869573855170358,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 9.043806496473627e-05,
+      "loss": 0.4767,
+      "step": 178580
+    },
+    {
+      "epoch": 8.870070527465979,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 9.039833118108673e-05,
+      "loss": 0.5076,
+      "step": 178590
+    },
+    {
+      "epoch": 8.870567199761597,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.035859739743717e-05,
+      "loss": 0.4949,
+      "step": 178600
+    },
+    {
+      "epoch": 8.871063872057217,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.031886361378763e-05,
+      "loss": 0.5077,
+      "step": 178610
+    },
+    {
+      "epoch": 8.871560544352835,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.027912983013808e-05,
+      "loss": 0.4732,
+      "step": 178620
+    },
+    {
+      "epoch": 8.872057216648455,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.023939604648853e-05,
+      "loss": 0.4695,
+      "step": 178630
+    },
+    {
+      "epoch": 8.872553888944076,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.019966226283898e-05,
+      "loss": 0.4374,
+      "step": 178640
+    },
+    {
+      "epoch": 8.873050561239694,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 9.015992847918944e-05,
+      "loss": 0.4909,
+      "step": 178650
+    },
+    {
+      "epoch": 8.873547233535314,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.012019469553988e-05,
+      "loss": 0.4758,
+      "step": 178660
+    },
+    {
+      "epoch": 8.874043905830932,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.008046091189034e-05,
+      "loss": 0.4834,
+      "step": 178670
+    },
+    {
+      "epoch": 8.874540578126553,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.004072712824078e-05,
+      "loss": 0.5022,
+      "step": 178680
+    },
+    {
+      "epoch": 8.87503725042217,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.000099334459124e-05,
+      "loss": 0.4888,
+      "step": 178690
+    },
+    {
+      "epoch": 8.875533922717791,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 8.996125956094168e-05,
+      "loss": 0.4553,
+      "step": 178700
+    },
+    {
+      "epoch": 8.876030595013411,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 8.992152577729214e-05,
+      "loss": 0.4859,
+      "step": 178710
+    },
+    {
+      "epoch": 8.87652726730903,
+      "grad_norm": 0.216796875,
+      "learning_rate": 8.98817919936426e-05,
+      "loss": 0.5082,
+      "step": 178720
+    },
+    {
+      "epoch": 8.87702393960465,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.984205820999306e-05,
+      "loss": 0.508,
+      "step": 178730
+    },
+    {
+      "epoch": 8.877520611900268,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 8.98023244263435e-05,
+      "loss": 0.4816,
+      "step": 178740
+    },
+    {
+      "epoch": 8.878017284195888,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.976259064269396e-05,
+      "loss": 0.4698,
+      "step": 178750
+    },
+    {
+      "epoch": 8.878513956491506,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.97228568590444e-05,
+      "loss": 0.4767,
+      "step": 178760
+    },
+    {
+      "epoch": 8.879010628787126,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.968312307539486e-05,
+      "loss": 0.4936,
+      "step": 178770
+    },
+    {
+      "epoch": 8.879507301082745,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 8.96433892917453e-05,
+      "loss": 0.4932,
+      "step": 178780
+    },
+    {
+      "epoch": 8.880003973378365,
+      "grad_norm": 0.216796875,
+      "learning_rate": 8.960365550809576e-05,
+      "loss": 0.5014,
+      "step": 178790
+    },
+    {
+      "epoch": 8.880500645673985,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.956392172444621e-05,
+      "loss": 0.4737,
+      "step": 178800
+    },
+    {
+      "epoch": 8.880997317969603,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.952418794079667e-05,
+      "loss": 0.4558,
+      "step": 178810
+    },
+    {
+      "epoch": 8.881493990265223,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.948445415714712e-05,
+      "loss": 0.511,
+      "step": 178820
+    },
+    {
+      "epoch": 8.881990662560842,
+      "grad_norm": 0.2265625,
+      "learning_rate": 8.944472037349757e-05,
+      "loss": 0.4751,
+      "step": 178830
+    },
+    {
+      "epoch": 8.882487334856462,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.940498658984803e-05,
+      "loss": 0.4934,
+      "step": 178840
+    },
+    {
+      "epoch": 8.88298400715208,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 8.936525280619847e-05,
+      "loss": 0.5163,
+      "step": 178850
+    },
+    {
+      "epoch": 8.8834806794477,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.932551902254893e-05,
+      "loss": 0.5046,
+      "step": 178860
+    },
+    {
+      "epoch": 8.88397735174332,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 8.928578523889937e-05,
+      "loss": 0.4613,
+      "step": 178870
+    },
+    {
+      "epoch": 8.884474024038939,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.924605145524983e-05,
+      "loss": 0.5034,
+      "step": 178880
+    },
+    {
+      "epoch": 8.884970696334559,
+      "grad_norm": 0.19921875,
+      "learning_rate": 8.920631767160029e-05,
+      "loss": 0.5281,
+      "step": 178890
+    },
+    {
+      "epoch": 8.885467368630177,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.916658388795075e-05,
+      "loss": 0.5001,
+      "step": 178900
+    },
+    {
+      "epoch": 8.885964040925797,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.912685010430119e-05,
+      "loss": 0.4436,
+      "step": 178910
+    },
+    {
+      "epoch": 8.886460713221416,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 8.908711632065165e-05,
+      "loss": 0.4834,
+      "step": 178920
+    },
+    {
+      "epoch": 8.886957385517036,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 8.904738253700209e-05,
+      "loss": 0.4922,
+      "step": 178930
+    },
+    {
+      "epoch": 8.887454057812656,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.900764875335255e-05,
+      "loss": 0.4558,
+      "step": 178940
+    },
+    {
+      "epoch": 8.887950730108274,
+      "grad_norm": 0.19921875,
+      "learning_rate": 8.8967914969703e-05,
+      "loss": 0.5002,
+      "step": 178950
+    },
+    {
+      "epoch": 8.888447402403894,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.892818118605345e-05,
+      "loss": 0.4759,
+      "step": 178960
+    },
+    {
+      "epoch": 8.888944074699513,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 8.88884474024039e-05,
+      "loss": 0.4968,
+      "step": 178970
+    },
+    {
+      "epoch": 8.889440746995133,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 8.884871361875435e-05,
+      "loss": 0.4812,
+      "step": 178980
+    },
+    {
+      "epoch": 8.889937419290751,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.88089798351048e-05,
+      "loss": 0.4873,
+      "step": 178990
+    },
+    {
+      "epoch": 8.890434091586371,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.876924605145526e-05,
+      "loss": 0.4781,
+      "step": 179000
+    },
+    {
+      "epoch": 8.890930763881991,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 8.87295122678057e-05,
+      "loss": 0.4674,
+      "step": 179010
+    },
+    {
+      "epoch": 8.89142743617761,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 8.868977848415616e-05,
+      "loss": 0.4771,
+      "step": 179020
+    },
+    {
+      "epoch": 8.89192410847323,
+      "grad_norm": 0.1953125,
+      "learning_rate": 8.86500447005066e-05,
+      "loss": 0.4913,
+      "step": 179030
+    },
+    {
+      "epoch": 8.892420780768848,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 8.861031091685706e-05,
+      "loss": 0.5114,
+      "step": 179040
+    },
+    {
+      "epoch": 8.892917453064468,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 8.857057713320752e-05,
+      "loss": 0.4954,
+      "step": 179050
+    },
+    {
+      "epoch": 8.893414125360087,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.853084334955796e-05,
+      "loss": 0.5121,
+      "step": 179060
+    },
+    {
+      "epoch": 8.893910797655707,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.849110956590842e-05,
+      "loss": 0.5159,
+      "step": 179070
+    },
+    {
+      "epoch": 8.894407469951327,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.845137578225888e-05,
+      "loss": 0.4863,
+      "step": 179080
+    },
+    {
+      "epoch": 8.894904142246945,
+      "grad_norm": 0.1875,
+      "learning_rate": 8.841164199860932e-05,
+      "loss": 0.4863,
+      "step": 179090
+    },
+    {
+      "epoch": 8.895400814542565,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.837190821495978e-05,
+      "loss": 0.516,
+      "step": 179100
+    },
+    {
+      "epoch": 8.895897486838184,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.833217443131022e-05,
+      "loss": 0.5159,
+      "step": 179110
+    },
+    {
+      "epoch": 8.896394159133804,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.829244064766068e-05,
+      "loss": 0.4884,
+      "step": 179120
+    },
+    {
+      "epoch": 8.896890831429422,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.825270686401113e-05,
+      "loss": 0.5105,
+      "step": 179130
+    },
+    {
+      "epoch": 8.897387503725042,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.821297308036158e-05,
+      "loss": 0.5287,
+      "step": 179140
+    },
+    {
+      "epoch": 8.897884176020662,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.817323929671203e-05,
+      "loss": 0.4891,
+      "step": 179150
+    },
+    {
+      "epoch": 8.89838084831628,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.813350551306249e-05,
+      "loss": 0.488,
+      "step": 179160
+    },
+    {
+      "epoch": 8.8988775206119,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 8.809377172941293e-05,
+      "loss": 0.4921,
+      "step": 179170
+    },
+    {
+      "epoch": 8.89937419290752,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 8.805403794576339e-05,
+      "loss": 0.5335,
+      "step": 179180
+    },
+    {
+      "epoch": 8.89987086520314,
+      "grad_norm": 0.1796875,
+      "learning_rate": 8.801430416211383e-05,
+      "loss": 0.5144,
+      "step": 179190
+    },
+    {
+      "epoch": 8.900367537498758,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.797457037846429e-05,
+      "loss": 0.5105,
+      "step": 179200
+    },
+    {
+      "epoch": 8.900864209794378,
+      "grad_norm": 0.189453125,
+      "learning_rate": 8.793483659481473e-05,
+      "loss": 0.4774,
+      "step": 179210
+    },
+    {
+      "epoch": 8.901360882089996,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 8.789510281116519e-05,
+      "loss": 0.5074,
+      "step": 179220
+    },
+    {
+      "epoch": 8.901857554385616,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.785536902751565e-05,
+      "loss": 0.4918,
+      "step": 179230
+    },
+    {
+      "epoch": 8.902354226681236,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.781563524386611e-05,
+      "loss": 0.4707,
+      "step": 179240
+    },
+    {
+      "epoch": 8.902850898976855,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 8.777590146021657e-05,
+      "loss": 0.4949,
+      "step": 179250
+    },
+    {
+      "epoch": 8.903347571272475,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 8.773616767656701e-05,
+      "loss": 0.4835,
+      "step": 179260
+    },
+    {
+      "epoch": 8.903844243568093,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 8.769643389291747e-05,
+      "loss": 0.4933,
+      "step": 179270
+    },
+    {
+      "epoch": 8.904340915863713,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.765670010926791e-05,
+      "loss": 0.4601,
+      "step": 179280
+    },
+    {
+      "epoch": 8.904837588159332,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.761696632561837e-05,
+      "loss": 0.4871,
+      "step": 179290
+    },
+    {
+      "epoch": 8.905334260454952,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.757723254196881e-05,
+      "loss": 0.4939,
+      "step": 179300
+    },
+    {
+      "epoch": 8.905830932750572,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 8.753749875831927e-05,
+      "loss": 0.4796,
+      "step": 179310
+    },
+    {
+      "epoch": 8.90632760504619,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 8.749776497466972e-05,
+      "loss": 0.4798,
+      "step": 179320
+    },
+    {
+      "epoch": 8.90682427734181,
+      "grad_norm": 0.20703125,
+      "learning_rate": 8.745803119102017e-05,
+      "loss": 0.4824,
+      "step": 179330
+    },
+    {
+      "epoch": 8.907320949637429,
+      "grad_norm": 0.166015625,
+      "learning_rate": 8.741829740737062e-05,
+      "loss": 0.5187,
+      "step": 179340
+    },
+    {
+      "epoch": 8.907817621933049,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.737856362372108e-05,
+      "loss": 0.5074,
+      "step": 179350
+    },
+    {
+      "epoch": 8.908314294228667,
+      "grad_norm": 0.18359375,
+      "learning_rate": 8.733882984007152e-05,
+      "loss": 0.4614,
+      "step": 179360
+    },
+    {
+      "epoch": 8.908810966524287,
+      "grad_norm": 0.1796875,
+      "learning_rate": 8.729909605642198e-05,
+      "loss": 0.4654,
+      "step": 179370
+    },
+    {
+      "epoch": 8.909307638819907,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 8.725936227277242e-05,
+      "loss": 0.4884,
+      "step": 179380
+    },
+    {
+      "epoch": 8.909804311115526,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 8.721962848912288e-05,
+      "loss": 0.5112,
+      "step": 179390
+    },
+    {
+      "epoch": 8.910300983411146,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 8.717989470547334e-05,
+      "loss": 0.4858,
+      "step": 179400
+    },
+    {
+      "epoch": 8.910797655706764,
+      "grad_norm": 0.189453125,
+      "learning_rate": 8.714016092182378e-05,
+      "loss": 0.5039,
+      "step": 179410
+    },
+    {
+      "epoch": 8.911294328002384,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.710042713817424e-05,
+      "loss": 0.4878,
+      "step": 179420
+    },
+    {
+      "epoch": 8.911791000298003,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 8.70606933545247e-05,
+      "loss": 0.4789,
+      "step": 179430
+    },
+    {
+      "epoch": 8.912287672593623,
+      "grad_norm": 0.18359375,
+      "learning_rate": 8.702095957087514e-05,
+      "loss": 0.4978,
+      "step": 179440
+    },
+    {
+      "epoch": 8.912784344889243,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 8.69812257872256e-05,
+      "loss": 0.4867,
+      "step": 179450
+    },
+    {
+      "epoch": 8.913281017184861,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 8.694149200357604e-05,
+      "loss": 0.4864,
+      "step": 179460
+    },
+    {
+      "epoch": 8.913777689480481,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.69017582199265e-05,
+      "loss": 0.489,
+      "step": 179470
+    },
+    {
+      "epoch": 8.9142743617761,
+      "grad_norm": 0.19921875,
+      "learning_rate": 8.686202443627695e-05,
+      "loss": 0.487,
+      "step": 179480
+    },
+    {
+      "epoch": 8.91477103407172,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 8.68222906526274e-05,
+      "loss": 0.4775,
+      "step": 179490
+    },
+    {
+      "epoch": 8.915267706367338,
+      "grad_norm": 0.197265625,
+      "learning_rate": 8.678255686897785e-05,
+      "loss": 0.4653,
+      "step": 179500
+    },
+    {
+      "epoch": 8.915764378662958,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.67428230853283e-05,
+      "loss": 0.4398,
+      "step": 179510
+    },
+    {
+      "epoch": 8.916261050958578,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.670308930167875e-05,
+      "loss": 0.507,
+      "step": 179520
+    },
+    {
+      "epoch": 8.916757723254197,
+      "grad_norm": 0.189453125,
+      "learning_rate": 8.666335551802921e-05,
+      "loss": 0.4761,
+      "step": 179530
+    },
+    {
+      "epoch": 8.917254395549817,
+      "grad_norm": 0.19921875,
+      "learning_rate": 8.662362173437965e-05,
+      "loss": 0.4468,
+      "step": 179540
+    },
+    {
+      "epoch": 8.917751067845435,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.658388795073011e-05,
+      "loss": 0.4619,
+      "step": 179550
+    },
+    {
+      "epoch": 8.918247740141055,
+      "grad_norm": 0.203125,
+      "learning_rate": 8.654415416708055e-05,
+      "loss": 0.4887,
+      "step": 179560
+    },
+    {
+      "epoch": 8.918744412436673,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 8.650442038343101e-05,
+      "loss": 0.4776,
+      "step": 179570
+    },
+    {
+      "epoch": 8.919241084732294,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 8.646468659978147e-05,
+      "loss": 0.5162,
+      "step": 179580
+    },
+    {
+      "epoch": 8.919737757027914,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.642495281613193e-05,
+      "loss": 0.4993,
+      "step": 179590
+    },
+    {
+      "epoch": 8.920234429323532,
+      "grad_norm": 0.1953125,
+      "learning_rate": 8.638521903248237e-05,
+      "loss": 0.4697,
+      "step": 179600
+    },
+    {
+      "epoch": 8.920731101619152,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 8.634548524883283e-05,
+      "loss": 0.4912,
+      "step": 179610
+    },
+    {
+      "epoch": 8.92122777391477,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.630575146518327e-05,
+      "loss": 0.4815,
+      "step": 179620
+    },
+    {
+      "epoch": 8.92172444621039,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.626601768153373e-05,
+      "loss": 0.4812,
+      "step": 179630
+    },
+    {
+      "epoch": 8.922221118506009,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.622628389788418e-05,
+      "loss": 0.4815,
+      "step": 179640
+    },
+    {
+      "epoch": 8.922717790801629,
+      "grad_norm": 0.1875,
+      "learning_rate": 8.618655011423463e-05,
+      "loss": 0.5083,
+      "step": 179650
+    },
+    {
+      "epoch": 8.92321446309725,
+      "grad_norm": 0.20703125,
+      "learning_rate": 8.614681633058509e-05,
+      "loss": 0.4818,
+      "step": 179660
+    },
+    {
+      "epoch": 8.923711135392868,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.610708254693554e-05,
+      "loss": 0.4998,
+      "step": 179670
+    },
+    {
+      "epoch": 8.924207807688488,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.606734876328599e-05,
+      "loss": 0.48,
+      "step": 179680
+    },
+    {
+      "epoch": 8.924704479984106,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.602761497963644e-05,
+      "loss": 0.503,
+      "step": 179690
+    },
+    {
+      "epoch": 8.925201152279726,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.59878811959869e-05,
+      "loss": 0.4443,
+      "step": 179700
+    },
+    {
+      "epoch": 8.925697824575344,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.594814741233734e-05,
+      "loss": 0.4841,
+      "step": 179710
+    },
+    {
+      "epoch": 8.926194496870965,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.59084136286878e-05,
+      "loss": 0.5048,
+      "step": 179720
+    },
+    {
+      "epoch": 8.926691169166585,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.586867984503824e-05,
+      "loss": 0.5001,
+      "step": 179730
+    },
+    {
+      "epoch": 8.927187841462203,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.58289460613887e-05,
+      "loss": 0.4688,
+      "step": 179740
+    },
+    {
+      "epoch": 8.927684513757823,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 8.578921227773916e-05,
+      "loss": 0.4602,
+      "step": 179750
+    },
+    {
+      "epoch": 8.928181186053441,
+      "grad_norm": 0.236328125,
+      "learning_rate": 8.57494784940896e-05,
+      "loss": 0.505,
+      "step": 179760
+    },
+    {
+      "epoch": 8.928677858349062,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.570974471044006e-05,
+      "loss": 0.4495,
+      "step": 179770
+    },
+    {
+      "epoch": 8.92917453064468,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 8.567001092679052e-05,
+      "loss": 0.4841,
+      "step": 179780
+    },
+    {
+      "epoch": 8.9296712029403,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 8.563027714314096e-05,
+      "loss": 0.4862,
+      "step": 179790
+    },
+    {
+      "epoch": 8.93016787523592,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.559054335949142e-05,
+      "loss": 0.5101,
+      "step": 179800
+    },
+    {
+      "epoch": 8.930664547531538,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 8.555080957584186e-05,
+      "loss": 0.4651,
+      "step": 179810
+    },
+    {
+      "epoch": 8.931161219827159,
+      "grad_norm": 0.20703125,
+      "learning_rate": 8.551107579219232e-05,
+      "loss": 0.4696,
+      "step": 179820
+    },
+    {
+      "epoch": 8.931657892122777,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 8.547134200854277e-05,
+      "loss": 0.5091,
+      "step": 179830
+    },
+    {
+      "epoch": 8.932154564418397,
+      "grad_norm": 0.19140625,
+      "learning_rate": 8.543160822489322e-05,
+      "loss": 0.4927,
+      "step": 179840
+    },
+    {
+      "epoch": 8.932651236714015,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 8.539187444124367e-05,
+      "loss": 0.4911,
+      "step": 179850
+    },
+    {
+      "epoch": 8.933147909009636,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 8.535214065759413e-05,
+      "loss": 0.52,
+      "step": 179860
+    },
+    {
+      "epoch": 8.933644581305256,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 8.531240687394457e-05,
+      "loss": 0.4781,
+      "step": 179870
+    },
+    {
+      "epoch": 8.934141253600874,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.527267309029503e-05,
+      "loss": 0.4909,
+      "step": 179880
+    },
+    {
+      "epoch": 8.934637925896494,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.523293930664547e-05,
+      "loss": 0.4733,
+      "step": 179890
+    },
+    {
+      "epoch": 8.935134598192112,
+      "grad_norm": 0.203125,
+      "learning_rate": 8.519320552299593e-05,
+      "loss": 0.499,
+      "step": 179900
+    },
+    {
+      "epoch": 8.935631270487733,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.515347173934637e-05,
+      "loss": 0.4652,
+      "step": 179910
+    },
+    {
+      "epoch": 8.93612794278335,
+      "grad_norm": 0.1875,
+      "learning_rate": 8.511373795569683e-05,
+      "loss": 0.4774,
+      "step": 179920
+    },
+    {
+      "epoch": 8.936624615078971,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 8.507400417204729e-05,
+      "loss": 0.5377,
+      "step": 179930
+    },
+    {
+      "epoch": 8.937121287374591,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.503427038839775e-05,
+      "loss": 0.4984,
+      "step": 179940
+    },
+    {
+      "epoch": 8.93761795967021,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.499453660474819e-05,
+      "loss": 0.4954,
+      "step": 179950
+    },
+    {
+      "epoch": 8.93811463196583,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.495480282109865e-05,
+      "loss": 0.4871,
+      "step": 179960
+    },
+    {
+      "epoch": 8.938611304261448,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.49150690374491e-05,
+      "loss": 0.4923,
+      "step": 179970
+    },
+    {
+      "epoch": 8.939107976557068,
+      "grad_norm": 0.232421875,
+      "learning_rate": 8.487533525379955e-05,
+      "loss": 0.4952,
+      "step": 179980
+    },
+    {
+      "epoch": 8.939604648852686,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.483560147015e-05,
+      "loss": 0.5052,
+      "step": 179990
+    },
+    {
+      "epoch": 8.940101321148306,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.479586768650045e-05,
+      "loss": 0.4791,
+      "step": 180000
+    },
+    {
+      "epoch": 8.940597993443927,
+      "grad_norm": 0.1875,
+      "learning_rate": 8.47561339028509e-05,
+      "loss": 0.4606,
+      "step": 180010
+    },
+    {
+      "epoch": 8.941094665739545,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.471640011920136e-05,
+      "loss": 0.4786,
+      "step": 180020
+    },
+    {
+      "epoch": 8.941591338035165,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 8.46766663355518e-05,
+      "loss": 0.468,
+      "step": 180030
+    },
+    {
+      "epoch": 8.942088010330783,
+      "grad_norm": 0.1796875,
+      "learning_rate": 8.463693255190226e-05,
+      "loss": 0.4834,
+      "step": 180040
+    },
+    {
+      "epoch": 8.942584682626403,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 8.45971987682527e-05,
+      "loss": 0.4883,
+      "step": 180050
+    },
+    {
+      "epoch": 8.943081354922022,
+      "grad_norm": 0.19140625,
+      "learning_rate": 8.455746498460316e-05,
+      "loss": 0.4961,
+      "step": 180060
+    },
+    {
+      "epoch": 8.943578027217642,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.451773120095362e-05,
+      "loss": 0.4694,
+      "step": 180070
+    },
+    {
+      "epoch": 8.944074699513262,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.447799741730406e-05,
+      "loss": 0.4684,
+      "step": 180080
+    },
+    {
+      "epoch": 8.94457137180888,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 8.443826363365452e-05,
+      "loss": 0.4963,
+      "step": 180090
+    },
+    {
+      "epoch": 8.9450680441045,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 8.439852985000498e-05,
+      "loss": 0.482,
+      "step": 180100
+    },
+    {
+      "epoch": 8.945564716400119,
+      "grad_norm": 0.20703125,
+      "learning_rate": 8.435879606635542e-05,
+      "loss": 0.5205,
+      "step": 180110
+    },
+    {
+      "epoch": 8.946061388695739,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.431906228270588e-05,
+      "loss": 0.4627,
+      "step": 180120
+    },
+    {
+      "epoch": 8.946558060991357,
+      "grad_norm": 0.197265625,
+      "learning_rate": 8.427932849905634e-05,
+      "loss": 0.4897,
+      "step": 180130
+    },
+    {
+      "epoch": 8.947054733286977,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.423959471540678e-05,
+      "loss": 0.4652,
+      "step": 180140
+    },
+    {
+      "epoch": 8.947551405582598,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.419986093175724e-05,
+      "loss": 0.5093,
+      "step": 180150
+    },
+    {
+      "epoch": 8.948048077878216,
+      "grad_norm": 0.18359375,
+      "learning_rate": 8.416012714810768e-05,
+      "loss": 0.4991,
+      "step": 180160
+    },
+    {
+      "epoch": 8.948544750173836,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.412039336445814e-05,
+      "loss": 0.4651,
+      "step": 180170
+    },
+    {
+      "epoch": 8.949041422469454,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.408065958080859e-05,
+      "loss": 0.4559,
+      "step": 180180
+    },
+    {
+      "epoch": 8.949538094765074,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.404092579715904e-05,
+      "loss": 0.5146,
+      "step": 180190
+    },
+    {
+      "epoch": 8.950034767060693,
+      "grad_norm": 0.1953125,
+      "learning_rate": 8.400119201350949e-05,
+      "loss": 0.495,
+      "step": 180200
+    },
+    {
+      "epoch": 8.950531439356313,
+      "grad_norm": 0.19921875,
+      "learning_rate": 8.396145822985994e-05,
+      "loss": 0.4924,
+      "step": 180210
+    },
+    {
+      "epoch": 8.951028111651931,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.392172444621039e-05,
+      "loss": 0.4889,
+      "step": 180220
+    },
+    {
+      "epoch": 8.951524783947551,
+      "grad_norm": 0.18359375,
+      "learning_rate": 8.388199066256085e-05,
+      "loss": 0.4817,
+      "step": 180230
+    },
+    {
+      "epoch": 8.952021456243171,
+      "grad_norm": 0.1796875,
+      "learning_rate": 8.384225687891129e-05,
+      "loss": 0.4716,
+      "step": 180240
+    },
+    {
+      "epoch": 8.95251812853879,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 8.380252309526175e-05,
+      "loss": 0.501,
+      "step": 180250
+    },
+    {
+      "epoch": 8.95301480083441,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 8.37627893116122e-05,
+      "loss": 0.4933,
+      "step": 180260
+    },
+    {
+      "epoch": 8.953511473130028,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.372305552796265e-05,
+      "loss": 0.4752,
+      "step": 180270
+    },
+    {
+      "epoch": 8.954008145425648,
+      "grad_norm": 0.18359375,
+      "learning_rate": 8.368332174431311e-05,
+      "loss": 0.4896,
+      "step": 180280
+    },
+    {
+      "epoch": 8.954504817721267,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8.364358796066357e-05,
+      "loss": 0.4925,
+      "step": 180290
+    },
+    {
+      "epoch": 8.955001490016887,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.360385417701401e-05,
+      "loss": 0.5158,
+      "step": 180300
+    },
+    {
+      "epoch": 8.955498162312507,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 8.356412039336447e-05,
+      "loss": 0.4804,
+      "step": 180310
+    },
+    {
+      "epoch": 8.955994834608125,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 8.352438660971491e-05,
+      "loss": 0.4729,
+      "step": 180320
+    },
+    {
+      "epoch": 8.956491506903745,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.348465282606537e-05,
+      "loss": 0.5196,
+      "step": 180330
+    },
+    {
+      "epoch": 8.956988179199364,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.344491904241582e-05,
+      "loss": 0.4813,
+      "step": 180340
+    },
+    {
+      "epoch": 8.957484851494984,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.340518525876627e-05,
+      "loss": 0.4854,
+      "step": 180350
+    },
+    {
+      "epoch": 8.957981523790602,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.336545147511672e-05,
+      "loss": 0.4981,
+      "step": 180360
+    },
+    {
+      "epoch": 8.958478196086222,
+      "grad_norm": 0.1953125,
+      "learning_rate": 8.332571769146717e-05,
+      "loss": 0.5089,
+      "step": 180370
+    },
+    {
+      "epoch": 8.958974868381842,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 8.328598390781762e-05,
+      "loss": 0.4921,
+      "step": 180380
+    },
+    {
+      "epoch": 8.95947154067746,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.324625012416808e-05,
+      "loss": 0.534,
+      "step": 180390
+    },
+    {
+      "epoch": 8.95996821297308,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 8.320651634051852e-05,
+      "loss": 0.5196,
+      "step": 180400
+    },
+    {
+      "epoch": 8.9604648852687,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.316678255686898e-05,
+      "loss": 0.4891,
+      "step": 180410
+    },
+    {
+      "epoch": 8.96096155756432,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 8.312704877321942e-05,
+      "loss": 0.5117,
+      "step": 180420
+    },
+    {
+      "epoch": 8.961458229859938,
+      "grad_norm": 0.1875,
+      "learning_rate": 8.308731498956988e-05,
+      "loss": 0.4737,
+      "step": 180430
+    },
+    {
+      "epoch": 8.961954902155558,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.304758120592034e-05,
+      "loss": 0.4806,
+      "step": 180440
+    },
+    {
+      "epoch": 8.962451574451178,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.30078474222708e-05,
+      "loss": 0.5053,
+      "step": 180450
+    },
+    {
+      "epoch": 8.962948246746796,
+      "grad_norm": 0.19140625,
+      "learning_rate": 8.296811363862124e-05,
+      "loss": 0.5095,
+      "step": 180460
+    },
+    {
+      "epoch": 8.963444919042416,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.29283798549717e-05,
+      "loss": 0.4786,
+      "step": 180470
+    },
+    {
+      "epoch": 8.963941591338035,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.288864607132214e-05,
+      "loss": 0.5075,
+      "step": 180480
+    },
+    {
+      "epoch": 8.964438263633655,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.28489122876726e-05,
+      "loss": 0.4938,
+      "step": 180490
+    },
+    {
+      "epoch": 8.964934935929273,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 8.280917850402306e-05,
+      "loss": 0.5019,
+      "step": 180500
+    },
+    {
+      "epoch": 8.965431608224893,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.27694447203735e-05,
+      "loss": 0.475,
+      "step": 180510
+    },
+    {
+      "epoch": 8.965928280520513,
+      "grad_norm": 0.19921875,
+      "learning_rate": 8.272971093672396e-05,
+      "loss": 0.4328,
+      "step": 180520
+    },
+    {
+      "epoch": 8.966424952816132,
+      "grad_norm": 0.1875,
+      "learning_rate": 8.26899771530744e-05,
+      "loss": 0.5268,
+      "step": 180530
+    },
+    {
+      "epoch": 8.966921625111752,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.265024336942486e-05,
+      "loss": 0.4776,
+      "step": 180540
+    },
+    {
+      "epoch": 8.96741829740737,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.261050958577531e-05,
+      "loss": 0.4855,
+      "step": 180550
+    },
+    {
+      "epoch": 8.96791496970299,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.257077580212576e-05,
+      "loss": 0.5095,
+      "step": 180560
+    },
+    {
+      "epoch": 8.968411641998609,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 8.253104201847621e-05,
+      "loss": 0.4727,
+      "step": 180570
+    },
+    {
+      "epoch": 8.968908314294229,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.249130823482667e-05,
+      "loss": 0.5032,
+      "step": 180580
+    },
+    {
+      "epoch": 8.969404986589847,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.245157445117711e-05,
+      "loss": 0.4954,
+      "step": 180590
+    },
+    {
+      "epoch": 8.969901658885467,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 8.241184066752757e-05,
+      "loss": 0.4854,
+      "step": 180600
+    },
+    {
+      "epoch": 8.970398331181087,
+      "grad_norm": 0.1953125,
+      "learning_rate": 8.237210688387801e-05,
+      "loss": 0.495,
+      "step": 180610
+    },
+    {
+      "epoch": 8.970895003476706,
+      "grad_norm": 0.203125,
+      "learning_rate": 8.233237310022847e-05,
+      "loss": 0.4903,
+      "step": 180620
+    },
+    {
+      "epoch": 8.971391675772326,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 8.229263931657893e-05,
+      "loss": 0.4724,
+      "step": 180630
+    },
+    {
+      "epoch": 8.971888348067944,
+      "grad_norm": 0.19140625,
+      "learning_rate": 8.225290553292939e-05,
+      "loss": 0.5086,
+      "step": 180640
+    },
+    {
+      "epoch": 8.972385020363564,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 8.221317174927983e-05,
+      "loss": 0.496,
+      "step": 180650
+    },
+    {
+      "epoch": 8.972881692659183,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 8.217343796563029e-05,
+      "loss": 0.4937,
+      "step": 180660
+    },
+    {
+      "epoch": 8.973378364954803,
+      "grad_norm": 0.232421875,
+      "learning_rate": 8.213370418198073e-05,
+      "loss": 0.4648,
+      "step": 180670
+    },
+    {
+      "epoch": 8.973875037250423,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 8.209397039833119e-05,
+      "loss": 0.4768,
+      "step": 180680
+    },
+    {
+      "epoch": 8.974371709546041,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 8.205423661468163e-05,
+      "loss": 0.4783,
+      "step": 180690
+    },
+    {
+      "epoch": 8.974868381841661,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 8.201450283103209e-05,
+      "loss": 0.4717,
+      "step": 180700
+    },
+    {
+      "epoch": 8.97536505413728,
+      "grad_norm": 0.244140625,
+      "learning_rate": 8.197476904738254e-05,
+      "loss": 0.4951,
+      "step": 180710
+    },
+    {
+      "epoch": 8.9758617264329,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.1935035263733e-05,
+      "loss": 0.4896,
+      "step": 180720
+    },
+    {
+      "epoch": 8.976358398728518,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 8.189530148008344e-05,
+      "loss": 0.4844,
+      "step": 180730
+    },
+    {
+      "epoch": 8.976855071024138,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 8.18555676964339e-05,
+      "loss": 0.4644,
+      "step": 180740
+    },
+    {
+      "epoch": 8.977351743319758,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 8.181583391278434e-05,
+      "loss": 0.4718,
+      "step": 180750
+    },
+    {
+      "epoch": 8.977848415615377,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.17761001291348e-05,
+      "loss": 0.4921,
+      "step": 180760
+    },
+    {
+      "epoch": 8.978345087910997,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.173636634548524e-05,
+      "loss": 0.467,
+      "step": 180770
+    },
+    {
+      "epoch": 8.978841760206615,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.16966325618357e-05,
+      "loss": 0.5066,
+      "step": 180780
+    },
+    {
+      "epoch": 8.979338432502235,
+      "grad_norm": 0.18359375,
+      "learning_rate": 8.165689877818616e-05,
+      "loss": 0.5071,
+      "step": 180790
+    },
+    {
+      "epoch": 8.979835104797854,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 8.161716499453662e-05,
+      "loss": 0.4895,
+      "step": 180800
+    },
+    {
+      "epoch": 8.980331777093474,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.157743121088706e-05,
+      "loss": 0.5055,
+      "step": 180810
+    },
+    {
+      "epoch": 8.980828449389094,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.153769742723752e-05,
+      "loss": 0.4669,
+      "step": 180820
+    },
+    {
+      "epoch": 8.981325121684712,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 8.149796364358796e-05,
+      "loss": 0.5243,
+      "step": 180830
+    },
+    {
+      "epoch": 8.981821793980332,
+      "grad_norm": 0.189453125,
+      "learning_rate": 8.145822985993842e-05,
+      "loss": 0.4916,
+      "step": 180840
+    },
+    {
+      "epoch": 8.98231846627595,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.141849607628886e-05,
+      "loss": 0.4769,
+      "step": 180850
+    },
+    {
+      "epoch": 8.98281513857157,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.137876229263932e-05,
+      "loss": 0.466,
+      "step": 180860
+    },
+    {
+      "epoch": 8.983311810867189,
+      "grad_norm": 0.22265625,
+      "learning_rate": 8.133902850898977e-05,
+      "loss": 0.524,
+      "step": 180870
+    },
+    {
+      "epoch": 8.983808483162809,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.129929472534022e-05,
+      "loss": 0.5201,
+      "step": 180880
+    },
+    {
+      "epoch": 8.98430515545843,
+      "grad_norm": 0.18359375,
+      "learning_rate": 8.125956094169067e-05,
+      "loss": 0.5019,
+      "step": 180890
+    },
+    {
+      "epoch": 8.984801827754048,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.121982715804113e-05,
+      "loss": 0.4948,
+      "step": 180900
+    },
+    {
+      "epoch": 8.985298500049668,
+      "grad_norm": 0.19921875,
+      "learning_rate": 8.118009337439157e-05,
+      "loss": 0.5,
+      "step": 180910
+    },
+    {
+      "epoch": 8.985795172345286,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.114035959074203e-05,
+      "loss": 0.4749,
+      "step": 180920
+    },
+    {
+      "epoch": 8.986291844640906,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.110062580709249e-05,
+      "loss": 0.47,
+      "step": 180930
+    },
+    {
+      "epoch": 8.986788516936524,
+      "grad_norm": 0.197265625,
+      "learning_rate": 8.106089202344293e-05,
+      "loss": 0.4902,
+      "step": 180940
+    },
+    {
+      "epoch": 8.987285189232145,
+      "grad_norm": 0.19140625,
+      "learning_rate": 8.102115823979339e-05,
+      "loss": 0.5165,
+      "step": 180950
+    },
+    {
+      "epoch": 8.987781861527765,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.098142445614383e-05,
+      "loss": 0.4855,
+      "step": 180960
+    },
+    {
+      "epoch": 8.988278533823383,
+      "grad_norm": 0.2109375,
+      "learning_rate": 8.094169067249429e-05,
+      "loss": 0.5052,
+      "step": 180970
+    },
+    {
+      "epoch": 8.988775206119003,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 8.090195688884475e-05,
+      "loss": 0.4728,
+      "step": 180980
+    },
+    {
+      "epoch": 8.989271878414621,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 8.08622231051952e-05,
+      "loss": 0.483,
+      "step": 180990
+    },
+    {
+      "epoch": 8.989768550710242,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.082248932154565e-05,
+      "loss": 0.4724,
+      "step": 181000
+    },
+    {
+      "epoch": 8.99026522300586,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 8.078275553789611e-05,
+      "loss": 0.4849,
+      "step": 181010
+    },
+    {
+      "epoch": 8.99076189530148,
+      "grad_norm": 0.228515625,
+      "learning_rate": 8.074302175424655e-05,
+      "loss": 0.486,
+      "step": 181020
+    },
+    {
+      "epoch": 8.9912585675971,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.070328797059701e-05,
+      "loss": 0.5072,
+      "step": 181030
+    },
+    {
+      "epoch": 8.991755239892719,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.066355418694745e-05,
+      "loss": 0.525,
+      "step": 181040
+    },
+    {
+      "epoch": 8.992251912188339,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.062382040329791e-05,
+      "loss": 0.4892,
+      "step": 181050
+    },
+    {
+      "epoch": 8.992748584483957,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.058408661964836e-05,
+      "loss": 0.4923,
+      "step": 181060
+    },
+    {
+      "epoch": 8.993245256779577,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 8.054435283599881e-05,
+      "loss": 0.4851,
+      "step": 181070
+    },
+    {
+      "epoch": 8.993741929075195,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.050461905234926e-05,
+      "loss": 0.4845,
+      "step": 181080
+    },
+    {
+      "epoch": 8.994238601370816,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 8.046488526869972e-05,
+      "loss": 0.4927,
+      "step": 181090
+    },
+    {
+      "epoch": 8.994735273666436,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.042515148505016e-05,
+      "loss": 0.5307,
+      "step": 181100
+    },
+    {
+      "epoch": 8.995231945962054,
+      "grad_norm": 0.18359375,
+      "learning_rate": 8.038541770140062e-05,
+      "loss": 0.4963,
+      "step": 181110
+    },
+    {
+      "epoch": 8.995728618257674,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.034568391775106e-05,
+      "loss": 0.5034,
+      "step": 181120
+    },
+    {
+      "epoch": 8.996225290553292,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.030595013410152e-05,
+      "loss": 0.5045,
+      "step": 181130
+    },
+    {
+      "epoch": 8.996721962848913,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.026621635045198e-05,
+      "loss": 0.4621,
+      "step": 181140
+    },
+    {
+      "epoch": 8.997218635144531,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.022648256680244e-05,
+      "loss": 0.5041,
+      "step": 181150
+    },
+    {
+      "epoch": 8.997715307440151,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.018674878315288e-05,
+      "loss": 0.477,
+      "step": 181160
+    },
+    {
+      "epoch": 8.998211979735771,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 8.014701499950334e-05,
+      "loss": 0.5047,
+      "step": 181170
+    },
+    {
+      "epoch": 8.99870865203139,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.010728121585378e-05,
+      "loss": 0.5087,
+      "step": 181180
+    },
+    {
+      "epoch": 8.99920532432701,
+      "grad_norm": 0.1875,
+      "learning_rate": 8.006754743220424e-05,
+      "loss": 0.495,
+      "step": 181190
+    },
+    {
+      "epoch": 8.999701996622628,
+      "grad_norm": 0.197265625,
+      "learning_rate": 8.002781364855468e-05,
+      "loss": 0.4698,
+      "step": 181200
+    },
+    {
+      "epoch": 9.000198668918248,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.998807986490514e-05,
+      "loss": 0.4873,
+      "step": 181210
+    },
+    {
+      "epoch": 9.000695341213866,
+      "grad_norm": 0.19921875,
+      "learning_rate": 7.994834608125559e-05,
+      "loss": 0.4846,
+      "step": 181220
+    },
+    {
+      "epoch": 9.001192013509487,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.990861229760604e-05,
+      "loss": 0.4822,
+      "step": 181230
+    },
+    {
+      "epoch": 9.001688685805107,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 7.986887851395649e-05,
+      "loss": 0.4589,
+      "step": 181240
+    },
+    {
+      "epoch": 9.002185358100725,
+      "grad_norm": 0.228515625,
+      "learning_rate": 7.982914473030695e-05,
+      "loss": 0.5168,
+      "step": 181250
+    },
+    {
+      "epoch": 9.002682030396345,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.978941094665739e-05,
+      "loss": 0.5155,
+      "step": 181260
+    },
+    {
+      "epoch": 9.003178702691963,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 7.974967716300785e-05,
+      "loss": 0.4748,
+      "step": 181270
+    },
+    {
+      "epoch": 9.003675374987584,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.970994337935829e-05,
+      "loss": 0.4775,
+      "step": 181280
+    },
+    {
+      "epoch": 9.004172047283202,
+      "grad_norm": 0.236328125,
+      "learning_rate": 7.967020959570875e-05,
+      "loss": 0.4732,
+      "step": 181290
+    },
+    {
+      "epoch": 9.004668719578822,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.963047581205921e-05,
+      "loss": 0.4981,
+      "step": 181300
+    },
+    {
+      "epoch": 9.005165391874442,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.959074202840965e-05,
+      "loss": 0.4724,
+      "step": 181310
+    },
+    {
+      "epoch": 9.00566206417006,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.955100824476011e-05,
+      "loss": 0.5017,
+      "step": 181320
+    },
+    {
+      "epoch": 9.00615873646568,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.951127446111057e-05,
+      "loss": 0.4853,
+      "step": 181330
+    },
+    {
+      "epoch": 9.006655408761299,
+      "grad_norm": 0.2109375,
+      "learning_rate": 7.947154067746103e-05,
+      "loss": 0.4691,
+      "step": 181340
+    },
+    {
+      "epoch": 9.007152081056919,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.943180689381147e-05,
+      "loss": 0.5132,
+      "step": 181350
+    },
+    {
+      "epoch": 9.007648753352537,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 7.939207311016193e-05,
+      "loss": 0.4841,
+      "step": 181360
+    },
+    {
+      "epoch": 9.008145425648157,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.935233932651237e-05,
+      "loss": 0.4567,
+      "step": 181370
+    },
+    {
+      "epoch": 9.008642097943778,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 7.931260554286283e-05,
+      "loss": 0.4849,
+      "step": 181380
+    },
+    {
+      "epoch": 9.009138770239396,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 7.927287175921327e-05,
+      "loss": 0.4844,
+      "step": 181390
+    },
+    {
+      "epoch": 9.009635442535016,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 7.923313797556373e-05,
+      "loss": 0.4968,
+      "step": 181400
+    },
+    {
+      "epoch": 9.010132114830634,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.919340419191418e-05,
+      "loss": 0.4743,
+      "step": 181410
+    },
+    {
+      "epoch": 9.010628787126254,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.915367040826463e-05,
+      "loss": 0.512,
+      "step": 181420
+    },
+    {
+      "epoch": 9.011125459421873,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 7.911393662461508e-05,
+      "loss": 0.4897,
+      "step": 181430
+    },
+    {
+      "epoch": 9.011622131717493,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.907420284096554e-05,
+      "loss": 0.4648,
+      "step": 181440
+    },
+    {
+      "epoch": 9.012118804013111,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.903446905731598e-05,
+      "loss": 0.5017,
+      "step": 181450
+    },
+    {
+      "epoch": 9.012615476308731,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.899473527366644e-05,
+      "loss": 0.4461,
+      "step": 181460
+    },
+    {
+      "epoch": 9.013112148604352,
+      "grad_norm": 0.2109375,
+      "learning_rate": 7.895500149001688e-05,
+      "loss": 0.5197,
+      "step": 181470
+    },
+    {
+      "epoch": 9.01360882089997,
+      "grad_norm": 0.19921875,
+      "learning_rate": 7.891526770636734e-05,
+      "loss": 0.5218,
+      "step": 181480
+    },
+    {
+      "epoch": 9.01410549319559,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.88755339227178e-05,
+      "loss": 0.4734,
+      "step": 181490
+    },
+    {
+      "epoch": 9.014602165491208,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 7.883580013906826e-05,
+      "loss": 0.4906,
+      "step": 181500
+    },
+    {
+      "epoch": 9.015098837786828,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.87960663554187e-05,
+      "loss": 0.4702,
+      "step": 181510
+    },
+    {
+      "epoch": 9.015595510082447,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.875633257176916e-05,
+      "loss": 0.4654,
+      "step": 181520
+    },
+    {
+      "epoch": 9.016092182378067,
+      "grad_norm": 0.1953125,
+      "learning_rate": 7.87165987881196e-05,
+      "loss": 0.4803,
+      "step": 181530
+    },
+    {
+      "epoch": 9.016588854673687,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.867686500447006e-05,
+      "loss": 0.4961,
+      "step": 181540
+    },
+    {
+      "epoch": 9.017085526969305,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.86371312208205e-05,
+      "loss": 0.4313,
+      "step": 181550
+    },
+    {
+      "epoch": 9.017582199264925,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 7.859739743717096e-05,
+      "loss": 0.4801,
+      "step": 181560
+    },
+    {
+      "epoch": 9.018078871560544,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.85576636535214e-05,
+      "loss": 0.4918,
+      "step": 181570
+    },
+    {
+      "epoch": 9.018575543856164,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 7.851792986987186e-05,
+      "loss": 0.4472,
+      "step": 181580
+    },
+    {
+      "epoch": 9.019072216151782,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.847819608622231e-05,
+      "loss": 0.455,
+      "step": 181590
+    },
+    {
+      "epoch": 9.019568888447402,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.843846230257277e-05,
+      "loss": 0.5086,
+      "step": 181600
+    },
+    {
+      "epoch": 9.020065560743022,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.839872851892321e-05,
+      "loss": 0.4726,
+      "step": 181610
+    },
+    {
+      "epoch": 9.02056223303864,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 7.835899473527367e-05,
+      "loss": 0.4827,
+      "step": 181620
+    },
+    {
+      "epoch": 9.021058905334261,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.831926095162411e-05,
+      "loss": 0.5131,
+      "step": 181630
+    },
+    {
+      "epoch": 9.02155557762988,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 7.827952716797457e-05,
+      "loss": 0.4827,
+      "step": 181640
+    },
+    {
+      "epoch": 9.0220522499255,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.823979338432503e-05,
+      "loss": 0.4884,
+      "step": 181650
+    },
+    {
+      "epoch": 9.022548922221118,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.820005960067547e-05,
+      "loss": 0.482,
+      "step": 181660
+    },
+    {
+      "epoch": 9.023045594516738,
+      "grad_norm": 0.220703125,
+      "learning_rate": 7.816032581702593e-05,
+      "loss": 0.5079,
+      "step": 181670
+    },
+    {
+      "epoch": 9.023542266812358,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.812059203337639e-05,
+      "loss": 0.4843,
+      "step": 181680
+    },
+    {
+      "epoch": 9.024038939107976,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.808085824972683e-05,
+      "loss": 0.4598,
+      "step": 181690
+    },
+    {
+      "epoch": 9.024535611403596,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.804112446607729e-05,
+      "loss": 0.4772,
+      "step": 181700
+    },
+    {
+      "epoch": 9.025032283699215,
+      "grad_norm": 0.201171875,
+      "learning_rate": 7.800139068242773e-05,
+      "loss": 0.5073,
+      "step": 181710
+    },
+    {
+      "epoch": 9.025528955994835,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 7.796165689877819e-05,
+      "loss": 0.4913,
+      "step": 181720
+    },
+    {
+      "epoch": 9.026025628290453,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 7.792192311512864e-05,
+      "loss": 0.4884,
+      "step": 181730
+    },
+    {
+      "epoch": 9.026522300586073,
+      "grad_norm": 0.1953125,
+      "learning_rate": 7.78821893314791e-05,
+      "loss": 0.4637,
+      "step": 181740
+    },
+    {
+      "epoch": 9.027018972881693,
+      "grad_norm": 0.2109375,
+      "learning_rate": 7.784245554782954e-05,
+      "loss": 0.4977,
+      "step": 181750
+    },
+    {
+      "epoch": 9.027515645177312,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.780272176418e-05,
+      "loss": 0.479,
+      "step": 181760
+    },
+    {
+      "epoch": 9.028012317472932,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 7.776298798053045e-05,
+      "loss": 0.5042,
+      "step": 181770
+    },
+    {
+      "epoch": 9.02850898976855,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 7.77232541968809e-05,
+      "loss": 0.5161,
+      "step": 181780
+    },
+    {
+      "epoch": 9.02900566206417,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.768352041323136e-05,
+      "loss": 0.4433,
+      "step": 181790
+    },
+    {
+      "epoch": 9.029502334359789,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.76437866295818e-05,
+      "loss": 0.4592,
+      "step": 181800
+    },
+    {
+      "epoch": 9.029999006655409,
+      "grad_norm": 0.224609375,
+      "learning_rate": 7.760405284593226e-05,
+      "loss": 0.4839,
+      "step": 181810
+    },
+    {
+      "epoch": 9.030495678951029,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 7.75643190622827e-05,
+      "loss": 0.5169,
+      "step": 181820
+    },
+    {
+      "epoch": 9.030992351246647,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 7.752458527863316e-05,
+      "loss": 0.5034,
+      "step": 181830
+    },
+    {
+      "epoch": 9.031489023542267,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.748485149498362e-05,
+      "loss": 0.4805,
+      "step": 181840
+    },
+    {
+      "epoch": 9.031985695837886,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 7.744511771133408e-05,
+      "loss": 0.474,
+      "step": 181850
+    },
+    {
+      "epoch": 9.032482368133506,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.740538392768452e-05,
+      "loss": 0.4976,
+      "step": 181860
+    },
+    {
+      "epoch": 9.032979040429124,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.736565014403498e-05,
+      "loss": 0.4408,
+      "step": 181870
+    },
+    {
+      "epoch": 9.033475712724744,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.732591636038542e-05,
+      "loss": 0.5109,
+      "step": 181880
+    },
+    {
+      "epoch": 9.033972385020364,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 7.728618257673588e-05,
+      "loss": 0.4535,
+      "step": 181890
+    },
+    {
+      "epoch": 9.034469057315983,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.724644879308632e-05,
+      "loss": 0.4903,
+      "step": 181900
+    },
+    {
+      "epoch": 9.034965729611603,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.720671500943678e-05,
+      "loss": 0.5,
+      "step": 181910
+    },
+    {
+      "epoch": 9.035462401907221,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 7.716698122578723e-05,
+      "loss": 0.5069,
+      "step": 181920
+    },
+    {
+      "epoch": 9.035959074202841,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.712724744213768e-05,
+      "loss": 0.4912,
+      "step": 181930
+    },
+    {
+      "epoch": 9.03645574649846,
+      "grad_norm": 0.228515625,
+      "learning_rate": 7.708751365848813e-05,
+      "loss": 0.4778,
+      "step": 181940
+    },
+    {
+      "epoch": 9.03695241879408,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 7.704777987483859e-05,
+      "loss": 0.5301,
+      "step": 181950
+    },
+    {
+      "epoch": 9.0374490910897,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.700804609118903e-05,
+      "loss": 0.4735,
+      "step": 181960
+    },
+    {
+      "epoch": 9.037945763385318,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 7.696831230753949e-05,
+      "loss": 0.4786,
+      "step": 181970
+    },
+    {
+      "epoch": 9.038442435680938,
+      "grad_norm": 0.201171875,
+      "learning_rate": 7.692857852388993e-05,
+      "loss": 0.4826,
+      "step": 181980
+    },
+    {
+      "epoch": 9.038939107976557,
+      "grad_norm": 0.1953125,
+      "learning_rate": 7.688884474024039e-05,
+      "loss": 0.4855,
+      "step": 181990
+    },
+    {
+      "epoch": 9.039435780272177,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 7.684911095659085e-05,
+      "loss": 0.4959,
+      "step": 182000
+    },
+    {
+      "epoch": 9.039932452567795,
+      "grad_norm": 0.20703125,
+      "learning_rate": 7.680937717294129e-05,
+      "loss": 0.4911,
+      "step": 182010
+    },
+    {
+      "epoch": 9.040429124863415,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 7.676964338929175e-05,
+      "loss": 0.4749,
+      "step": 182020
+    },
+    {
+      "epoch": 9.040925797159035,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.672990960564221e-05,
+      "loss": 0.5086,
+      "step": 182030
+    },
+    {
+      "epoch": 9.041422469454654,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.669017582199265e-05,
+      "loss": 0.4991,
+      "step": 182040
+    },
+    {
+      "epoch": 9.041919141750274,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.665044203834311e-05,
+      "loss": 0.4976,
+      "step": 182050
+    },
+    {
+      "epoch": 9.042415814045892,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.661070825469355e-05,
+      "loss": 0.4819,
+      "step": 182060
+    },
+    {
+      "epoch": 9.042912486341512,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 7.657097447104401e-05,
+      "loss": 0.4784,
+      "step": 182070
+    },
+    {
+      "epoch": 9.04340915863713,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.653124068739446e-05,
+      "loss": 0.5272,
+      "step": 182080
+    },
+    {
+      "epoch": 9.04390583093275,
+      "grad_norm": 0.20703125,
+      "learning_rate": 7.649150690374491e-05,
+      "loss": 0.4936,
+      "step": 182090
+    },
+    {
+      "epoch": 9.04440250322837,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.645177312009536e-05,
+      "loss": 0.479,
+      "step": 182100
+    },
+    {
+      "epoch": 9.04489917552399,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 7.641203933644582e-05,
+      "loss": 0.4642,
+      "step": 182110
+    },
+    {
+      "epoch": 9.04539584781961,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.637230555279626e-05,
+      "loss": 0.4864,
+      "step": 182120
+    },
+    {
+      "epoch": 9.045892520115228,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.633257176914672e-05,
+      "loss": 0.5005,
+      "step": 182130
+    },
+    {
+      "epoch": 9.046389192410848,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.629283798549716e-05,
+      "loss": 0.4765,
+      "step": 182140
+    },
+    {
+      "epoch": 9.046885864706466,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.625310420184762e-05,
+      "loss": 0.4727,
+      "step": 182150
+    },
+    {
+      "epoch": 9.047382537002086,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.621337041819806e-05,
+      "loss": 0.4515,
+      "step": 182160
+    },
+    {
+      "epoch": 9.047879209297705,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.617363663454852e-05,
+      "loss": 0.505,
+      "step": 182170
+    },
+    {
+      "epoch": 9.048375881593325,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 7.613390285089898e-05,
+      "loss": 0.4773,
+      "step": 182180
+    },
+    {
+      "epoch": 9.048872553888945,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.609416906724944e-05,
+      "loss": 0.4876,
+      "step": 182190
+    },
+    {
+      "epoch": 9.049369226184563,
+      "grad_norm": 0.2109375,
+      "learning_rate": 7.60544352835999e-05,
+      "loss": 0.4807,
+      "step": 182200
+    },
+    {
+      "epoch": 9.049865898480183,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.601470149995034e-05,
+      "loss": 0.5126,
+      "step": 182210
+    },
+    {
+      "epoch": 9.050362570775802,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.59749677163008e-05,
+      "loss": 0.5207,
+      "step": 182220
+    },
+    {
+      "epoch": 9.050859243071422,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 7.593523393265124e-05,
+      "loss": 0.4763,
+      "step": 182230
+    },
+    {
+      "epoch": 9.05135591536704,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 7.58955001490017e-05,
+      "loss": 0.483,
+      "step": 182240
+    },
+    {
+      "epoch": 9.05185258766266,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.585576636535214e-05,
+      "loss": 0.4933,
+      "step": 182250
+    },
+    {
+      "epoch": 9.05234925995828,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.58160325817026e-05,
+      "loss": 0.4884,
+      "step": 182260
+    },
+    {
+      "epoch": 9.052845932253899,
+      "grad_norm": 0.197265625,
+      "learning_rate": 7.577629879805305e-05,
+      "loss": 0.4849,
+      "step": 182270
+    },
+    {
+      "epoch": 9.053342604549519,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.57365650144035e-05,
+      "loss": 0.5134,
+      "step": 182280
+    },
+    {
+      "epoch": 9.053839276845137,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 7.569683123075395e-05,
+      "loss": 0.4907,
+      "step": 182290
+    },
+    {
+      "epoch": 9.054335949140757,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.56570974471044e-05,
+      "loss": 0.4962,
+      "step": 182300
+    },
+    {
+      "epoch": 9.054832621436375,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.561736366345485e-05,
+      "loss": 0.48,
+      "step": 182310
+    },
+    {
+      "epoch": 9.055329293731996,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.557762987980531e-05,
+      "loss": 0.4899,
+      "step": 182320
+    },
+    {
+      "epoch": 9.055825966027616,
+      "grad_norm": 0.205078125,
+      "learning_rate": 7.553789609615575e-05,
+      "loss": 0.4762,
+      "step": 182330
+    },
+    {
+      "epoch": 9.056322638323234,
+      "grad_norm": 0.220703125,
+      "learning_rate": 7.549816231250621e-05,
+      "loss": 0.4973,
+      "step": 182340
+    },
+    {
+      "epoch": 9.056819310618854,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.545842852885667e-05,
+      "loss": 0.4652,
+      "step": 182350
+    },
+    {
+      "epoch": 9.057315982914472,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.541869474520713e-05,
+      "loss": 0.4807,
+      "step": 182360
+    },
+    {
+      "epoch": 9.057812655210093,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.537896096155757e-05,
+      "loss": 0.4666,
+      "step": 182370
+    },
+    {
+      "epoch": 9.058309327505711,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 7.533922717790803e-05,
+      "loss": 0.4926,
+      "step": 182380
+    },
+    {
+      "epoch": 9.058805999801331,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.529949339425847e-05,
+      "loss": 0.4793,
+      "step": 182390
+    },
+    {
+      "epoch": 9.059302672096951,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 7.525975961060893e-05,
+      "loss": 0.4722,
+      "step": 182400
+    },
+    {
+      "epoch": 9.05979934439257,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.522002582695937e-05,
+      "loss": 0.4739,
+      "step": 182410
+    },
+    {
+      "epoch": 9.06029601668819,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.518029204330983e-05,
+      "loss": 0.4611,
+      "step": 182420
+    },
+    {
+      "epoch": 9.060792688983808,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.514055825966028e-05,
+      "loss": 0.4728,
+      "step": 182430
+    },
+    {
+      "epoch": 9.061289361279428,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 7.510082447601073e-05,
+      "loss": 0.5058,
+      "step": 182440
+    },
+    {
+      "epoch": 9.061786033575046,
+      "grad_norm": 0.20703125,
+      "learning_rate": 7.506109069236118e-05,
+      "loss": 0.5037,
+      "step": 182450
+    },
+    {
+      "epoch": 9.062282705870667,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.502135690871164e-05,
+      "loss": 0.4864,
+      "step": 182460
+    },
+    {
+      "epoch": 9.062779378166287,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.498162312506208e-05,
+      "loss": 0.5376,
+      "step": 182470
+    },
+    {
+      "epoch": 9.063276050461905,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.494188934141254e-05,
+      "loss": 0.4829,
+      "step": 182480
+    },
+    {
+      "epoch": 9.063772722757525,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 7.490215555776298e-05,
+      "loss": 0.4903,
+      "step": 182490
+    },
+    {
+      "epoch": 9.064269395053143,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.486242177411344e-05,
+      "loss": 0.4897,
+      "step": 182500
+    },
+    {
+      "epoch": 9.064766067348764,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 7.482268799046388e-05,
+      "loss": 0.4932,
+      "step": 182510
+    },
+    {
+      "epoch": 9.065262739644382,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 7.478295420681434e-05,
+      "loss": 0.5258,
+      "step": 182520
+    },
+    {
+      "epoch": 9.065759411940002,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.47432204231648e-05,
+      "loss": 0.5126,
+      "step": 182530
+    },
+    {
+      "epoch": 9.066256084235622,
+      "grad_norm": 0.20703125,
+      "learning_rate": 7.470348663951526e-05,
+      "loss": 0.4997,
+      "step": 182540
+    },
+    {
+      "epoch": 9.06675275653124,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.46637528558657e-05,
+      "loss": 0.4666,
+      "step": 182550
+    },
+    {
+      "epoch": 9.06724942882686,
+      "grad_norm": 0.19921875,
+      "learning_rate": 7.462401907221616e-05,
+      "loss": 0.4605,
+      "step": 182560
+    },
+    {
+      "epoch": 9.067746101122479,
+      "grad_norm": 0.23046875,
+      "learning_rate": 7.45842852885666e-05,
+      "loss": 0.4722,
+      "step": 182570
+    },
+    {
+      "epoch": 9.068242773418099,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 7.454455150491706e-05,
+      "loss": 0.5031,
+      "step": 182580
+    },
+    {
+      "epoch": 9.068739445713717,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.45048177212675e-05,
+      "loss": 0.4937,
+      "step": 182590
+    },
+    {
+      "epoch": 9.069236118009337,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 7.446508393761796e-05,
+      "loss": 0.4644,
+      "step": 182600
+    },
+    {
+      "epoch": 9.069732790304958,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 7.442535015396842e-05,
+      "loss": 0.4606,
+      "step": 182610
+    },
+    {
+      "epoch": 9.070229462600576,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 7.438561637031887e-05,
+      "loss": 0.5453,
+      "step": 182620
+    },
+    {
+      "epoch": 9.070726134896196,
+      "grad_norm": 0.208984375,
+      "learning_rate": 7.434588258666932e-05,
+      "loss": 0.4805,
+      "step": 182630
+    },
+    {
+      "epoch": 9.071222807191814,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.430614880301977e-05,
+      "loss": 0.4781,
+      "step": 182640
+    },
+    {
+      "epoch": 9.071719479487435,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.426641501937023e-05,
+      "loss": 0.4755,
+      "step": 182650
+    },
+    {
+      "epoch": 9.072216151783053,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 7.422668123572067e-05,
+      "loss": 0.487,
+      "step": 182660
+    },
+    {
+      "epoch": 9.072712824078673,
+      "grad_norm": 0.197265625,
+      "learning_rate": 7.418694745207113e-05,
+      "loss": 0.5061,
+      "step": 182670
+    },
+    {
+      "epoch": 9.073209496374293,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 7.414721366842157e-05,
+      "loss": 0.4973,
+      "step": 182680
+    },
+    {
+      "epoch": 9.073706168669911,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 7.410747988477203e-05,
+      "loss": 0.4624,
+      "step": 182690
+    },
+    {
+      "epoch": 9.074202840965532,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 7.406774610112249e-05,
+      "loss": 0.4932,
+      "step": 182700
+    },
+    {
+      "epoch": 9.07469951326115,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.402801231747295e-05,
+      "loss": 0.5181,
+      "step": 182710
+    },
+    {
+      "epoch": 9.07519618555677,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 7.398827853382339e-05,
+      "loss": 0.494,
+      "step": 182720
+    },
+    {
+      "epoch": 9.075692857852388,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.394854475017385e-05,
+      "loss": 0.4742,
+      "step": 182730
+    },
+    {
+      "epoch": 9.076189530148008,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.390881096652429e-05,
+      "loss": 0.4949,
+      "step": 182740
+    },
+    {
+      "epoch": 9.076686202443629,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.386907718287475e-05,
+      "loss": 0.4609,
+      "step": 182750
+    },
+    {
+      "epoch": 9.077182874739247,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.38293433992252e-05,
+      "loss": 0.4786,
+      "step": 182760
+    },
+    {
+      "epoch": 9.077679547034867,
+      "grad_norm": 0.1953125,
+      "learning_rate": 7.378960961557565e-05,
+      "loss": 0.5006,
+      "step": 182770
+    },
+    {
+      "epoch": 9.078176219330485,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.37498758319261e-05,
+      "loss": 0.478,
+      "step": 182780
+    },
+    {
+      "epoch": 9.078672891626105,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 7.371014204827655e-05,
+      "loss": 0.5064,
+      "step": 182790
+    },
+    {
+      "epoch": 9.079169563921724,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.3670408264627e-05,
+      "loss": 0.482,
+      "step": 182800
+    },
+    {
+      "epoch": 9.079666236217344,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 7.363067448097746e-05,
+      "loss": 0.4831,
+      "step": 182810
+    },
+    {
+      "epoch": 9.080162908512964,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 7.35909406973279e-05,
+      "loss": 0.4944,
+      "step": 182820
+    },
+    {
+      "epoch": 9.080659580808582,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.355120691367836e-05,
+      "loss": 0.4607,
+      "step": 182830
+    },
+    {
+      "epoch": 9.081156253104202,
+      "grad_norm": 0.203125,
+      "learning_rate": 7.35114731300288e-05,
+      "loss": 0.4474,
+      "step": 182840
+    },
+    {
+      "epoch": 9.08165292539982,
+      "grad_norm": 0.21484375,
+      "learning_rate": 7.347173934637926e-05,
+      "loss": 0.4873,
+      "step": 182850
+    },
+    {
+      "epoch": 9.082149597695441,
+      "grad_norm": 0.208984375,
+      "learning_rate": 7.34320055627297e-05,
+      "loss": 0.4632,
+      "step": 182860
+    },
+    {
+      "epoch": 9.08264626999106,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.339227177908016e-05,
+      "loss": 0.4593,
+      "step": 182870
+    },
+    {
+      "epoch": 9.08314294228668,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.335253799543062e-05,
+      "loss": 0.464,
+      "step": 182880
+    },
+    {
+      "epoch": 9.083639614582298,
+      "grad_norm": 0.24609375,
+      "learning_rate": 7.331280421178108e-05,
+      "loss": 0.4516,
+      "step": 182890
+    },
+    {
+      "epoch": 9.084136286877918,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.327307042813152e-05,
+      "loss": 0.4705,
+      "step": 182900
+    },
+    {
+      "epoch": 9.084632959173538,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 7.323333664448198e-05,
+      "loss": 0.5121,
+      "step": 182910
+    },
+    {
+      "epoch": 9.085129631469156,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 7.319360286083242e-05,
+      "loss": 0.4887,
+      "step": 182920
+    },
+    {
+      "epoch": 9.085626303764776,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 7.315386907718288e-05,
+      "loss": 0.4915,
+      "step": 182930
+    },
+    {
+      "epoch": 9.086122976060395,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 7.311413529353333e-05,
+      "loss": 0.4873,
+      "step": 182940
+    },
+    {
+      "epoch": 9.086619648356015,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.307440150988378e-05,
+      "loss": 0.4949,
+      "step": 182950
+    },
+    {
+      "epoch": 9.087116320651633,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.303466772623423e-05,
+      "loss": 0.4797,
+      "step": 182960
+    },
+    {
+      "epoch": 9.087612992947253,
+      "grad_norm": 0.203125,
+      "learning_rate": 7.299493394258469e-05,
+      "loss": 0.4649,
+      "step": 182970
+    },
+    {
+      "epoch": 9.088109665242873,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.295520015893513e-05,
+      "loss": 0.4682,
+      "step": 182980
+    },
+    {
+      "epoch": 9.088606337538492,
+      "grad_norm": 0.197265625,
+      "learning_rate": 7.291546637528559e-05,
+      "loss": 0.5045,
+      "step": 182990
+    },
+    {
+      "epoch": 9.089103009834112,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.287573259163603e-05,
+      "loss": 0.4728,
+      "step": 183000
+    },
+    {
+      "epoch": 9.08959968212973,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 7.283599880798649e-05,
+      "loss": 0.4804,
+      "step": 183010
+    },
+    {
+      "epoch": 9.09009635442535,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.279626502433695e-05,
+      "loss": 0.4665,
+      "step": 183020
+    },
+    {
+      "epoch": 9.090593026720969,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.275653124068739e-05,
+      "loss": 0.4453,
+      "step": 183030
+    },
+    {
+      "epoch": 9.091089699016589,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.271679745703785e-05,
+      "loss": 0.5243,
+      "step": 183040
+    },
+    {
+      "epoch": 9.091586371312209,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 7.267706367338831e-05,
+      "loss": 0.5243,
+      "step": 183050
+    },
+    {
+      "epoch": 9.092083043607827,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.263732988973877e-05,
+      "loss": 0.4825,
+      "step": 183060
+    },
+    {
+      "epoch": 9.092579715903447,
+      "grad_norm": 0.203125,
+      "learning_rate": 7.259759610608921e-05,
+      "loss": 0.4764,
+      "step": 183070
+    },
+    {
+      "epoch": 9.093076388199066,
+      "grad_norm": 0.201171875,
+      "learning_rate": 7.255786232243967e-05,
+      "loss": 0.4724,
+      "step": 183080
+    },
+    {
+      "epoch": 9.093573060494686,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.251812853879011e-05,
+      "loss": 0.4651,
+      "step": 183090
+    },
+    {
+      "epoch": 9.094069732790304,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.247839475514057e-05,
+      "loss": 0.5167,
+      "step": 183100
+    },
+    {
+      "epoch": 9.094566405085924,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.243866097149101e-05,
+      "loss": 0.4953,
+      "step": 183110
+    },
+    {
+      "epoch": 9.095063077381544,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.239892718784147e-05,
+      "loss": 0.5026,
+      "step": 183120
+    },
+    {
+      "epoch": 9.095559749677163,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 7.235919340419192e-05,
+      "loss": 0.4616,
+      "step": 183130
+    },
+    {
+      "epoch": 9.096056421972783,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 7.231945962054237e-05,
+      "loss": 0.4756,
+      "step": 183140
+    },
+    {
+      "epoch": 9.096553094268401,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 7.227972583689282e-05,
+      "loss": 0.503,
+      "step": 183150
+    },
+    {
+      "epoch": 9.097049766564021,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.223999205324328e-05,
+      "loss": 0.5279,
+      "step": 183160
+    },
+    {
+      "epoch": 9.09754643885964,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.220025826959372e-05,
+      "loss": 0.4922,
+      "step": 183170
+    },
+    {
+      "epoch": 9.09804311115526,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 7.216052448594418e-05,
+      "loss": 0.4952,
+      "step": 183180
+    },
+    {
+      "epoch": 9.09853978345088,
+      "grad_norm": 0.23828125,
+      "learning_rate": 7.212079070229462e-05,
+      "loss": 0.5045,
+      "step": 183190
+    },
+    {
+      "epoch": 9.099036455746498,
+      "grad_norm": 0.25,
+      "learning_rate": 7.208105691864508e-05,
+      "loss": 0.5088,
+      "step": 183200
+    },
+    {
+      "epoch": 9.099533128042118,
+      "grad_norm": 0.2109375,
+      "learning_rate": 7.204132313499554e-05,
+      "loss": 0.4668,
+      "step": 183210
+    },
+    {
+      "epoch": 9.100029800337737,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.200158935134598e-05,
+      "loss": 0.4921,
+      "step": 183220
+    },
+    {
+      "epoch": 9.100526472633357,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.196185556769644e-05,
+      "loss": 0.4922,
+      "step": 183230
+    },
+    {
+      "epoch": 9.101023144928975,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.19221217840469e-05,
+      "loss": 0.4779,
+      "step": 183240
+    },
+    {
+      "epoch": 9.101519817224595,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 7.188238800039734e-05,
+      "loss": 0.477,
+      "step": 183250
+    },
+    {
+      "epoch": 9.102016489520215,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.18426542167478e-05,
+      "loss": 0.5114,
+      "step": 183260
+    },
+    {
+      "epoch": 9.102513161815834,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.180292043309824e-05,
+      "loss": 0.4597,
+      "step": 183270
+    },
+    {
+      "epoch": 9.103009834111454,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 7.17631866494487e-05,
+      "loss": 0.5096,
+      "step": 183280
+    },
+    {
+      "epoch": 9.103506506407072,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.172345286579915e-05,
+      "loss": 0.5081,
+      "step": 183290
+    },
+    {
+      "epoch": 9.104003178702692,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.16837190821496e-05,
+      "loss": 0.474,
+      "step": 183300
+    },
+    {
+      "epoch": 9.10449985099831,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.164398529850005e-05,
+      "loss": 0.467,
+      "step": 183310
+    },
+    {
+      "epoch": 9.10499652329393,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 7.16042515148505e-05,
+      "loss": 0.4756,
+      "step": 183320
+    },
+    {
+      "epoch": 9.10549319558955,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.156451773120095e-05,
+      "loss": 0.4683,
+      "step": 183330
+    },
+    {
+      "epoch": 9.10598986788517,
+      "grad_norm": 0.205078125,
+      "learning_rate": 7.152478394755141e-05,
+      "loss": 0.4684,
+      "step": 183340
+    },
+    {
+      "epoch": 9.10648654018079,
+      "grad_norm": 0.232421875,
+      "learning_rate": 7.148505016390185e-05,
+      "loss": 0.5029,
+      "step": 183350
+    },
+    {
+      "epoch": 9.106983212476408,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.144531638025231e-05,
+      "loss": 0.5172,
+      "step": 183360
+    },
+    {
+      "epoch": 9.107479884772028,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.140558259660275e-05,
+      "loss": 0.496,
+      "step": 183370
+    },
+    {
+      "epoch": 9.107976557067646,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.136584881295321e-05,
+      "loss": 0.4647,
+      "step": 183380
+    },
+    {
+      "epoch": 9.108473229363266,
+      "grad_norm": 0.212890625,
+      "learning_rate": 7.132611502930367e-05,
+      "loss": 0.4909,
+      "step": 183390
+    },
+    {
+      "epoch": 9.108969901658886,
+      "grad_norm": 0.212890625,
+      "learning_rate": 7.128638124565413e-05,
+      "loss": 0.5088,
+      "step": 183400
+    },
+    {
+      "epoch": 9.109466573954505,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 7.124664746200457e-05,
+      "loss": 0.4997,
+      "step": 183410
+    },
+    {
+      "epoch": 9.109963246250125,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.120691367835503e-05,
+      "loss": 0.4552,
+      "step": 183420
+    },
+    {
+      "epoch": 9.110459918545743,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.116717989470547e-05,
+      "loss": 0.4588,
+      "step": 183430
+    },
+    {
+      "epoch": 9.110956590841363,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.112744611105593e-05,
+      "loss": 0.4673,
+      "step": 183440
+    },
+    {
+      "epoch": 9.111453263136982,
+      "grad_norm": 0.244140625,
+      "learning_rate": 7.108771232740639e-05,
+      "loss": 0.498,
+      "step": 183450
+    },
+    {
+      "epoch": 9.111949935432602,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.104797854375683e-05,
+      "loss": 0.4668,
+      "step": 183460
+    },
+    {
+      "epoch": 9.112446607728222,
+      "grad_norm": 0.21875,
+      "learning_rate": 7.100824476010729e-05,
+      "loss": 0.4676,
+      "step": 183470
+    },
+    {
+      "epoch": 9.11294328002384,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 7.096851097645774e-05,
+      "loss": 0.5052,
+      "step": 183480
+    },
+    {
+      "epoch": 9.11343995231946,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.09287771928082e-05,
+      "loss": 0.4709,
+      "step": 183490
+    },
+    {
+      "epoch": 9.113936624615079,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.088904340915864e-05,
+      "loss": 0.4904,
+      "step": 183500
+    },
+    {
+      "epoch": 9.114433296910699,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.08493096255091e-05,
+      "loss": 0.4782,
+      "step": 183510
+    },
+    {
+      "epoch": 9.114929969206317,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.080957584185954e-05,
+      "loss": 0.4843,
+      "step": 183520
+    },
+    {
+      "epoch": 9.115426641501937,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.076984205821e-05,
+      "loss": 0.4844,
+      "step": 183530
+    },
+    {
+      "epoch": 9.115923313797555,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.073010827456044e-05,
+      "loss": 0.4814,
+      "step": 183540
+    },
+    {
+      "epoch": 9.116419986093176,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 7.06903744909109e-05,
+      "loss": 0.5254,
+      "step": 183550
+    },
+    {
+      "epoch": 9.116916658388796,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.065064070726136e-05,
+      "loss": 0.495,
+      "step": 183560
+    },
+    {
+      "epoch": 9.117413330684414,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.06109069236118e-05,
+      "loss": 0.4925,
+      "step": 183570
+    },
+    {
+      "epoch": 9.117910002980034,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.057117313996226e-05,
+      "loss": 0.5212,
+      "step": 183580
+    },
+    {
+      "epoch": 9.118406675275653,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.053143935631272e-05,
+      "loss": 0.4711,
+      "step": 183590
+    },
+    {
+      "epoch": 9.118903347571273,
+      "grad_norm": 0.201171875,
+      "learning_rate": 7.049170557266316e-05,
+      "loss": 0.4479,
+      "step": 183600
+    },
+    {
+      "epoch": 9.119400019866891,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.045197178901362e-05,
+      "loss": 0.4634,
+      "step": 183610
+    },
+    {
+      "epoch": 9.119896692162511,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 7.041223800536406e-05,
+      "loss": 0.5072,
+      "step": 183620
+    },
+    {
+      "epoch": 9.120393364458131,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.037250422171452e-05,
+      "loss": 0.4743,
+      "step": 183630
+    },
+    {
+      "epoch": 9.12089003675375,
+      "grad_norm": 0.20703125,
+      "learning_rate": 7.033277043806497e-05,
+      "loss": 0.4725,
+      "step": 183640
+    },
+    {
+      "epoch": 9.12138670904937,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 7.029303665441542e-05,
+      "loss": 0.483,
+      "step": 183650
+    },
+    {
+      "epoch": 9.121883381344988,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.025330287076587e-05,
+      "loss": 0.4595,
+      "step": 183660
+    },
+    {
+      "epoch": 9.122380053640608,
+      "grad_norm": 0.205078125,
+      "learning_rate": 7.021356908711633e-05,
+      "loss": 0.4898,
+      "step": 183670
+    },
+    {
+      "epoch": 9.122876725936226,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.017383530346677e-05,
+      "loss": 0.4713,
+      "step": 183680
+    },
+    {
+      "epoch": 9.123373398231847,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.013410151981723e-05,
+      "loss": 0.4793,
+      "step": 183690
+    },
+    {
+      "epoch": 9.123870070527467,
+      "grad_norm": 0.205078125,
+      "learning_rate": 7.009436773616767e-05,
+      "loss": 0.4977,
+      "step": 183700
+    },
+    {
+      "epoch": 9.124366742823085,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.005463395251813e-05,
+      "loss": 0.4537,
+      "step": 183710
+    },
+    {
+      "epoch": 9.124863415118705,
+      "grad_norm": 0.197265625,
+      "learning_rate": 7.001490016886857e-05,
+      "loss": 0.5065,
+      "step": 183720
+    },
+    {
+      "epoch": 9.125360087414323,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 6.997516638521903e-05,
+      "loss": 0.5102,
+      "step": 183730
+    },
+    {
+      "epoch": 9.125856759709944,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 6.993543260156949e-05,
+      "loss": 0.5043,
+      "step": 183740
+    },
+    {
+      "epoch": 9.126353432005562,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 6.989569881791995e-05,
+      "loss": 0.5042,
+      "step": 183750
+    },
+    {
+      "epoch": 9.126850104301182,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.985596503427039e-05,
+      "loss": 0.5023,
+      "step": 183760
+    },
+    {
+      "epoch": 9.127346776596802,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.981623125062085e-05,
+      "loss": 0.4962,
+      "step": 183770
+    },
+    {
+      "epoch": 9.12784344889242,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.97764974669713e-05,
+      "loss": 0.4831,
+      "step": 183780
+    },
+    {
+      "epoch": 9.12834012118804,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 6.973676368332175e-05,
+      "loss": 0.5234,
+      "step": 183790
+    },
+    {
+      "epoch": 9.128836793483659,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 6.96970298996722e-05,
+      "loss": 0.4678,
+      "step": 183800
+    },
+    {
+      "epoch": 9.129333465779279,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 6.965729611602265e-05,
+      "loss": 0.4922,
+      "step": 183810
+    },
+    {
+      "epoch": 9.129830138074897,
+      "grad_norm": 0.18359375,
+      "learning_rate": 6.96175623323731e-05,
+      "loss": 0.4929,
+      "step": 183820
+    },
+    {
+      "epoch": 9.130326810370518,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.957782854872356e-05,
+      "loss": 0.4406,
+      "step": 183830
+    },
+    {
+      "epoch": 9.130823482666138,
+      "grad_norm": 0.201171875,
+      "learning_rate": 6.9538094765074e-05,
+      "loss": 0.4779,
+      "step": 183840
+    },
+    {
+      "epoch": 9.131320154961756,
+      "grad_norm": 0.197265625,
+      "learning_rate": 6.949836098142446e-05,
+      "loss": 0.4758,
+      "step": 183850
+    },
+    {
+      "epoch": 9.131816827257376,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.94586271977749e-05,
+      "loss": 0.4949,
+      "step": 183860
+    },
+    {
+      "epoch": 9.132313499552994,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.941889341412536e-05,
+      "loss": 0.4358,
+      "step": 183870
+    },
+    {
+      "epoch": 9.132810171848615,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.937915963047582e-05,
+      "loss": 0.4765,
+      "step": 183880
+    },
+    {
+      "epoch": 9.133306844144233,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.933942584682626e-05,
+      "loss": 0.5001,
+      "step": 183890
+    },
+    {
+      "epoch": 9.133803516439853,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.929969206317672e-05,
+      "loss": 0.5032,
+      "step": 183900
+    },
+    {
+      "epoch": 9.134300188735473,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.925995827952718e-05,
+      "loss": 0.4726,
+      "step": 183910
+    },
+    {
+      "epoch": 9.134796861031091,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 6.922022449587762e-05,
+      "loss": 0.4912,
+      "step": 183920
+    },
+    {
+      "epoch": 9.135293533326712,
+      "grad_norm": 0.220703125,
+      "learning_rate": 6.918049071222808e-05,
+      "loss": 0.4891,
+      "step": 183930
+    },
+    {
+      "epoch": 9.13579020562233,
+      "grad_norm": 0.224609375,
+      "learning_rate": 6.914075692857854e-05,
+      "loss": 0.4766,
+      "step": 183940
+    },
+    {
+      "epoch": 9.13628687791795,
+      "grad_norm": 0.21875,
+      "learning_rate": 6.910102314492898e-05,
+      "loss": 0.4845,
+      "step": 183950
+    },
+    {
+      "epoch": 9.136783550213568,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 6.906128936127944e-05,
+      "loss": 0.4646,
+      "step": 183960
+    },
+    {
+      "epoch": 9.137280222509188,
+      "grad_norm": 0.208984375,
+      "learning_rate": 6.902155557762988e-05,
+      "loss": 0.4531,
+      "step": 183970
+    },
+    {
+      "epoch": 9.137776894804809,
+      "grad_norm": 0.19921875,
+      "learning_rate": 6.898182179398034e-05,
+      "loss": 0.4454,
+      "step": 183980
+    },
+    {
+      "epoch": 9.138273567100427,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 6.894208801033079e-05,
+      "loss": 0.5006,
+      "step": 183990
+    },
+    {
+      "epoch": 9.138770239396047,
+      "grad_norm": 0.236328125,
+      "learning_rate": 6.890235422668124e-05,
+      "loss": 0.4959,
+      "step": 184000
+    },
+    {
+      "epoch": 9.139266911691665,
+      "grad_norm": 0.220703125,
+      "learning_rate": 6.886262044303169e-05,
+      "loss": 0.4727,
+      "step": 184010
+    },
+    {
+      "epoch": 9.139763583987286,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.882288665938215e-05,
+      "loss": 0.4789,
+      "step": 184020
+    },
+    {
+      "epoch": 9.140260256282904,
+      "grad_norm": 0.23046875,
+      "learning_rate": 6.878315287573259e-05,
+      "loss": 0.5327,
+      "step": 184030
+    },
+    {
+      "epoch": 9.140756928578524,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.874341909208305e-05,
+      "loss": 0.5216,
+      "step": 184040
+    },
+    {
+      "epoch": 9.141253600874144,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.870368530843349e-05,
+      "loss": 0.4922,
+      "step": 184050
+    },
+    {
+      "epoch": 9.141750273169762,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 6.866395152478395e-05,
+      "loss": 0.4965,
+      "step": 184060
+    },
+    {
+      "epoch": 9.142246945465383,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 6.86242177411344e-05,
+      "loss": 0.495,
+      "step": 184070
+    },
+    {
+      "epoch": 9.142743617761,
+      "grad_norm": 0.19921875,
+      "learning_rate": 6.858448395748485e-05,
+      "loss": 0.4731,
+      "step": 184080
+    },
+    {
+      "epoch": 9.143240290056621,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.854475017383531e-05,
+      "loss": 0.4728,
+      "step": 184090
+    },
+    {
+      "epoch": 9.14373696235224,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.850501639018577e-05,
+      "loss": 0.4996,
+      "step": 184100
+    },
+    {
+      "epoch": 9.14423363464786,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.846528260653621e-05,
+      "loss": 0.4944,
+      "step": 184110
+    },
+    {
+      "epoch": 9.14473030694348,
+      "grad_norm": 0.216796875,
+      "learning_rate": 6.842554882288667e-05,
+      "loss": 0.5061,
+      "step": 184120
+    },
+    {
+      "epoch": 9.145226979239098,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 6.838581503923711e-05,
+      "loss": 0.482,
+      "step": 184130
+    },
+    {
+      "epoch": 9.145723651534718,
+      "grad_norm": 0.203125,
+      "learning_rate": 6.834608125558757e-05,
+      "loss": 0.4966,
+      "step": 184140
+    },
+    {
+      "epoch": 9.146220323830336,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 6.830634747193802e-05,
+      "loss": 0.4577,
+      "step": 184150
+    },
+    {
+      "epoch": 9.146716996125956,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.826661368828847e-05,
+      "loss": 0.468,
+      "step": 184160
+    },
+    {
+      "epoch": 9.147213668421575,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.822687990463892e-05,
+      "loss": 0.4917,
+      "step": 184170
+    },
+    {
+      "epoch": 9.147710340717195,
+      "grad_norm": 0.201171875,
+      "learning_rate": 6.818714612098938e-05,
+      "loss": 0.446,
+      "step": 184180
+    },
+    {
+      "epoch": 9.148207013012815,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 6.814741233733982e-05,
+      "loss": 0.4935,
+      "step": 184190
+    },
+    {
+      "epoch": 9.148703685308433,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.810767855369028e-05,
+      "loss": 0.4845,
+      "step": 184200
+    },
+    {
+      "epoch": 9.149200357604053,
+      "grad_norm": 0.234375,
+      "learning_rate": 6.806794477004072e-05,
+      "loss": 0.512,
+      "step": 184210
+    },
+    {
+      "epoch": 9.149697029899672,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.802821098639118e-05,
+      "loss": 0.4937,
+      "step": 184220
+    },
+    {
+      "epoch": 9.150193702195292,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.798847720274162e-05,
+      "loss": 0.4869,
+      "step": 184230
+    },
+    {
+      "epoch": 9.15069037449091,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.794874341909208e-05,
+      "loss": 0.5046,
+      "step": 184240
+    },
+    {
+      "epoch": 9.15118704678653,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.790900963544254e-05,
+      "loss": 0.4638,
+      "step": 184250
+    },
+    {
+      "epoch": 9.151683719082149,
+      "grad_norm": 0.208984375,
+      "learning_rate": 6.7869275851793e-05,
+      "loss": 0.4879,
+      "step": 184260
+    },
+    {
+      "epoch": 9.152180391377769,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.782954206814344e-05,
+      "loss": 0.4779,
+      "step": 184270
+    },
+    {
+      "epoch": 9.152677063673389,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.77898082844939e-05,
+      "loss": 0.4877,
+      "step": 184280
+    },
+    {
+      "epoch": 9.153173735969007,
+      "grad_norm": 0.23828125,
+      "learning_rate": 6.775007450084436e-05,
+      "loss": 0.4687,
+      "step": 184290
+    },
+    {
+      "epoch": 9.153670408264627,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.77103407171948e-05,
+      "loss": 0.474,
+      "step": 184300
+    },
+    {
+      "epoch": 9.154167080560246,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 6.767060693354526e-05,
+      "loss": 0.4974,
+      "step": 184310
+    },
+    {
+      "epoch": 9.154663752855866,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 6.76308731498957e-05,
+      "loss": 0.4704,
+      "step": 184320
+    },
+    {
+      "epoch": 9.155160425151484,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.759113936624616e-05,
+      "loss": 0.494,
+      "step": 184330
+    },
+    {
+      "epoch": 9.155657097447104,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 6.75514055825966e-05,
+      "loss": 0.5045,
+      "step": 184340
+    },
+    {
+      "epoch": 9.156153769742724,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.751167179894706e-05,
+      "loss": 0.462,
+      "step": 184350
+    },
+    {
+      "epoch": 9.156650442038343,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.747193801529751e-05,
+      "loss": 0.4745,
+      "step": 184360
+    },
+    {
+      "epoch": 9.157147114333963,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 6.743220423164797e-05,
+      "loss": 0.5212,
+      "step": 184370
+    },
+    {
+      "epoch": 9.157643786629581,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.739247044799841e-05,
+      "loss": 0.457,
+      "step": 184380
+    },
+    {
+      "epoch": 9.158140458925201,
+      "grad_norm": 0.20703125,
+      "learning_rate": 6.735273666434887e-05,
+      "loss": 0.4669,
+      "step": 184390
+    },
+    {
+      "epoch": 9.15863713122082,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.731300288069931e-05,
+      "loss": 0.4686,
+      "step": 184400
+    },
+    {
+      "epoch": 9.15913380351644,
+      "grad_norm": 0.21484375,
+      "learning_rate": 6.727326909704977e-05,
+      "loss": 0.4725,
+      "step": 184410
+    },
+    {
+      "epoch": 9.15963047581206,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 6.723353531340021e-05,
+      "loss": 0.4822,
+      "step": 184420
+    },
+    {
+      "epoch": 9.160127148107678,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.719380152975067e-05,
+      "loss": 0.4776,
+      "step": 184430
+    },
+    {
+      "epoch": 9.160623820403298,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 6.715406774610113e-05,
+      "loss": 0.4736,
+      "step": 184440
+    },
+    {
+      "epoch": 9.161120492698917,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 6.711433396245159e-05,
+      "loss": 0.4684,
+      "step": 184450
+    },
+    {
+      "epoch": 9.161617164994537,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.707460017880203e-05,
+      "loss": 0.4736,
+      "step": 184460
+    },
+    {
+      "epoch": 9.162113837290155,
+      "grad_norm": 0.224609375,
+      "learning_rate": 6.703486639515249e-05,
+      "loss": 0.4846,
+      "step": 184470
+    },
+    {
+      "epoch": 9.162610509585775,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.699513261150293e-05,
+      "loss": 0.4632,
+      "step": 184480
+    },
+    {
+      "epoch": 9.163107181881395,
+      "grad_norm": 0.251953125,
+      "learning_rate": 6.695539882785339e-05,
+      "loss": 0.4721,
+      "step": 184490
+    },
+    {
+      "epoch": 9.163603854177014,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 6.691566504420384e-05,
+      "loss": 0.5025,
+      "step": 184500
+    },
+    {
+      "epoch": 9.164100526472634,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.687593126055429e-05,
+      "loss": 0.4663,
+      "step": 184510
+    },
+    {
+      "epoch": 9.164597198768252,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.683619747690474e-05,
+      "loss": 0.4928,
+      "step": 184520
+    },
+    {
+      "epoch": 9.165093871063872,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 6.67964636932552e-05,
+      "loss": 0.4811,
+      "step": 184530
+    },
+    {
+      "epoch": 9.16559054335949,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 6.675672990960564e-05,
+      "loss": 0.4769,
+      "step": 184540
+    },
+    {
+      "epoch": 9.16608721565511,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.67169961259561e-05,
+      "loss": 0.4968,
+      "step": 184550
+    },
+    {
+      "epoch": 9.16658388795073,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 6.667726234230654e-05,
+      "loss": 0.4564,
+      "step": 184560
+    },
+    {
+      "epoch": 9.16708056024635,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 6.6637528558657e-05,
+      "loss": 0.4507,
+      "step": 184570
+    },
+    {
+      "epoch": 9.16757723254197,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 6.659779477500744e-05,
+      "loss": 0.5085,
+      "step": 184580
+    },
+    {
+      "epoch": 9.168073904837588,
+      "grad_norm": 0.20703125,
+      "learning_rate": 6.65580609913579e-05,
+      "loss": 0.4818,
+      "step": 184590
+    },
+    {
+      "epoch": 9.168570577133208,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.651832720770836e-05,
+      "loss": 0.4917,
+      "step": 184600
+    },
+    {
+      "epoch": 9.169067249428826,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 6.647859342405882e-05,
+      "loss": 0.4645,
+      "step": 184610
+    },
+    {
+      "epoch": 9.169563921724446,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 6.643885964040926e-05,
+      "loss": 0.4783,
+      "step": 184620
+    },
+    {
+      "epoch": 9.170060594020066,
+      "grad_norm": 0.203125,
+      "learning_rate": 6.639912585675972e-05,
+      "loss": 0.4996,
+      "step": 184630
+    },
+    {
+      "epoch": 9.170557266315685,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 6.635939207311016e-05,
+      "loss": 0.5024,
+      "step": 184640
+    },
+    {
+      "epoch": 9.171053938611305,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.631965828946062e-05,
+      "loss": 0.4781,
+      "step": 184650
+    },
+    {
+      "epoch": 9.171550610906923,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 6.627992450581107e-05,
+      "loss": 0.4547,
+      "step": 184660
+    },
+    {
+      "epoch": 9.172047283202543,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.624019072216152e-05,
+      "loss": 0.4807,
+      "step": 184670
+    },
+    {
+      "epoch": 9.172543955498162,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 6.620045693851197e-05,
+      "loss": 0.504,
+      "step": 184680
+    },
+    {
+      "epoch": 9.173040627793782,
+      "grad_norm": 0.2265625,
+      "learning_rate": 6.616072315486243e-05,
+      "loss": 0.4856,
+      "step": 184690
+    },
+    {
+      "epoch": 9.173537300089402,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.612098937121287e-05,
+      "loss": 0.441,
+      "step": 184700
+    },
+    {
+      "epoch": 9.17403397238502,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 6.608125558756333e-05,
+      "loss": 0.4968,
+      "step": 184710
+    },
+    {
+      "epoch": 9.17453064468064,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 6.604152180391378e-05,
+      "loss": 0.5115,
+      "step": 184720
+    },
+    {
+      "epoch": 9.175027316976259,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.600178802026423e-05,
+      "loss": 0.4634,
+      "step": 184730
+    },
+    {
+      "epoch": 9.175523989271879,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.596205423661469e-05,
+      "loss": 0.4893,
+      "step": 184740
+    },
+    {
+      "epoch": 9.176020661567497,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.592232045296513e-05,
+      "loss": 0.4928,
+      "step": 184750
+    },
+    {
+      "epoch": 9.176517333863117,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.588258666931559e-05,
+      "loss": 0.4845,
+      "step": 184760
+    },
+    {
+      "epoch": 9.177014006158737,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.584285288566603e-05,
+      "loss": 0.4834,
+      "step": 184770
+    },
+    {
+      "epoch": 9.177510678454356,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.580311910201649e-05,
+      "loss": 0.4973,
+      "step": 184780
+    },
+    {
+      "epoch": 9.178007350749976,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 6.576338531836695e-05,
+      "loss": 0.4579,
+      "step": 184790
+    },
+    {
+      "epoch": 9.178504023045594,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 6.57236515347174e-05,
+      "loss": 0.5235,
+      "step": 184800
+    },
+    {
+      "epoch": 9.179000695341214,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 6.568391775106785e-05,
+      "loss": 0.4999,
+      "step": 184810
+    },
+    {
+      "epoch": 9.179497367636833,
+      "grad_norm": 0.203125,
+      "learning_rate": 6.564418396741831e-05,
+      "loss": 0.4872,
+      "step": 184820
+    },
+    {
+      "epoch": 9.179994039932453,
+      "grad_norm": 0.234375,
+      "learning_rate": 6.560445018376875e-05,
+      "loss": 0.4969,
+      "step": 184830
+    },
+    {
+      "epoch": 9.180490712228073,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 6.556471640011921e-05,
+      "loss": 0.4938,
+      "step": 184840
+    },
+    {
+      "epoch": 9.180987384523691,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.552498261646966e-05,
+      "loss": 0.4687,
+      "step": 184850
+    },
+    {
+      "epoch": 9.181484056819311,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.548524883282011e-05,
+      "loss": 0.5009,
+      "step": 184860
+    },
+    {
+      "epoch": 9.18198072911493,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.544551504917056e-05,
+      "loss": 0.491,
+      "step": 184870
+    },
+    {
+      "epoch": 9.18247740141055,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.540578126552101e-05,
+      "loss": 0.4583,
+      "step": 184880
+    },
+    {
+      "epoch": 9.182974073706168,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.536604748187146e-05,
+      "loss": 0.4648,
+      "step": 184890
+    },
+    {
+      "epoch": 9.183470746001788,
+      "grad_norm": 0.240234375,
+      "learning_rate": 6.532631369822192e-05,
+      "loss": 0.5045,
+      "step": 184900
+    },
+    {
+      "epoch": 9.183967418297406,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.528657991457236e-05,
+      "loss": 0.4865,
+      "step": 184910
+    },
+    {
+      "epoch": 9.184464090593027,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.524684613092282e-05,
+      "loss": 0.4882,
+      "step": 184920
+    },
+    {
+      "epoch": 9.184960762888647,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.520711234727326e-05,
+      "loss": 0.4672,
+      "step": 184930
+    },
+    {
+      "epoch": 9.185457435184265,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 6.516737856362372e-05,
+      "loss": 0.4775,
+      "step": 184940
+    },
+    {
+      "epoch": 9.185954107479885,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.512764477997418e-05,
+      "loss": 0.4545,
+      "step": 184950
+    },
+    {
+      "epoch": 9.186450779775504,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 6.508791099632464e-05,
+      "loss": 0.4564,
+      "step": 184960
+    },
+    {
+      "epoch": 9.186947452071124,
+      "grad_norm": 0.2265625,
+      "learning_rate": 6.504817721267508e-05,
+      "loss": 0.4562,
+      "step": 184970
+    },
+    {
+      "epoch": 9.187444124366742,
+      "grad_norm": 0.197265625,
+      "learning_rate": 6.500844342902554e-05,
+      "loss": 0.4695,
+      "step": 184980
+    },
+    {
+      "epoch": 9.187940796662362,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.496870964537598e-05,
+      "loss": 0.4632,
+      "step": 184990
+    },
+    {
+      "epoch": 9.188437468957982,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.492897586172644e-05,
+      "loss": 0.4847,
+      "step": 185000
+    },
+    {
+      "epoch": 9.1889341412536,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 6.488924207807689e-05,
+      "loss": 0.5039,
+      "step": 185010
+    },
+    {
+      "epoch": 9.18943081354922,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.484950829442734e-05,
+      "loss": 0.4899,
+      "step": 185020
+    },
+    {
+      "epoch": 9.189927485844839,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.480977451077779e-05,
+      "loss": 0.4832,
+      "step": 185030
+    },
+    {
+      "epoch": 9.190424158140459,
+      "grad_norm": 0.201171875,
+      "learning_rate": 6.477004072712824e-05,
+      "loss": 0.5118,
+      "step": 185040
+    },
+    {
+      "epoch": 9.190920830436077,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.473030694347869e-05,
+      "loss": 0.4991,
+      "step": 185050
+    },
+    {
+      "epoch": 9.191417502731698,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.469057315982915e-05,
+      "loss": 0.4967,
+      "step": 185060
+    },
+    {
+      "epoch": 9.191914175027318,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.465083937617959e-05,
+      "loss": 0.5004,
+      "step": 185070
+    },
+    {
+      "epoch": 9.192410847322936,
+      "grad_norm": 0.22265625,
+      "learning_rate": 6.461110559253005e-05,
+      "loss": 0.4704,
+      "step": 185080
+    },
+    {
+      "epoch": 9.192907519618556,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 6.45713718088805e-05,
+      "loss": 0.4667,
+      "step": 185090
+    },
+    {
+      "epoch": 9.193404191914174,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.453163802523095e-05,
+      "loss": 0.4669,
+      "step": 185100
+    },
+    {
+      "epoch": 9.193900864209795,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.449190424158141e-05,
+      "loss": 0.4886,
+      "step": 185110
+    },
+    {
+      "epoch": 9.194397536505413,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 6.445217045793185e-05,
+      "loss": 0.4788,
+      "step": 185120
+    },
+    {
+      "epoch": 9.194894208801033,
+      "grad_norm": 0.2265625,
+      "learning_rate": 6.441243667428231e-05,
+      "loss": 0.4927,
+      "step": 185130
+    },
+    {
+      "epoch": 9.195390881096653,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.437270289063277e-05,
+      "loss": 0.4908,
+      "step": 185140
+    },
+    {
+      "epoch": 9.195887553392271,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.433296910698323e-05,
+      "loss": 0.4802,
+      "step": 185150
+    },
+    {
+      "epoch": 9.196384225687892,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 6.429323532333367e-05,
+      "loss": 0.4572,
+      "step": 185160
+    },
+    {
+      "epoch": 9.19688089798351,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.425350153968413e-05,
+      "loss": 0.4916,
+      "step": 185170
+    },
+    {
+      "epoch": 9.19737757027913,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.421376775603457e-05,
+      "loss": 0.4472,
+      "step": 185180
+    },
+    {
+      "epoch": 9.197874242574748,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 6.417403397238503e-05,
+      "loss": 0.4867,
+      "step": 185190
+    },
+    {
+      "epoch": 9.198370914870369,
+      "grad_norm": 0.212890625,
+      "learning_rate": 6.413430018873547e-05,
+      "loss": 0.4517,
+      "step": 185200
+    },
+    {
+      "epoch": 9.198867587165989,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.409456640508593e-05,
+      "loss": 0.4819,
+      "step": 185210
+    },
+    {
+      "epoch": 9.199364259461607,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.405483262143638e-05,
+      "loss": 0.4861,
+      "step": 185220
+    },
+    {
+      "epoch": 9.199860931757227,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.401509883778683e-05,
+      "loss": 0.4791,
+      "step": 185230
+    },
+    {
+      "epoch": 9.200357604052845,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.397536505413728e-05,
+      "loss": 0.5115,
+      "step": 185240
+    },
+    {
+      "epoch": 9.200854276348466,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.393563127048774e-05,
+      "loss": 0.5009,
+      "step": 185250
+    },
+    {
+      "epoch": 9.201350948644084,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.389589748683818e-05,
+      "loss": 0.4814,
+      "step": 185260
+    },
+    {
+      "epoch": 9.201847620939704,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.385616370318864e-05,
+      "loss": 0.4707,
+      "step": 185270
+    },
+    {
+      "epoch": 9.202344293235324,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.381642991953908e-05,
+      "loss": 0.4635,
+      "step": 185280
+    },
+    {
+      "epoch": 9.202840965530942,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 6.377669613588954e-05,
+      "loss": 0.446,
+      "step": 185290
+    },
+    {
+      "epoch": 9.203337637826563,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 6.373696235224e-05,
+      "loss": 0.5201,
+      "step": 185300
+    },
+    {
+      "epoch": 9.203834310122181,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 6.369722856859046e-05,
+      "loss": 0.5005,
+      "step": 185310
+    },
+    {
+      "epoch": 9.204330982417801,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 6.36574947849409e-05,
+      "loss": 0.475,
+      "step": 185320
+    },
+    {
+      "epoch": 9.20482765471342,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.361776100129136e-05,
+      "loss": 0.4586,
+      "step": 185330
+    },
+    {
+      "epoch": 9.20532432700904,
+      "grad_norm": 0.201171875,
+      "learning_rate": 6.35780272176418e-05,
+      "loss": 0.4828,
+      "step": 185340
+    },
+    {
+      "epoch": 9.20582099930466,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 6.353829343399226e-05,
+      "loss": 0.4793,
+      "step": 185350
+    },
+    {
+      "epoch": 9.206317671600278,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.34985596503427e-05,
+      "loss": 0.477,
+      "step": 185360
+    },
+    {
+      "epoch": 9.206814343895898,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.345882586669316e-05,
+      "loss": 0.4984,
+      "step": 185370
+    },
+    {
+      "epoch": 9.207311016191516,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 6.341909208304361e-05,
+      "loss": 0.4755,
+      "step": 185380
+    },
+    {
+      "epoch": 9.207807688487136,
+      "grad_norm": 0.208984375,
+      "learning_rate": 6.337935829939406e-05,
+      "loss": 0.5122,
+      "step": 185390
+    },
+    {
+      "epoch": 9.208304360782755,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.333962451574451e-05,
+      "loss": 0.4924,
+      "step": 185400
+    },
+    {
+      "epoch": 9.208801033078375,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.329989073209497e-05,
+      "loss": 0.4954,
+      "step": 185410
+    },
+    {
+      "epoch": 9.209297705373995,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.326015694844541e-05,
+      "loss": 0.469,
+      "step": 185420
+    },
+    {
+      "epoch": 9.209794377669613,
+      "grad_norm": 0.29296875,
+      "learning_rate": 6.322042316479587e-05,
+      "loss": 0.4755,
+      "step": 185430
+    },
+    {
+      "epoch": 9.210291049965234,
+      "grad_norm": 0.19921875,
+      "learning_rate": 6.318068938114631e-05,
+      "loss": 0.4923,
+      "step": 185440
+    },
+    {
+      "epoch": 9.210787722260852,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.314095559749677e-05,
+      "loss": 0.5002,
+      "step": 185450
+    },
+    {
+      "epoch": 9.211284394556472,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.310122181384723e-05,
+      "loss": 0.5146,
+      "step": 185460
+    },
+    {
+      "epoch": 9.21178106685209,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.306148803019767e-05,
+      "loss": 0.5041,
+      "step": 185470
+    },
+    {
+      "epoch": 9.21227773914771,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 6.302175424654813e-05,
+      "loss": 0.4766,
+      "step": 185480
+    },
+    {
+      "epoch": 9.21277441144333,
+      "grad_norm": 0.212890625,
+      "learning_rate": 6.298202046289859e-05,
+      "loss": 0.4516,
+      "step": 185490
+    },
+    {
+      "epoch": 9.213271083738949,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 6.294228667924903e-05,
+      "loss": 0.4993,
+      "step": 185500
+    },
+    {
+      "epoch": 9.213767756034569,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 6.290255289559949e-05,
+      "loss": 0.496,
+      "step": 185510
+    },
+    {
+      "epoch": 9.214264428330187,
+      "grad_norm": 0.203125,
+      "learning_rate": 6.286281911194993e-05,
+      "loss": 0.4954,
+      "step": 185520
+    },
+    {
+      "epoch": 9.214761100625807,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.282308532830039e-05,
+      "loss": 0.4816,
+      "step": 185530
+    },
+    {
+      "epoch": 9.215257772921426,
+      "grad_norm": 0.21484375,
+      "learning_rate": 6.278335154465084e-05,
+      "loss": 0.4819,
+      "step": 185540
+    },
+    {
+      "epoch": 9.215754445217046,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.27436177610013e-05,
+      "loss": 0.4628,
+      "step": 185550
+    },
+    {
+      "epoch": 9.216251117512666,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 6.270388397735175e-05,
+      "loss": 0.4623,
+      "step": 185560
+    },
+    {
+      "epoch": 9.216747789808284,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.26641501937022e-05,
+      "loss": 0.4967,
+      "step": 185570
+    },
+    {
+      "epoch": 9.217244462103904,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.262441641005265e-05,
+      "loss": 0.4698,
+      "step": 185580
+    },
+    {
+      "epoch": 9.217741134399523,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 6.25846826264031e-05,
+      "loss": 0.4738,
+      "step": 185590
+    },
+    {
+      "epoch": 9.218237806695143,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 6.254494884275356e-05,
+      "loss": 0.4926,
+      "step": 185600
+    },
+    {
+      "epoch": 9.218734478990761,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 6.2505215059104e-05,
+      "loss": 0.496,
+      "step": 185610
+    },
+    {
+      "epoch": 9.219231151286381,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.246548127545446e-05,
+      "loss": 0.4931,
+      "step": 185620
+    },
+    {
+      "epoch": 9.219727823582001,
+      "grad_norm": 0.224609375,
+      "learning_rate": 6.24257474918049e-05,
+      "loss": 0.5287,
+      "step": 185630
+    },
+    {
+      "epoch": 9.22022449587762,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 6.238601370815536e-05,
+      "loss": 0.5118,
+      "step": 185640
+    },
+    {
+      "epoch": 9.22072116817324,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.234627992450582e-05,
+      "loss": 0.4663,
+      "step": 185650
+    },
+    {
+      "epoch": 9.221217840468858,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.230654614085628e-05,
+      "loss": 0.5234,
+      "step": 185660
+    },
+    {
+      "epoch": 9.221714512764478,
+      "grad_norm": 0.2265625,
+      "learning_rate": 6.226681235720672e-05,
+      "loss": 0.4939,
+      "step": 185670
+    },
+    {
+      "epoch": 9.222211185060097,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 6.222707857355718e-05,
+      "loss": 0.499,
+      "step": 185680
+    },
+    {
+      "epoch": 9.222707857355717,
+      "grad_norm": 0.19921875,
+      "learning_rate": 6.218734478990762e-05,
+      "loss": 0.4977,
+      "step": 185690
+    },
+    {
+      "epoch": 9.223204529651335,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 6.214761100625808e-05,
+      "loss": 0.4894,
+      "step": 185700
+    },
+    {
+      "epoch": 9.223701201946955,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 6.210787722260852e-05,
+      "loss": 0.4917,
+      "step": 185710
+    },
+    {
+      "epoch": 9.224197874242575,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.206814343895898e-05,
+      "loss": 0.5219,
+      "step": 185720
+    },
+    {
+      "epoch": 9.224694546538194,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.202840965530943e-05,
+      "loss": 0.4857,
+      "step": 185730
+    },
+    {
+      "epoch": 9.225191218833814,
+      "grad_norm": 0.212890625,
+      "learning_rate": 6.198867587165988e-05,
+      "loss": 0.4869,
+      "step": 185740
+    },
+    {
+      "epoch": 9.225687891129432,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.194894208801033e-05,
+      "loss": 0.4897,
+      "step": 185750
+    },
+    {
+      "epoch": 9.226184563425052,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.190920830436079e-05,
+      "loss": 0.4721,
+      "step": 185760
+    },
+    {
+      "epoch": 9.22668123572067,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 6.186947452071123e-05,
+      "loss": 0.4598,
+      "step": 185770
+    },
+    {
+      "epoch": 9.22717790801629,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 6.182974073706169e-05,
+      "loss": 0.5066,
+      "step": 185780
+    },
+    {
+      "epoch": 9.227674580311911,
+      "grad_norm": 0.201171875,
+      "learning_rate": 6.179000695341213e-05,
+      "loss": 0.4993,
+      "step": 185790
+    },
+    {
+      "epoch": 9.22817125260753,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.175027316976259e-05,
+      "loss": 0.4807,
+      "step": 185800
+    },
+    {
+      "epoch": 9.22866792490315,
+      "grad_norm": 0.201171875,
+      "learning_rate": 6.171053938611305e-05,
+      "loss": 0.4626,
+      "step": 185810
+    },
+    {
+      "epoch": 9.229164597198768,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.167080560246349e-05,
+      "loss": 0.4757,
+      "step": 185820
+    },
+    {
+      "epoch": 9.229661269494388,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.163107181881395e-05,
+      "loss": 0.4612,
+      "step": 185830
+    },
+    {
+      "epoch": 9.230157941790006,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 6.159133803516441e-05,
+      "loss": 0.4929,
+      "step": 185840
+    },
+    {
+      "epoch": 9.230654614085626,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.155160425151485e-05,
+      "loss": 0.483,
+      "step": 185850
+    },
+    {
+      "epoch": 9.231151286381246,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.151187046786531e-05,
+      "loss": 0.4785,
+      "step": 185860
+    },
+    {
+      "epoch": 9.231647958676865,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.147213668421575e-05,
+      "loss": 0.4799,
+      "step": 185870
+    },
+    {
+      "epoch": 9.232144630972485,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 6.143240290056621e-05,
+      "loss": 0.49,
+      "step": 185880
+    },
+    {
+      "epoch": 9.232641303268103,
+      "grad_norm": 0.197265625,
+      "learning_rate": 6.139266911691666e-05,
+      "loss": 0.4554,
+      "step": 185890
+    },
+    {
+      "epoch": 9.233137975563723,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.135293533326711e-05,
+      "loss": 0.4884,
+      "step": 185900
+    },
+    {
+      "epoch": 9.233634647859342,
+      "grad_norm": 0.19921875,
+      "learning_rate": 6.131320154961756e-05,
+      "loss": 0.4668,
+      "step": 185910
+    },
+    {
+      "epoch": 9.234131320154962,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.127346776596802e-05,
+      "loss": 0.4803,
+      "step": 185920
+    },
+    {
+      "epoch": 9.234627992450582,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.123373398231846e-05,
+      "loss": 0.5186,
+      "step": 185930
+    },
+    {
+      "epoch": 9.2351246647462,
+      "grad_norm": 0.244140625,
+      "learning_rate": 6.119400019866892e-05,
+      "loss": 0.4741,
+      "step": 185940
+    },
+    {
+      "epoch": 9.23562133704182,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 6.115426641501936e-05,
+      "loss": 0.4883,
+      "step": 185950
+    },
+    {
+      "epoch": 9.236118009337439,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.111453263136982e-05,
+      "loss": 0.4836,
+      "step": 185960
+    },
+    {
+      "epoch": 9.236614681633059,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.107479884772028e-05,
+      "loss": 0.4724,
+      "step": 185970
+    },
+    {
+      "epoch": 9.237111353928677,
+      "grad_norm": 0.216796875,
+      "learning_rate": 6.103506506407073e-05,
+      "loss": 0.5036,
+      "step": 185980
+    },
+    {
+      "epoch": 9.237608026224297,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 6.099533128042119e-05,
+      "loss": 0.4576,
+      "step": 185990
+    },
+    {
+      "epoch": 9.238104698519917,
+      "grad_norm": 0.232421875,
+      "learning_rate": 6.095559749677163e-05,
+      "loss": 0.4848,
+      "step": 186000
+    },
+    {
+      "epoch": 9.238601370815536,
+      "grad_norm": 0.20703125,
+      "learning_rate": 6.091586371312209e-05,
+      "loss": 0.4708,
+      "step": 186010
+    },
+    {
+      "epoch": 9.239098043111156,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.0876129929472534e-05,
+      "loss": 0.4908,
+      "step": 186020
+    },
+    {
+      "epoch": 9.239594715406774,
+      "grad_norm": 0.203125,
+      "learning_rate": 6.083639614582299e-05,
+      "loss": 0.5043,
+      "step": 186030
+    },
+    {
+      "epoch": 9.240091387702394,
+      "grad_norm": 0.216796875,
+      "learning_rate": 6.079666236217344e-05,
+      "loss": 0.4702,
+      "step": 186040
+    },
+    {
+      "epoch": 9.240588059998013,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.07569285785239e-05,
+      "loss": 0.4784,
+      "step": 186050
+    },
+    {
+      "epoch": 9.241084732293633,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 6.0717194794874345e-05,
+      "loss": 0.485,
+      "step": 186060
+    },
+    {
+      "epoch": 9.241581404589253,
+      "grad_norm": 0.25390625,
+      "learning_rate": 6.06774610112248e-05,
+      "loss": 0.4819,
+      "step": 186070
+    },
+    {
+      "epoch": 9.242078076884871,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.063772722757525e-05,
+      "loss": 0.4849,
+      "step": 186080
+    },
+    {
+      "epoch": 9.242574749180491,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.0597993443925704e-05,
+      "loss": 0.5043,
+      "step": 186090
+    },
+    {
+      "epoch": 9.24307142147611,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.055825966027615e-05,
+      "loss": 0.4819,
+      "step": 186100
+    },
+    {
+      "epoch": 9.24356809377173,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 6.0518525876626606e-05,
+      "loss": 0.4952,
+      "step": 186110
+    },
+    {
+      "epoch": 9.244064766067348,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 6.047879209297706e-05,
+      "loss": 0.4589,
+      "step": 186120
+    },
+    {
+      "epoch": 9.244561438362968,
+      "grad_norm": 0.23046875,
+      "learning_rate": 6.0439058309327515e-05,
+      "loss": 0.5142,
+      "step": 186130
+    },
+    {
+      "epoch": 9.245058110658588,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 6.039932452567796e-05,
+      "loss": 0.4881,
+      "step": 186140
+    },
+    {
+      "epoch": 9.245554782954207,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 6.035959074202842e-05,
+      "loss": 0.4849,
+      "step": 186150
+    },
+    {
+      "epoch": 9.246051455249827,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 6.031985695837886e-05,
+      "loss": 0.4902,
+      "step": 186160
+    },
+    {
+      "epoch": 9.246548127545445,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.028012317472932e-05,
+      "loss": 0.5227,
+      "step": 186170
+    },
+    {
+      "epoch": 9.247044799841065,
+      "grad_norm": 0.203125,
+      "learning_rate": 6.0240389391079764e-05,
+      "loss": 0.5197,
+      "step": 186180
+    },
+    {
+      "epoch": 9.247541472136684,
+      "grad_norm": 0.21875,
+      "learning_rate": 6.020065560743022e-05,
+      "loss": 0.5058,
+      "step": 186190
+    },
+    {
+      "epoch": 9.248038144432304,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.016092182378067e-05,
+      "loss": 0.4697,
+      "step": 186200
+    },
+    {
+      "epoch": 9.248534816727924,
+      "grad_norm": 0.21484375,
+      "learning_rate": 6.0121188040131124e-05,
+      "loss": 0.5106,
+      "step": 186210
+    },
+    {
+      "epoch": 9.249031489023542,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 6.0081454256481575e-05,
+      "loss": 0.4796,
+      "step": 186220
+    },
+    {
+      "epoch": 9.249528161319162,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.004172047283203e-05,
+      "loss": 0.4907,
+      "step": 186230
+    },
+    {
+      "epoch": 9.25002483361478,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.0001986689182477e-05,
+      "loss": 0.4646,
+      "step": 186240
+    },
+    {
+      "epoch": 9.2505215059104,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.9962252905532934e-05,
+      "loss": 0.4674,
+      "step": 186250
+    },
+    {
+      "epoch": 9.251018178206019,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 5.992251912188338e-05,
+      "loss": 0.4699,
+      "step": 186260
+    },
+    {
+      "epoch": 9.25151485050164,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 5.9882785338233836e-05,
+      "loss": 0.4921,
+      "step": 186270
+    },
+    {
+      "epoch": 9.252011522797257,
+      "grad_norm": 0.20703125,
+      "learning_rate": 5.984305155458428e-05,
+      "loss": 0.5019,
+      "step": 186280
+    },
+    {
+      "epoch": 9.252508195092878,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.980331777093474e-05,
+      "loss": 0.4835,
+      "step": 186290
+    },
+    {
+      "epoch": 9.253004867388498,
+      "grad_norm": 0.19921875,
+      "learning_rate": 5.976358398728519e-05,
+      "loss": 0.5084,
+      "step": 186300
+    },
+    {
+      "epoch": 9.253501539684116,
+      "grad_norm": 0.232421875,
+      "learning_rate": 5.972385020363565e-05,
+      "loss": 0.5069,
+      "step": 186310
+    },
+    {
+      "epoch": 9.253998211979736,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 5.968411641998609e-05,
+      "loss": 0.4888,
+      "step": 186320
+    },
+    {
+      "epoch": 9.254494884275354,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 5.964438263633655e-05,
+      "loss": 0.4543,
+      "step": 186330
+    },
+    {
+      "epoch": 9.254991556570975,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 5.9604648852686994e-05,
+      "loss": 0.4822,
+      "step": 186340
+    },
+    {
+      "epoch": 9.255488228866593,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 5.956491506903745e-05,
+      "loss": 0.4528,
+      "step": 186350
+    },
+    {
+      "epoch": 9.255984901162213,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 5.9525181285387896e-05,
+      "loss": 0.5025,
+      "step": 186360
+    },
+    {
+      "epoch": 9.256481573457833,
+      "grad_norm": 0.18359375,
+      "learning_rate": 5.9485447501738353e-05,
+      "loss": 0.4753,
+      "step": 186370
+    },
+    {
+      "epoch": 9.256978245753452,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 5.9445713718088805e-05,
+      "loss": 0.4525,
+      "step": 186380
+    },
+    {
+      "epoch": 9.257474918049072,
+      "grad_norm": 0.212890625,
+      "learning_rate": 5.940597993443926e-05,
+      "loss": 0.4734,
+      "step": 186390
+    },
+    {
+      "epoch": 9.25797159034469,
+      "grad_norm": 0.18359375,
+      "learning_rate": 5.936624615078972e-05,
+      "loss": 0.467,
+      "step": 186400
+    },
+    {
+      "epoch": 9.25846826264031,
+      "grad_norm": 0.23046875,
+      "learning_rate": 5.9326512367140164e-05,
+      "loss": 0.5307,
+      "step": 186410
+    },
+    {
+      "epoch": 9.258964934935928,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5.928677858349062e-05,
+      "loss": 0.492,
+      "step": 186420
+    },
+    {
+      "epoch": 9.259461607231549,
+      "grad_norm": 0.20703125,
+      "learning_rate": 5.9247044799841066e-05,
+      "loss": 0.4876,
+      "step": 186430
+    },
+    {
+      "epoch": 9.259958279527169,
+      "grad_norm": 0.181640625,
+      "learning_rate": 5.9207311016191524e-05,
+      "loss": 0.4827,
+      "step": 186440
+    },
+    {
+      "epoch": 9.260454951822787,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5.916757723254197e-05,
+      "loss": 0.5002,
+      "step": 186450
+    },
+    {
+      "epoch": 9.260951624118407,
+      "grad_norm": 0.1953125,
+      "learning_rate": 5.9127843448892426e-05,
+      "loss": 0.4513,
+      "step": 186460
+    },
+    {
+      "epoch": 9.261448296414025,
+      "grad_norm": 0.23046875,
+      "learning_rate": 5.908810966524288e-05,
+      "loss": 0.5047,
+      "step": 186470
+    },
+    {
+      "epoch": 9.261944968709646,
+      "grad_norm": 0.1953125,
+      "learning_rate": 5.9048375881593335e-05,
+      "loss": 0.4695,
+      "step": 186480
+    },
+    {
+      "epoch": 9.262441641005264,
+      "grad_norm": 0.205078125,
+      "learning_rate": 5.900864209794378e-05,
+      "loss": 0.4897,
+      "step": 186490
+    },
+    {
+      "epoch": 9.262938313300884,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 5.896890831429424e-05,
+      "loss": 0.4734,
+      "step": 186500
+    },
+    {
+      "epoch": 9.263434985596504,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 5.892917453064468e-05,
+      "loss": 0.4814,
+      "step": 186510
+    },
+    {
+      "epoch": 9.263931657892122,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.888944074699514e-05,
+      "loss": 0.5046,
+      "step": 186520
+    },
+    {
+      "epoch": 9.264428330187743,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 5.8849706963345583e-05,
+      "loss": 0.4948,
+      "step": 186530
+    },
+    {
+      "epoch": 9.264925002483361,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 5.880997317969604e-05,
+      "loss": 0.5241,
+      "step": 186540
+    },
+    {
+      "epoch": 9.265421674778981,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 5.877023939604649e-05,
+      "loss": 0.5078,
+      "step": 186550
+    },
+    {
+      "epoch": 9.2659183470746,
+      "grad_norm": 0.212890625,
+      "learning_rate": 5.873050561239694e-05,
+      "loss": 0.4903,
+      "step": 186560
+    },
+    {
+      "epoch": 9.26641501937022,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.8690771828747394e-05,
+      "loss": 0.4595,
+      "step": 186570
+    },
+    {
+      "epoch": 9.26691169166584,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 5.865103804509785e-05,
+      "loss": 0.4921,
+      "step": 186580
+    },
+    {
+      "epoch": 9.267408363961458,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 5.8611304261448296e-05,
+      "loss": 0.4667,
+      "step": 186590
+    },
+    {
+      "epoch": 9.267905036257078,
+      "grad_norm": 0.21484375,
+      "learning_rate": 5.8571570477798754e-05,
+      "loss": 0.4944,
+      "step": 186600
+    },
+    {
+      "epoch": 9.268401708552696,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 5.85318366941492e-05,
+      "loss": 0.4866,
+      "step": 186610
+    },
+    {
+      "epoch": 9.268898380848317,
+      "grad_norm": 0.23046875,
+      "learning_rate": 5.8492102910499656e-05,
+      "loss": 0.454,
+      "step": 186620
+    },
+    {
+      "epoch": 9.269395053143935,
+      "grad_norm": 0.220703125,
+      "learning_rate": 5.84523691268501e-05,
+      "loss": 0.4836,
+      "step": 186630
+    },
+    {
+      "epoch": 9.269891725439555,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 5.841263534320056e-05,
+      "loss": 0.4713,
+      "step": 186640
+    },
+    {
+      "epoch": 9.270388397735175,
+      "grad_norm": 0.224609375,
+      "learning_rate": 5.837290155955101e-05,
+      "loss": 0.5056,
+      "step": 186650
+    },
+    {
+      "epoch": 9.270885070030793,
+      "grad_norm": 0.193359375,
+      "learning_rate": 5.833316777590147e-05,
+      "loss": 0.4788,
+      "step": 186660
+    },
+    {
+      "epoch": 9.271381742326414,
+      "grad_norm": 0.216796875,
+      "learning_rate": 5.829343399225191e-05,
+      "loss": 0.4942,
+      "step": 186670
+    },
+    {
+      "epoch": 9.271878414622032,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 5.825370020860237e-05,
+      "loss": 0.4962,
+      "step": 186680
+    },
+    {
+      "epoch": 9.272375086917652,
+      "grad_norm": 0.259765625,
+      "learning_rate": 5.8213966424952813e-05,
+      "loss": 0.5012,
+      "step": 186690
+    },
+    {
+      "epoch": 9.27287175921327,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 5.817423264130327e-05,
+      "loss": 0.4848,
+      "step": 186700
+    },
+    {
+      "epoch": 9.27336843150889,
+      "grad_norm": 0.1796875,
+      "learning_rate": 5.8134498857653716e-05,
+      "loss": 0.5006,
+      "step": 186710
+    },
+    {
+      "epoch": 9.27386510380451,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 5.809476507400417e-05,
+      "loss": 0.4955,
+      "step": 186720
+    },
+    {
+      "epoch": 9.274361776100129,
+      "grad_norm": 0.205078125,
+      "learning_rate": 5.8055031290354624e-05,
+      "loss": 0.4762,
+      "step": 186730
+    },
+    {
+      "epoch": 9.274858448395749,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.801529750670508e-05,
+      "loss": 0.5008,
+      "step": 186740
+    },
+    {
+      "epoch": 9.275355120691367,
+      "grad_norm": 0.224609375,
+      "learning_rate": 5.7975563723055526e-05,
+      "loss": 0.501,
+      "step": 186750
+    },
+    {
+      "epoch": 9.275851792986987,
+      "grad_norm": 0.1796875,
+      "learning_rate": 5.7935829939405984e-05,
+      "loss": 0.4476,
+      "step": 186760
+    },
+    {
+      "epoch": 9.276348465282606,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 5.789609615575643e-05,
+      "loss": 0.4489,
+      "step": 186770
+    },
+    {
+      "epoch": 9.276845137578226,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 5.7856362372106886e-05,
+      "loss": 0.4733,
+      "step": 186780
+    },
+    {
+      "epoch": 9.277341809873846,
+      "grad_norm": 0.23046875,
+      "learning_rate": 5.781662858845733e-05,
+      "loss": 0.5189,
+      "step": 186790
+    },
+    {
+      "epoch": 9.277838482169464,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 5.777689480480779e-05,
+      "loss": 0.5196,
+      "step": 186800
+    },
+    {
+      "epoch": 9.278335154465084,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.7737161021158246e-05,
+      "loss": 0.4705,
+      "step": 186810
+    },
+    {
+      "epoch": 9.278831826760703,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 5.76974272375087e-05,
+      "loss": 0.4956,
+      "step": 186820
+    },
+    {
+      "epoch": 9.279328499056323,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 5.7657693453859155e-05,
+      "loss": 0.5072,
+      "step": 186830
+    },
+    {
+      "epoch": 9.279825171351941,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.76179596702096e-05,
+      "loss": 0.4981,
+      "step": 186840
+    },
+    {
+      "epoch": 9.280321843647561,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.757822588656006e-05,
+      "loss": 0.47,
+      "step": 186850
+    },
+    {
+      "epoch": 9.280818515943182,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 5.75384921029105e-05,
+      "loss": 0.5055,
+      "step": 186860
+    },
+    {
+      "epoch": 9.2813151882388,
+      "grad_norm": 0.19921875,
+      "learning_rate": 5.749875831926096e-05,
+      "loss": 0.4947,
+      "step": 186870
+    },
+    {
+      "epoch": 9.28181186053442,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 5.74590245356114e-05,
+      "loss": 0.5082,
+      "step": 186880
+    },
+    {
+      "epoch": 9.282308532830038,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5.741929075196186e-05,
+      "loss": 0.4985,
+      "step": 186890
+    },
+    {
+      "epoch": 9.282805205125658,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.737955696831231e-05,
+      "loss": 0.4571,
+      "step": 186900
+    },
+    {
+      "epoch": 9.283301877421277,
+      "grad_norm": 0.1796875,
+      "learning_rate": 5.733982318466276e-05,
+      "loss": 0.4831,
+      "step": 186910
+    },
+    {
+      "epoch": 9.283798549716897,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 5.7300089401013214e-05,
+      "loss": 0.4659,
+      "step": 186920
+    },
+    {
+      "epoch": 9.284295222012517,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.726035561736367e-05,
+      "loss": 0.4556,
+      "step": 186930
+    },
+    {
+      "epoch": 9.284791894308135,
+      "grad_norm": 0.236328125,
+      "learning_rate": 5.7220621833714116e-05,
+      "loss": 0.4494,
+      "step": 186940
+    },
+    {
+      "epoch": 9.285288566603755,
+      "grad_norm": 0.1953125,
+      "learning_rate": 5.7180888050064574e-05,
+      "loss": 0.4987,
+      "step": 186950
+    },
+    {
+      "epoch": 9.285785238899374,
+      "grad_norm": 0.19921875,
+      "learning_rate": 5.714115426641502e-05,
+      "loss": 0.4772,
+      "step": 186960
+    },
+    {
+      "epoch": 9.286281911194994,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 5.7101420482765476e-05,
+      "loss": 0.4859,
+      "step": 186970
+    },
+    {
+      "epoch": 9.286778583490612,
+      "grad_norm": 0.20703125,
+      "learning_rate": 5.706168669911593e-05,
+      "loss": 0.5106,
+      "step": 186980
+    },
+    {
+      "epoch": 9.287275255786232,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.702195291546638e-05,
+      "loss": 0.4717,
+      "step": 186990
+    },
+    {
+      "epoch": 9.287771928081852,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 5.698221913181683e-05,
+      "loss": 0.4641,
+      "step": 187000
+    },
+    {
+      "epoch": 9.28826860037747,
+      "grad_norm": 0.193359375,
+      "learning_rate": 5.694248534816729e-05,
+      "loss": 0.5042,
+      "step": 187010
+    },
+    {
+      "epoch": 9.288765272673091,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 5.690275156451773e-05,
+      "loss": 0.4724,
+      "step": 187020
+    },
+    {
+      "epoch": 9.28926194496871,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5.686301778086819e-05,
+      "loss": 0.4697,
+      "step": 187030
+    },
+    {
+      "epoch": 9.28975861726433,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 5.682328399721863e-05,
+      "loss": 0.4926,
+      "step": 187040
+    },
+    {
+      "epoch": 9.290255289559948,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 5.678355021356909e-05,
+      "loss": 0.4981,
+      "step": 187050
+    },
+    {
+      "epoch": 9.290751961855568,
+      "grad_norm": 0.216796875,
+      "learning_rate": 5.6743816429919535e-05,
+      "loss": 0.4962,
+      "step": 187060
+    },
+    {
+      "epoch": 9.291248634151188,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 5.670408264626999e-05,
+      "loss": 0.4751,
+      "step": 187070
+    },
+    {
+      "epoch": 9.291745306446806,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 5.6664348862620444e-05,
+      "loss": 0.489,
+      "step": 187080
+    },
+    {
+      "epoch": 9.292241978742426,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 5.66246150789709e-05,
+      "loss": 0.4979,
+      "step": 187090
+    },
+    {
+      "epoch": 9.292738651038045,
+      "grad_norm": 0.19140625,
+      "learning_rate": 5.6584881295321346e-05,
+      "loss": 0.4927,
+      "step": 187100
+    },
+    {
+      "epoch": 9.293235323333665,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.6545147511671804e-05,
+      "loss": 0.489,
+      "step": 187110
+    },
+    {
+      "epoch": 9.293731995629283,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 5.650541372802225e-05,
+      "loss": 0.4744,
+      "step": 187120
+    },
+    {
+      "epoch": 9.294228667924903,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 5.6465679944372706e-05,
+      "loss": 0.5103,
+      "step": 187130
+    },
+    {
+      "epoch": 9.294725340220522,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 5.642594616072315e-05,
+      "loss": 0.493,
+      "step": 187140
+    },
+    {
+      "epoch": 9.295222012516142,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.638621237707361e-05,
+      "loss": 0.4323,
+      "step": 187150
+    },
+    {
+      "epoch": 9.295718684811762,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 5.634647859342406e-05,
+      "loss": 0.4903,
+      "step": 187160
+    },
+    {
+      "epoch": 9.29621535710738,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 5.630674480977452e-05,
+      "loss": 0.4894,
+      "step": 187170
+    },
+    {
+      "epoch": 9.296712029403,
+      "grad_norm": 0.208984375,
+      "learning_rate": 5.626701102612496e-05,
+      "loss": 0.5074,
+      "step": 187180
+    },
+    {
+      "epoch": 9.297208701698619,
+      "grad_norm": 0.20703125,
+      "learning_rate": 5.622727724247542e-05,
+      "loss": 0.5152,
+      "step": 187190
+    },
+    {
+      "epoch": 9.297705373994239,
+      "grad_norm": 0.205078125,
+      "learning_rate": 5.618754345882586e-05,
+      "loss": 0.4853,
+      "step": 187200
+    },
+    {
+      "epoch": 9.298202046289857,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 5.614780967517632e-05,
+      "loss": 0.5112,
+      "step": 187210
+    },
+    {
+      "epoch": 9.298698718585477,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 5.6108075891526765e-05,
+      "loss": 0.5068,
+      "step": 187220
+    },
+    {
+      "epoch": 9.299195390881097,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 5.606834210787722e-05,
+      "loss": 0.4822,
+      "step": 187230
+    },
+    {
+      "epoch": 9.299692063176716,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.602860832422768e-05,
+      "loss": 0.4783,
+      "step": 187240
+    },
+    {
+      "epoch": 9.300188735472336,
+      "grad_norm": 0.18359375,
+      "learning_rate": 5.598887454057813e-05,
+      "loss": 0.4732,
+      "step": 187250
+    },
+    {
+      "epoch": 9.300685407767954,
+      "grad_norm": 0.20703125,
+      "learning_rate": 5.594914075692858e-05,
+      "loss": 0.4715,
+      "step": 187260
+    },
+    {
+      "epoch": 9.301182080063574,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.5909406973279034e-05,
+      "loss": 0.4939,
+      "step": 187270
+    },
+    {
+      "epoch": 9.301678752359193,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5.586967318962949e-05,
+      "loss": 0.4863,
+      "step": 187280
+    },
+    {
+      "epoch": 9.302175424654813,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.5829939405979936e-05,
+      "loss": 0.4837,
+      "step": 187290
+    },
+    {
+      "epoch": 9.302672096950433,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5.5790205622330394e-05,
+      "loss": 0.4811,
+      "step": 187300
+    },
+    {
+      "epoch": 9.303168769246051,
+      "grad_norm": 0.185546875,
+      "learning_rate": 5.575047183868084e-05,
+      "loss": 0.4762,
+      "step": 187310
+    },
+    {
+      "epoch": 9.303665441541671,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 5.5710738055031296e-05,
+      "loss": 0.511,
+      "step": 187320
+    },
+    {
+      "epoch": 9.30416211383729,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 5.567100427138175e-05,
+      "loss": 0.4655,
+      "step": 187330
+    },
+    {
+      "epoch": 9.30465878613291,
+      "grad_norm": 0.19921875,
+      "learning_rate": 5.56312704877322e-05,
+      "loss": 0.4847,
+      "step": 187340
+    },
+    {
+      "epoch": 9.305155458428528,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 5.559153670408265e-05,
+      "loss": 0.486,
+      "step": 187350
+    },
+    {
+      "epoch": 9.305652130724148,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 5.555180292043311e-05,
+      "loss": 0.4938,
+      "step": 187360
+    },
+    {
+      "epoch": 9.306148803019768,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 5.551206913678355e-05,
+      "loss": 0.4912,
+      "step": 187370
+    },
+    {
+      "epoch": 9.306645475315387,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 5.547233535313401e-05,
+      "loss": 0.5352,
+      "step": 187380
+    },
+    {
+      "epoch": 9.307142147611007,
+      "grad_norm": 0.203125,
+      "learning_rate": 5.543260156948445e-05,
+      "loss": 0.4706,
+      "step": 187390
+    },
+    {
+      "epoch": 9.307638819906625,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 5.539286778583491e-05,
+      "loss": 0.4798,
+      "step": 187400
+    },
+    {
+      "epoch": 9.308135492202245,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 5.5353134002185355e-05,
+      "loss": 0.4745,
+      "step": 187410
+    },
+    {
+      "epoch": 9.308632164497864,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 5.531340021853581e-05,
+      "loss": 0.4589,
+      "step": 187420
+    },
+    {
+      "epoch": 9.309128836793484,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 5.5273666434886264e-05,
+      "loss": 0.4995,
+      "step": 187430
+    },
+    {
+      "epoch": 9.309625509089104,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 5.523393265123672e-05,
+      "loss": 0.5002,
+      "step": 187440
+    },
+    {
+      "epoch": 9.310122181384722,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 5.5194198867587166e-05,
+      "loss": 0.4806,
+      "step": 187450
+    },
+    {
+      "epoch": 9.310618853680342,
+      "grad_norm": 0.22265625,
+      "learning_rate": 5.5154465083937624e-05,
+      "loss": 0.4971,
+      "step": 187460
+    },
+    {
+      "epoch": 9.31111552597596,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 5.511473130028807e-05,
+      "loss": 0.4762,
+      "step": 187470
+    },
+    {
+      "epoch": 9.31161219827158,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 5.5074997516638526e-05,
+      "loss": 0.5103,
+      "step": 187480
+    },
+    {
+      "epoch": 9.312108870567199,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 5.503526373298897e-05,
+      "loss": 0.498,
+      "step": 187490
+    },
+    {
+      "epoch": 9.31260554286282,
+      "grad_norm": 0.22265625,
+      "learning_rate": 5.499552994933943e-05,
+      "loss": 0.4498,
+      "step": 187500
+    },
+    {
+      "epoch": 9.31310221515844,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 5.495579616568988e-05,
+      "loss": 0.5271,
+      "step": 187510
+    },
+    {
+      "epoch": 9.313598887454058,
+      "grad_norm": 0.22265625,
+      "learning_rate": 5.491606238204034e-05,
+      "loss": 0.523,
+      "step": 187520
+    },
+    {
+      "epoch": 9.314095559749678,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 5.487632859839078e-05,
+      "loss": 0.4399,
+      "step": 187530
+    },
+    {
+      "epoch": 9.314592232045296,
+      "grad_norm": 0.19140625,
+      "learning_rate": 5.483659481474124e-05,
+      "loss": 0.4728,
+      "step": 187540
+    },
+    {
+      "epoch": 9.315088904340916,
+      "grad_norm": 0.232421875,
+      "learning_rate": 5.479686103109168e-05,
+      "loss": 0.4888,
+      "step": 187550
+    },
+    {
+      "epoch": 9.315585576636535,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 5.475712724744214e-05,
+      "loss": 0.4777,
+      "step": 187560
+    },
+    {
+      "epoch": 9.316082248932155,
+      "grad_norm": 0.234375,
+      "learning_rate": 5.4717393463792585e-05,
+      "loss": 0.5027,
+      "step": 187570
+    },
+    {
+      "epoch": 9.316578921227775,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 5.467765968014304e-05,
+      "loss": 0.4625,
+      "step": 187580
+    },
+    {
+      "epoch": 9.317075593523393,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.4637925896493494e-05,
+      "loss": 0.4635,
+      "step": 187590
+    },
+    {
+      "epoch": 9.317572265819013,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.459819211284395e-05,
+      "loss": 0.4702,
+      "step": 187600
+    },
+    {
+      "epoch": 9.318068938114632,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 5.4558458329194396e-05,
+      "loss": 0.4579,
+      "step": 187610
+    },
+    {
+      "epoch": 9.318565610410252,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 5.4518724545544854e-05,
+      "loss": 0.504,
+      "step": 187620
+    },
+    {
+      "epoch": 9.31906228270587,
+      "grad_norm": 0.240234375,
+      "learning_rate": 5.44789907618953e-05,
+      "loss": 0.4948,
+      "step": 187630
+    },
+    {
+      "epoch": 9.31955895500149,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 5.4439256978245756e-05,
+      "loss": 0.4738,
+      "step": 187640
+    },
+    {
+      "epoch": 9.320055627297108,
+      "grad_norm": 0.220703125,
+      "learning_rate": 5.43995231945962e-05,
+      "loss": 0.4768,
+      "step": 187650
+    },
+    {
+      "epoch": 9.320552299592729,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 5.435978941094666e-05,
+      "loss": 0.5013,
+      "step": 187660
+    },
+    {
+      "epoch": 9.321048971888349,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 5.4320055627297116e-05,
+      "loss": 0.4793,
+      "step": 187670
+    },
+    {
+      "epoch": 9.321545644183967,
+      "grad_norm": 0.185546875,
+      "learning_rate": 5.428032184364757e-05,
+      "loss": 0.4856,
+      "step": 187680
+    },
+    {
+      "epoch": 9.322042316479587,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 5.424058805999802e-05,
+      "loss": 0.4884,
+      "step": 187690
+    },
+    {
+      "epoch": 9.322538988775205,
+      "grad_norm": 0.201171875,
+      "learning_rate": 5.420085427634847e-05,
+      "loss": 0.5033,
+      "step": 187700
+    },
+    {
+      "epoch": 9.323035661070826,
+      "grad_norm": 0.20703125,
+      "learning_rate": 5.416112049269893e-05,
+      "loss": 0.4459,
+      "step": 187710
+    },
+    {
+      "epoch": 9.323532333366444,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 5.412138670904937e-05,
+      "loss": 0.4816,
+      "step": 187720
+    },
+    {
+      "epoch": 9.324029005662064,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5.408165292539983e-05,
+      "loss": 0.4857,
+      "step": 187730
+    },
+    {
+      "epoch": 9.324525677957684,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 5.404191914175027e-05,
+      "loss": 0.4951,
+      "step": 187740
+    },
+    {
+      "epoch": 9.325022350253303,
+      "grad_norm": 0.189453125,
+      "learning_rate": 5.400218535810073e-05,
+      "loss": 0.4738,
+      "step": 187750
+    },
+    {
+      "epoch": 9.325519022548923,
+      "grad_norm": 0.185546875,
+      "learning_rate": 5.3962451574451175e-05,
+      "loss": 0.4828,
+      "step": 187760
+    },
+    {
+      "epoch": 9.326015694844541,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 5.392271779080163e-05,
+      "loss": 0.4806,
+      "step": 187770
+    },
+    {
+      "epoch": 9.326512367140161,
+      "grad_norm": 0.19140625,
+      "learning_rate": 5.3882984007152084e-05,
+      "loss": 0.5165,
+      "step": 187780
+    },
+    {
+      "epoch": 9.32700903943578,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 5.384325022350254e-05,
+      "loss": 0.5057,
+      "step": 187790
+    },
+    {
+      "epoch": 9.3275057117314,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 5.3803516439852986e-05,
+      "loss": 0.4697,
+      "step": 187800
+    },
+    {
+      "epoch": 9.32800238402702,
+      "grad_norm": 0.201171875,
+      "learning_rate": 5.3763782656203444e-05,
+      "loss": 0.4986,
+      "step": 187810
+    },
+    {
+      "epoch": 9.328499056322638,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 5.372404887255389e-05,
+      "loss": 0.485,
+      "step": 187820
+    },
+    {
+      "epoch": 9.328995728618258,
+      "grad_norm": 0.18359375,
+      "learning_rate": 5.3684315088904346e-05,
+      "loss": 0.4799,
+      "step": 187830
+    },
+    {
+      "epoch": 9.329492400913876,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 5.364458130525479e-05,
+      "loss": 0.5037,
+      "step": 187840
+    },
+    {
+      "epoch": 9.329989073209497,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.360484752160525e-05,
+      "loss": 0.4841,
+      "step": 187850
+    },
+    {
+      "epoch": 9.330485745505115,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 5.35651137379557e-05,
+      "loss": 0.4872,
+      "step": 187860
+    },
+    {
+      "epoch": 9.330982417800735,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 5.3525379954306157e-05,
+      "loss": 0.4738,
+      "step": 187870
+    },
+    {
+      "epoch": 9.331479090096355,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 5.34856461706566e-05,
+      "loss": 0.5264,
+      "step": 187880
+    },
+    {
+      "epoch": 9.331975762391973,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 5.344591238700706e-05,
+      "loss": 0.4613,
+      "step": 187890
+    },
+    {
+      "epoch": 9.332472434687594,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 5.34061786033575e-05,
+      "loss": 0.4727,
+      "step": 187900
+    },
+    {
+      "epoch": 9.332969106983212,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 5.336644481970796e-05,
+      "loss": 0.5011,
+      "step": 187910
+    },
+    {
+      "epoch": 9.333465779278832,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.3326711036058405e-05,
+      "loss": 0.4775,
+      "step": 187920
+    },
+    {
+      "epoch": 9.33396245157445,
+      "grad_norm": 0.203125,
+      "learning_rate": 5.328697725240886e-05,
+      "loss": 0.4799,
+      "step": 187930
+    },
+    {
+      "epoch": 9.33445912387007,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.3247243468759314e-05,
+      "loss": 0.4657,
+      "step": 187940
+    },
+    {
+      "epoch": 9.33495579616569,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.320750968510977e-05,
+      "loss": 0.5079,
+      "step": 187950
+    },
+    {
+      "epoch": 9.335452468461309,
+      "grad_norm": 0.201171875,
+      "learning_rate": 5.3167775901460216e-05,
+      "loss": 0.5437,
+      "step": 187960
+    },
+    {
+      "epoch": 9.335949140756929,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.3128042117810674e-05,
+      "loss": 0.4687,
+      "step": 187970
+    },
+    {
+      "epoch": 9.336445813052547,
+      "grad_norm": 0.212890625,
+      "learning_rate": 5.308830833416112e-05,
+      "loss": 0.4709,
+      "step": 187980
+    },
+    {
+      "epoch": 9.336942485348168,
+      "grad_norm": 0.224609375,
+      "learning_rate": 5.3048574550511576e-05,
+      "loss": 0.4948,
+      "step": 187990
+    },
+    {
+      "epoch": 9.337439157643786,
+      "grad_norm": 0.19921875,
+      "learning_rate": 5.300884076686202e-05,
+      "loss": 0.4843,
+      "step": 188000
+    },
+    {
+      "epoch": 9.337935829939406,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.296910698321248e-05,
+      "loss": 0.4385,
+      "step": 188010
+    },
+    {
+      "epoch": 9.338432502235026,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 5.292937319956293e-05,
+      "loss": 0.4862,
+      "step": 188020
+    },
+    {
+      "epoch": 9.338929174530644,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.2889639415913387e-05,
+      "loss": 0.4811,
+      "step": 188030
+    },
+    {
+      "epoch": 9.339425846826265,
+      "grad_norm": 0.189453125,
+      "learning_rate": 5.284990563226383e-05,
+      "loss": 0.4694,
+      "step": 188040
+    },
+    {
+      "epoch": 9.339922519121883,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 5.281017184861429e-05,
+      "loss": 0.5002,
+      "step": 188050
+    },
+    {
+      "epoch": 9.340419191417503,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 5.277043806496473e-05,
+      "loss": 0.4877,
+      "step": 188060
+    },
+    {
+      "epoch": 9.340915863713121,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.273070428131519e-05,
+      "loss": 0.4835,
+      "step": 188070
+    },
+    {
+      "epoch": 9.341412536008741,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 5.269097049766565e-05,
+      "loss": 0.5024,
+      "step": 188080
+    },
+    {
+      "epoch": 9.341909208304362,
+      "grad_norm": 0.203125,
+      "learning_rate": 5.265123671401609e-05,
+      "loss": 0.4623,
+      "step": 188090
+    },
+    {
+      "epoch": 9.34240588059998,
+      "grad_norm": 0.228515625,
+      "learning_rate": 5.261150293036655e-05,
+      "loss": 0.5086,
+      "step": 188100
+    },
+    {
+      "epoch": 9.3429025528956,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5.2571769146716995e-05,
+      "loss": 0.4834,
+      "step": 188110
+    },
+    {
+      "epoch": 9.343399225191218,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 5.253203536306745e-05,
+      "loss": 0.4931,
+      "step": 188120
+    },
+    {
+      "epoch": 9.343895897486838,
+      "grad_norm": 0.201171875,
+      "learning_rate": 5.2492301579417904e-05,
+      "loss": 0.4896,
+      "step": 188130
+    },
+    {
+      "epoch": 9.344392569782457,
+      "grad_norm": 0.1953125,
+      "learning_rate": 5.245256779576836e-05,
+      "loss": 0.4814,
+      "step": 188140
+    },
+    {
+      "epoch": 9.344889242078077,
+      "grad_norm": 0.21484375,
+      "learning_rate": 5.2412834012118806e-05,
+      "loss": 0.4714,
+      "step": 188150
+    },
+    {
+      "epoch": 9.345385914373697,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 5.2373100228469263e-05,
+      "loss": 0.4631,
+      "step": 188160
+    },
+    {
+      "epoch": 9.345882586669315,
+      "grad_norm": 0.21484375,
+      "learning_rate": 5.233336644481971e-05,
+      "loss": 0.4694,
+      "step": 188170
+    },
+    {
+      "epoch": 9.346379258964935,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 5.2293632661170166e-05,
+      "loss": 0.5235,
+      "step": 188180
+    },
+    {
+      "epoch": 9.346875931260554,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 5.225389887752061e-05,
+      "loss": 0.4933,
+      "step": 188190
+    },
+    {
+      "epoch": 9.347372603556174,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.221416509387107e-05,
+      "loss": 0.4818,
+      "step": 188200
+    },
+    {
+      "epoch": 9.347869275851792,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.217443131022152e-05,
+      "loss": 0.4755,
+      "step": 188210
+    },
+    {
+      "epoch": 9.348365948147412,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 5.2134697526571976e-05,
+      "loss": 0.5092,
+      "step": 188220
+    },
+    {
+      "epoch": 9.348862620443033,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.209496374292242e-05,
+      "loss": 0.5006,
+      "step": 188230
+    },
+    {
+      "epoch": 9.34935929273865,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.205522995927288e-05,
+      "loss": 0.4579,
+      "step": 188240
+    },
+    {
+      "epoch": 9.349855965034271,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.201549617562332e-05,
+      "loss": 0.4728,
+      "step": 188250
+    },
+    {
+      "epoch": 9.35035263732989,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 5.197576239197378e-05,
+      "loss": 0.4806,
+      "step": 188260
+    },
+    {
+      "epoch": 9.35084930962551,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.1936028608324225e-05,
+      "loss": 0.4834,
+      "step": 188270
+    },
+    {
+      "epoch": 9.351345981921128,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 5.189629482467468e-05,
+      "loss": 0.4668,
+      "step": 188280
+    },
+    {
+      "epoch": 9.351842654216748,
+      "grad_norm": 0.1953125,
+      "learning_rate": 5.1856561041025134e-05,
+      "loss": 0.4714,
+      "step": 188290
+    },
+    {
+      "epoch": 9.352339326512368,
+      "grad_norm": 0.232421875,
+      "learning_rate": 5.181682725737559e-05,
+      "loss": 0.4732,
+      "step": 188300
+    },
+    {
+      "epoch": 9.352835998807986,
+      "grad_norm": 0.216796875,
+      "learning_rate": 5.1777093473726036e-05,
+      "loss": 0.5134,
+      "step": 188310
+    },
+    {
+      "epoch": 9.353332671103606,
+      "grad_norm": 0.22265625,
+      "learning_rate": 5.1737359690076493e-05,
+      "loss": 0.4875,
+      "step": 188320
+    },
+    {
+      "epoch": 9.353829343399225,
+      "grad_norm": 0.22265625,
+      "learning_rate": 5.169762590642694e-05,
+      "loss": 0.513,
+      "step": 188330
+    },
+    {
+      "epoch": 9.354326015694845,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 5.1657892122777396e-05,
+      "loss": 0.4639,
+      "step": 188340
+    },
+    {
+      "epoch": 9.354822687990463,
+      "grad_norm": 0.193359375,
+      "learning_rate": 5.161815833912784e-05,
+      "loss": 0.5193,
+      "step": 188350
+    },
+    {
+      "epoch": 9.355319360286083,
+      "grad_norm": 0.20703125,
+      "learning_rate": 5.15784245554783e-05,
+      "loss": 0.4933,
+      "step": 188360
+    },
+    {
+      "epoch": 9.355816032581703,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.153869077182875e-05,
+      "loss": 0.4735,
+      "step": 188370
+    },
+    {
+      "epoch": 9.356312704877322,
+      "grad_norm": 0.212890625,
+      "learning_rate": 5.1498956988179206e-05,
+      "loss": 0.4596,
+      "step": 188380
+    },
+    {
+      "epoch": 9.356809377172942,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 5.145922320452965e-05,
+      "loss": 0.4963,
+      "step": 188390
+    },
+    {
+      "epoch": 9.35730604946856,
+      "grad_norm": 0.189453125,
+      "learning_rate": 5.141948942088011e-05,
+      "loss": 0.5084,
+      "step": 188400
+    },
+    {
+      "epoch": 9.35780272176418,
+      "grad_norm": 0.19921875,
+      "learning_rate": 5.137975563723055e-05,
+      "loss": 0.4659,
+      "step": 188410
+    },
+    {
+      "epoch": 9.358299394059799,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 5.134002185358101e-05,
+      "loss": 0.4687,
+      "step": 188420
+    },
+    {
+      "epoch": 9.358796066355419,
+      "grad_norm": 0.205078125,
+      "learning_rate": 5.1300288069931455e-05,
+      "loss": 0.4858,
+      "step": 188430
+    },
+    {
+      "epoch": 9.359292738651039,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 5.126055428628191e-05,
+      "loss": 0.5013,
+      "step": 188440
+    },
+    {
+      "epoch": 9.359789410946657,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 5.1220820502632364e-05,
+      "loss": 0.5071,
+      "step": 188450
+    },
+    {
+      "epoch": 9.360286083242277,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.118108671898282e-05,
+      "loss": 0.5256,
+      "step": 188460
+    },
+    {
+      "epoch": 9.360782755537896,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 5.1141352935333266e-05,
+      "loss": 0.49,
+      "step": 188470
+    },
+    {
+      "epoch": 9.361279427833516,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 5.1101619151683723e-05,
+      "loss": 0.4787,
+      "step": 188480
+    },
+    {
+      "epoch": 9.361776100129134,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 5.106188536803417e-05,
+      "loss": 0.4634,
+      "step": 188490
+    },
+    {
+      "epoch": 9.362272772424754,
+      "grad_norm": 0.21484375,
+      "learning_rate": 5.1022151584384626e-05,
+      "loss": 0.4745,
+      "step": 188500
+    },
+    {
+      "epoch": 9.362769444720373,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 5.098241780073508e-05,
+      "loss": 0.4861,
+      "step": 188510
+    },
+    {
+      "epoch": 9.363266117015993,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 5.094268401708553e-05,
+      "loss": 0.5009,
+      "step": 188520
+    },
+    {
+      "epoch": 9.363762789311613,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 5.0902950233435985e-05,
+      "loss": 0.5162,
+      "step": 188530
+    },
+    {
+      "epoch": 9.364259461607231,
+      "grad_norm": 0.212890625,
+      "learning_rate": 5.086321644978643e-05,
+      "loss": 0.4731,
+      "step": 188540
+    },
+    {
+      "epoch": 9.364756133902851,
+      "grad_norm": 0.216796875,
+      "learning_rate": 5.082348266613689e-05,
+      "loss": 0.4853,
+      "step": 188550
+    },
+    {
+      "epoch": 9.36525280619847,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 5.078374888248734e-05,
+      "loss": 0.4853,
+      "step": 188560
+    },
+    {
+      "epoch": 9.36574947849409,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 5.0744015098837796e-05,
+      "loss": 0.519,
+      "step": 188570
+    },
+    {
+      "epoch": 9.366246150789708,
+      "grad_norm": 0.193359375,
+      "learning_rate": 5.070428131518824e-05,
+      "loss": 0.4911,
+      "step": 188580
+    },
+    {
+      "epoch": 9.366742823085328,
+      "grad_norm": 0.234375,
+      "learning_rate": 5.06645475315387e-05,
+      "loss": 0.4996,
+      "step": 188590
+    },
+    {
+      "epoch": 9.367239495380948,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 5.062481374788914e-05,
+      "loss": 0.4864,
+      "step": 188600
+    },
+    {
+      "epoch": 9.367736167676567,
+      "grad_norm": 0.203125,
+      "learning_rate": 5.05850799642396e-05,
+      "loss": 0.4807,
+      "step": 188610
+    },
+    {
+      "epoch": 9.368232839972187,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.0545346180590045e-05,
+      "loss": 0.4658,
+      "step": 188620
+    },
+    {
+      "epoch": 9.368729512267805,
+      "grad_norm": 0.19921875,
+      "learning_rate": 5.05056123969405e-05,
+      "loss": 0.4984,
+      "step": 188630
+    },
+    {
+      "epoch": 9.369226184563425,
+      "grad_norm": 0.212890625,
+      "learning_rate": 5.0465878613290953e-05,
+      "loss": 0.4678,
+      "step": 188640
+    },
+    {
+      "epoch": 9.369722856859044,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.042614482964141e-05,
+      "loss": 0.4779,
+      "step": 188650
+    },
+    {
+      "epoch": 9.370219529154664,
+      "grad_norm": 0.185546875,
+      "learning_rate": 5.0386411045991855e-05,
+      "loss": 0.4985,
+      "step": 188660
+    },
+    {
+      "epoch": 9.370716201450284,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.034667726234231e-05,
+      "loss": 0.4886,
+      "step": 188670
+    },
+    {
+      "epoch": 9.371212873745902,
+      "grad_norm": 0.212890625,
+      "learning_rate": 5.030694347869276e-05,
+      "loss": 0.5163,
+      "step": 188680
+    },
+    {
+      "epoch": 9.371709546041522,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.0267209695043215e-05,
+      "loss": 0.4977,
+      "step": 188690
+    },
+    {
+      "epoch": 9.37220621833714,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 5.022747591139366e-05,
+      "loss": 0.5211,
+      "step": 188700
+    },
+    {
+      "epoch": 9.37270289063276,
+      "grad_norm": 0.1953125,
+      "learning_rate": 5.018774212774412e-05,
+      "loss": 0.4903,
+      "step": 188710
+    },
+    {
+      "epoch": 9.373199562928379,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.014800834409457e-05,
+      "loss": 0.4677,
+      "step": 188720
+    },
+    {
+      "epoch": 9.373696235224,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 5.0108274560445026e-05,
+      "loss": 0.4746,
+      "step": 188730
+    },
+    {
+      "epoch": 9.37419290751962,
+      "grad_norm": 0.193359375,
+      "learning_rate": 5.006854077679547e-05,
+      "loss": 0.4586,
+      "step": 188740
+    },
+    {
+      "epoch": 9.374689579815238,
+      "grad_norm": 0.185546875,
+      "learning_rate": 5.002880699314593e-05,
+      "loss": 0.4692,
+      "step": 188750
+    },
+    {
+      "epoch": 9.375186252110858,
+      "grad_norm": 0.208984375,
+      "learning_rate": 4.998907320949638e-05,
+      "loss": 0.4933,
+      "step": 188760
+    },
+    {
+      "epoch": 9.375682924406476,
+      "grad_norm": 0.25,
+      "learning_rate": 4.994933942584683e-05,
+      "loss": 0.493,
+      "step": 188770
+    },
+    {
+      "epoch": 9.376179596702096,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.990960564219728e-05,
+      "loss": 0.4864,
+      "step": 188780
+    },
+    {
+      "epoch": 9.376676268997715,
+      "grad_norm": 0.21484375,
+      "learning_rate": 4.986987185854773e-05,
+      "loss": 0.4733,
+      "step": 188790
+    },
+    {
+      "epoch": 9.377172941293335,
+      "grad_norm": 0.2265625,
+      "learning_rate": 4.9830138074898183e-05,
+      "loss": 0.5183,
+      "step": 188800
+    },
+    {
+      "epoch": 9.377669613588955,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.979040429124864e-05,
+      "loss": 0.4848,
+      "step": 188810
+    },
+    {
+      "epoch": 9.378166285884573,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 4.975067050759909e-05,
+      "loss": 0.5142,
+      "step": 188820
+    },
+    {
+      "epoch": 9.378662958180193,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.971093672394954e-05,
+      "loss": 0.4965,
+      "step": 188830
+    },
+    {
+      "epoch": 9.379159630475812,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 4.9671202940299994e-05,
+      "loss": 0.4981,
+      "step": 188840
+    },
+    {
+      "epoch": 9.379656302771432,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.9631469156650445e-05,
+      "loss": 0.4926,
+      "step": 188850
+    },
+    {
+      "epoch": 9.38015297506705,
+      "grad_norm": 0.197265625,
+      "learning_rate": 4.9591735373000896e-05,
+      "loss": 0.5015,
+      "step": 188860
+    },
+    {
+      "epoch": 9.38064964736267,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.955200158935135e-05,
+      "loss": 0.5019,
+      "step": 188870
+    },
+    {
+      "epoch": 9.38114631965829,
+      "grad_norm": 0.201171875,
+      "learning_rate": 4.95122678057018e-05,
+      "loss": 0.5043,
+      "step": 188880
+    },
+    {
+      "epoch": 9.381642991953909,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 4.947253402205225e-05,
+      "loss": 0.4871,
+      "step": 188890
+    },
+    {
+      "epoch": 9.382139664249529,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 4.943280023840271e-05,
+      "loss": 0.499,
+      "step": 188900
+    },
+    {
+      "epoch": 9.382636336545147,
+      "grad_norm": 0.1796875,
+      "learning_rate": 4.939306645475316e-05,
+      "loss": 0.4528,
+      "step": 188910
+    },
+    {
+      "epoch": 9.383133008840767,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 4.935333267110361e-05,
+      "loss": 0.483,
+      "step": 188920
+    },
+    {
+      "epoch": 9.383629681136386,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.931359888745406e-05,
+      "loss": 0.4871,
+      "step": 188930
+    },
+    {
+      "epoch": 9.384126353432006,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.927386510380451e-05,
+      "loss": 0.469,
+      "step": 188940
+    },
+    {
+      "epoch": 9.384623025727626,
+      "grad_norm": 0.2109375,
+      "learning_rate": 4.923413132015496e-05,
+      "loss": 0.4758,
+      "step": 188950
+    },
+    {
+      "epoch": 9.385119698023244,
+      "grad_norm": 0.248046875,
+      "learning_rate": 4.9194397536505413e-05,
+      "loss": 0.4691,
+      "step": 188960
+    },
+    {
+      "epoch": 9.385616370318864,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 4.9154663752855864e-05,
+      "loss": 0.4728,
+      "step": 188970
+    },
+    {
+      "epoch": 9.386113042614483,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 4.911492996920632e-05,
+      "loss": 0.4822,
+      "step": 188980
+    },
+    {
+      "epoch": 9.386609714910103,
+      "grad_norm": 0.193359375,
+      "learning_rate": 4.907519618555677e-05,
+      "loss": 0.4673,
+      "step": 188990
+    },
+    {
+      "epoch": 9.387106387205721,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.9035462401907224e-05,
+      "loss": 0.4952,
+      "step": 189000
+    },
+    {
+      "epoch": 9.387603059501341,
+      "grad_norm": 0.19140625,
+      "learning_rate": 4.8995728618257675e-05,
+      "loss": 0.4755,
+      "step": 189010
+    },
+    {
+      "epoch": 9.388099731796961,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 4.8955994834608126e-05,
+      "loss": 0.5159,
+      "step": 189020
+    },
+    {
+      "epoch": 9.38859640409258,
+      "grad_norm": 0.205078125,
+      "learning_rate": 4.891626105095858e-05,
+      "loss": 0.4792,
+      "step": 189030
+    },
+    {
+      "epoch": 9.3890930763882,
+      "grad_norm": 0.201171875,
+      "learning_rate": 4.887652726730903e-05,
+      "loss": 0.4912,
+      "step": 189040
+    },
+    {
+      "epoch": 9.389589748683818,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 4.883679348365948e-05,
+      "loss": 0.4851,
+      "step": 189050
+    },
+    {
+      "epoch": 9.390086420979438,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.879705970000994e-05,
+      "loss": 0.5188,
+      "step": 189060
+    },
+    {
+      "epoch": 9.390583093275056,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 4.875732591636039e-05,
+      "loss": 0.4958,
+      "step": 189070
+    },
+    {
+      "epoch": 9.391079765570677,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 4.871759213271084e-05,
+      "loss": 0.4566,
+      "step": 189080
+    },
+    {
+      "epoch": 9.391576437866295,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 4.867785834906129e-05,
+      "loss": 0.485,
+      "step": 189090
+    },
+    {
+      "epoch": 9.392073110161915,
+      "grad_norm": 0.208984375,
+      "learning_rate": 4.863812456541174e-05,
+      "loss": 0.4859,
+      "step": 189100
+    },
+    {
+      "epoch": 9.392569782457535,
+      "grad_norm": 0.203125,
+      "learning_rate": 4.859839078176219e-05,
+      "loss": 0.4685,
+      "step": 189110
+    },
+    {
+      "epoch": 9.393066454753153,
+      "grad_norm": 0.203125,
+      "learning_rate": 4.855865699811264e-05,
+      "loss": 0.4739,
+      "step": 189120
+    },
+    {
+      "epoch": 9.393563127048774,
+      "grad_norm": 0.1875,
+      "learning_rate": 4.85189232144631e-05,
+      "loss": 0.464,
+      "step": 189130
+    },
+    {
+      "epoch": 9.394059799344392,
+      "grad_norm": 0.220703125,
+      "learning_rate": 4.847918943081355e-05,
+      "loss": 0.5047,
+      "step": 189140
+    },
+    {
+      "epoch": 9.394556471640012,
+      "grad_norm": 0.2421875,
+      "learning_rate": 4.8439455647164e-05,
+      "loss": 0.5102,
+      "step": 189150
+    },
+    {
+      "epoch": 9.39505314393563,
+      "grad_norm": 0.197265625,
+      "learning_rate": 4.839972186351446e-05,
+      "loss": 0.4791,
+      "step": 189160
+    },
+    {
+      "epoch": 9.39554981623125,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 4.835998807986491e-05,
+      "loss": 0.4723,
+      "step": 189170
+    },
+    {
+      "epoch": 9.39604648852687,
+      "grad_norm": 0.1875,
+      "learning_rate": 4.832025429621536e-05,
+      "loss": 0.5001,
+      "step": 189180
+    },
+    {
+      "epoch": 9.396543160822489,
+      "grad_norm": 0.208984375,
+      "learning_rate": 4.8280520512565814e-05,
+      "loss": 0.4919,
+      "step": 189190
+    },
+    {
+      "epoch": 9.397039833118109,
+      "grad_norm": 0.2421875,
+      "learning_rate": 4.8240786728916265e-05,
+      "loss": 0.531,
+      "step": 189200
+    },
+    {
+      "epoch": 9.397536505413727,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 4.8201052945266716e-05,
+      "loss": 0.4866,
+      "step": 189210
+    },
+    {
+      "epoch": 9.398033177709348,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 4.816131916161717e-05,
+      "loss": 0.4984,
+      "step": 189220
+    },
+    {
+      "epoch": 9.398529850004966,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 4.812158537796762e-05,
+      "loss": 0.466,
+      "step": 189230
+    },
+    {
+      "epoch": 9.399026522300586,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 4.808185159431807e-05,
+      "loss": 0.5262,
+      "step": 189240
+    },
+    {
+      "epoch": 9.399523194596206,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.804211781066853e-05,
+      "loss": 0.4608,
+      "step": 189250
+    },
+    {
+      "epoch": 9.400019866891824,
+      "grad_norm": 0.2109375,
+      "learning_rate": 4.800238402701898e-05,
+      "loss": 0.4591,
+      "step": 189260
+    },
+    {
+      "epoch": 9.400516539187445,
+      "grad_norm": 0.208984375,
+      "learning_rate": 4.796265024336943e-05,
+      "loss": 0.4883,
+      "step": 189270
+    },
+    {
+      "epoch": 9.401013211483063,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 4.792291645971988e-05,
+      "loss": 0.4759,
+      "step": 189280
+    },
+    {
+      "epoch": 9.401509883778683,
+      "grad_norm": 0.201171875,
+      "learning_rate": 4.788318267607033e-05,
+      "loss": 0.4968,
+      "step": 189290
+    },
+    {
+      "epoch": 9.402006556074301,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.784344889242078e-05,
+      "loss": 0.4804,
+      "step": 189300
+    },
+    {
+      "epoch": 9.402503228369921,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 4.780371510877123e-05,
+      "loss": 0.4754,
+      "step": 189310
+    },
+    {
+      "epoch": 9.402999900665542,
+      "grad_norm": 0.212890625,
+      "learning_rate": 4.7763981325121684e-05,
+      "loss": 0.4695,
+      "step": 189320
+    },
+    {
+      "epoch": 9.40349657296116,
+      "grad_norm": 0.205078125,
+      "learning_rate": 4.772424754147214e-05,
+      "loss": 0.4796,
+      "step": 189330
+    },
+    {
+      "epoch": 9.40399324525678,
+      "grad_norm": 0.208984375,
+      "learning_rate": 4.768451375782259e-05,
+      "loss": 0.4904,
+      "step": 189340
+    },
+    {
+      "epoch": 9.404489917552398,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 4.7644779974173044e-05,
+      "loss": 0.4853,
+      "step": 189350
+    },
+    {
+      "epoch": 9.404986589848018,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.7605046190523495e-05,
+      "loss": 0.513,
+      "step": 189360
+    },
+    {
+      "epoch": 9.405483262143637,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 4.7565312406873946e-05,
+      "loss": 0.4981,
+      "step": 189370
+    },
+    {
+      "epoch": 9.405979934439257,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.75255786232244e-05,
+      "loss": 0.4926,
+      "step": 189380
+    },
+    {
+      "epoch": 9.406476606734877,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.748584483957485e-05,
+      "loss": 0.4727,
+      "step": 189390
+    },
+    {
+      "epoch": 9.406973279030495,
+      "grad_norm": 0.193359375,
+      "learning_rate": 4.74461110559253e-05,
+      "loss": 0.4861,
+      "step": 189400
+    },
+    {
+      "epoch": 9.407469951326116,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.740637727227576e-05,
+      "loss": 0.5027,
+      "step": 189410
+    },
+    {
+      "epoch": 9.407966623621734,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.736664348862621e-05,
+      "loss": 0.5004,
+      "step": 189420
+    },
+    {
+      "epoch": 9.408463295917354,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 4.732690970497666e-05,
+      "loss": 0.5357,
+      "step": 189430
+    },
+    {
+      "epoch": 9.408959968212972,
+      "grad_norm": 0.19140625,
+      "learning_rate": 4.728717592132711e-05,
+      "loss": 0.4972,
+      "step": 189440
+    },
+    {
+      "epoch": 9.409456640508592,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 4.724744213767756e-05,
+      "loss": 0.5064,
+      "step": 189450
+    },
+    {
+      "epoch": 9.409953312804213,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.720770835402801e-05,
+      "loss": 0.4709,
+      "step": 189460
+    },
+    {
+      "epoch": 9.41044998509983,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.716797457037846e-05,
+      "loss": 0.4683,
+      "step": 189470
+    },
+    {
+      "epoch": 9.410946657395451,
+      "grad_norm": 0.271484375,
+      "learning_rate": 4.7128240786728914e-05,
+      "loss": 0.5011,
+      "step": 189480
+    },
+    {
+      "epoch": 9.41144332969107,
+      "grad_norm": 0.2109375,
+      "learning_rate": 4.7088507003079365e-05,
+      "loss": 0.5058,
+      "step": 189490
+    },
+    {
+      "epoch": 9.41194000198669,
+      "grad_norm": 0.208984375,
+      "learning_rate": 4.704877321942982e-05,
+      "loss": 0.5098,
+      "step": 189500
+    },
+    {
+      "epoch": 9.412436674282308,
+      "grad_norm": 0.22265625,
+      "learning_rate": 4.7009039435780274e-05,
+      "loss": 0.4796,
+      "step": 189510
+    },
+    {
+      "epoch": 9.412933346577928,
+      "grad_norm": 0.224609375,
+      "learning_rate": 4.6969305652130725e-05,
+      "loss": 0.5047,
+      "step": 189520
+    },
+    {
+      "epoch": 9.413430018873548,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.6929571868481176e-05,
+      "loss": 0.4659,
+      "step": 189530
+    },
+    {
+      "epoch": 9.413926691169166,
+      "grad_norm": 0.197265625,
+      "learning_rate": 4.688983808483163e-05,
+      "loss": 0.471,
+      "step": 189540
+    },
+    {
+      "epoch": 9.414423363464786,
+      "grad_norm": 0.1875,
+      "learning_rate": 4.6850104301182085e-05,
+      "loss": 0.4838,
+      "step": 189550
+    },
+    {
+      "epoch": 9.414920035760405,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.6810370517532536e-05,
+      "loss": 0.4829,
+      "step": 189560
+    },
+    {
+      "epoch": 9.415416708056025,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 4.677063673388299e-05,
+      "loss": 0.4788,
+      "step": 189570
+    },
+    {
+      "epoch": 9.415913380351643,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 4.673090295023344e-05,
+      "loss": 0.4913,
+      "step": 189580
+    },
+    {
+      "epoch": 9.416410052647263,
+      "grad_norm": 0.236328125,
+      "learning_rate": 4.669116916658389e-05,
+      "loss": 0.5042,
+      "step": 189590
+    },
+    {
+      "epoch": 9.416906724942883,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 4.665143538293435e-05,
+      "loss": 0.4967,
+      "step": 189600
+    },
+    {
+      "epoch": 9.417403397238502,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 4.66117015992848e-05,
+      "loss": 0.4489,
+      "step": 189610
+    },
+    {
+      "epoch": 9.417900069534122,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.657196781563525e-05,
+      "loss": 0.5097,
+      "step": 189620
+    },
+    {
+      "epoch": 9.41839674182974,
+      "grad_norm": 0.22265625,
+      "learning_rate": 4.65322340319857e-05,
+      "loss": 0.4542,
+      "step": 189630
+    },
+    {
+      "epoch": 9.41889341412536,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 4.649250024833615e-05,
+      "loss": 0.4861,
+      "step": 189640
+    },
+    {
+      "epoch": 9.419390086420979,
+      "grad_norm": 0.255859375,
+      "learning_rate": 4.64527664646866e-05,
+      "loss": 0.4727,
+      "step": 189650
+    },
+    {
+      "epoch": 9.419886758716599,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 4.641303268103705e-05,
+      "loss": 0.4634,
+      "step": 189660
+    },
+    {
+      "epoch": 9.420383431012219,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.6373298897387504e-05,
+      "loss": 0.4895,
+      "step": 189670
+    },
+    {
+      "epoch": 9.420880103307837,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 4.633356511373796e-05,
+      "loss": 0.5143,
+      "step": 189680
+    },
+    {
+      "epoch": 9.421376775603457,
+      "grad_norm": 0.193359375,
+      "learning_rate": 4.629383133008841e-05,
+      "loss": 0.4595,
+      "step": 189690
+    },
+    {
+      "epoch": 9.421873447899076,
+      "grad_norm": 0.193359375,
+      "learning_rate": 4.6254097546438864e-05,
+      "loss": 0.501,
+      "step": 189700
+    },
+    {
+      "epoch": 9.422370120194696,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 4.6214363762789315e-05,
+      "loss": 0.4674,
+      "step": 189710
+    },
+    {
+      "epoch": 9.422866792490314,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.6174629979139766e-05,
+      "loss": 0.5058,
+      "step": 189720
+    },
+    {
+      "epoch": 9.423363464785934,
+      "grad_norm": 0.224609375,
+      "learning_rate": 4.613489619549022e-05,
+      "loss": 0.4846,
+      "step": 189730
+    },
+    {
+      "epoch": 9.423860137081554,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 4.609516241184067e-05,
+      "loss": 0.4785,
+      "step": 189740
+    },
+    {
+      "epoch": 9.424356809377173,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.605542862819112e-05,
+      "loss": 0.5165,
+      "step": 189750
+    },
+    {
+      "epoch": 9.424853481672793,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.601569484454158e-05,
+      "loss": 0.4946,
+      "step": 189760
+    },
+    {
+      "epoch": 9.425350153968411,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 4.597596106089203e-05,
+      "loss": 0.5008,
+      "step": 189770
+    },
+    {
+      "epoch": 9.425846826264031,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 4.593622727724248e-05,
+      "loss": 0.5056,
+      "step": 189780
+    },
+    {
+      "epoch": 9.42634349855965,
+      "grad_norm": 0.224609375,
+      "learning_rate": 4.589649349359293e-05,
+      "loss": 0.4819,
+      "step": 189790
+    },
+    {
+      "epoch": 9.42684017085527,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.585675970994338e-05,
+      "loss": 0.4559,
+      "step": 189800
+    },
+    {
+      "epoch": 9.42733684315089,
+      "grad_norm": 0.2109375,
+      "learning_rate": 4.581702592629383e-05,
+      "loss": 0.4852,
+      "step": 189810
+    },
+    {
+      "epoch": 9.427833515446508,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 4.577729214264428e-05,
+      "loss": 0.4828,
+      "step": 189820
+    },
+    {
+      "epoch": 9.428330187742128,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 4.5737558358994734e-05,
+      "loss": 0.5205,
+      "step": 189830
+    },
+    {
+      "epoch": 9.428826860037747,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 4.5697824575345185e-05,
+      "loss": 0.4804,
+      "step": 189840
+    },
+    {
+      "epoch": 9.429323532333367,
+      "grad_norm": 0.201171875,
+      "learning_rate": 4.565809079169564e-05,
+      "loss": 0.5254,
+      "step": 189850
+    },
+    {
+      "epoch": 9.429820204628985,
+      "grad_norm": 0.18359375,
+      "learning_rate": 4.5618357008046094e-05,
+      "loss": 0.49,
+      "step": 189860
+    },
+    {
+      "epoch": 9.430316876924605,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.5578623224396545e-05,
+      "loss": 0.4915,
+      "step": 189870
+    },
+    {
+      "epoch": 9.430813549220225,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.5538889440746996e-05,
+      "loss": 0.4658,
+      "step": 189880
+    },
+    {
+      "epoch": 9.431310221515844,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 4.549915565709745e-05,
+      "loss": 0.5017,
+      "step": 189890
+    },
+    {
+      "epoch": 9.431806893811464,
+      "grad_norm": 0.212890625,
+      "learning_rate": 4.54594218734479e-05,
+      "loss": 0.5033,
+      "step": 189900
+    },
+    {
+      "epoch": 9.432303566107082,
+      "grad_norm": 0.224609375,
+      "learning_rate": 4.541968808979835e-05,
+      "loss": 0.4766,
+      "step": 189910
+    },
+    {
+      "epoch": 9.432800238402702,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 4.53799543061488e-05,
+      "loss": 0.4802,
+      "step": 189920
+    },
+    {
+      "epoch": 9.43329691069832,
+      "grad_norm": 0.203125,
+      "learning_rate": 4.534022052249926e-05,
+      "loss": 0.4642,
+      "step": 189930
+    },
+    {
+      "epoch": 9.43379358299394,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 4.530048673884971e-05,
+      "loss": 0.4749,
+      "step": 189940
+    },
+    {
+      "epoch": 9.43429025528956,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 4.526075295520016e-05,
+      "loss": 0.4763,
+      "step": 189950
+    },
+    {
+      "epoch": 9.43478692758518,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.522101917155061e-05,
+      "loss": 0.4624,
+      "step": 189960
+    },
+    {
+      "epoch": 9.4352835998808,
+      "grad_norm": 0.208984375,
+      "learning_rate": 4.518128538790107e-05,
+      "loss": 0.4999,
+      "step": 189970
+    },
+    {
+      "epoch": 9.435780272176418,
+      "grad_norm": 0.205078125,
+      "learning_rate": 4.514155160425152e-05,
+      "loss": 0.5081,
+      "step": 189980
+    },
+    {
+      "epoch": 9.436276944472038,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 4.510181782060197e-05,
+      "loss": 0.4671,
+      "step": 189990
+    },
+    {
+      "epoch": 9.436773616767656,
+      "grad_norm": 0.267578125,
+      "learning_rate": 4.506208403695242e-05,
+      "loss": 0.4814,
+      "step": 190000
+    },
+    {
+      "epoch": 9.437270289063276,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 4.502235025330287e-05,
+      "loss": 0.481,
+      "step": 190010
+    },
+    {
+      "epoch": 9.437766961358895,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 4.4982616469653324e-05,
+      "loss": 0.4711,
+      "step": 190020
+    },
+    {
+      "epoch": 9.438263633654515,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 4.494288268600378e-05,
+      "loss": 0.4816,
+      "step": 190030
+    },
+    {
+      "epoch": 9.438760305950135,
+      "grad_norm": 0.205078125,
+      "learning_rate": 4.490314890235423e-05,
+      "loss": 0.4984,
+      "step": 190040
+    },
+    {
+      "epoch": 9.439256978245753,
+      "grad_norm": 0.234375,
+      "learning_rate": 4.4863415118704684e-05,
+      "loss": 0.4892,
+      "step": 190050
+    },
+    {
+      "epoch": 9.439753650541373,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 4.4823681335055135e-05,
+      "loss": 0.4802,
+      "step": 190060
+    },
+    {
+      "epoch": 9.440250322836992,
+      "grad_norm": 0.21484375,
+      "learning_rate": 4.4783947551405586e-05,
+      "loss": 0.4801,
+      "step": 190070
+    },
+    {
+      "epoch": 9.440746995132612,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 4.474421376775604e-05,
+      "loss": 0.4821,
+      "step": 190080
+    },
+    {
+      "epoch": 9.44124366742823,
+      "grad_norm": 0.2421875,
+      "learning_rate": 4.470447998410649e-05,
+      "loss": 0.4888,
+      "step": 190090
+    },
+    {
+      "epoch": 9.44174033972385,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 4.466474620045694e-05,
+      "loss": 0.4631,
+      "step": 190100
+    },
+    {
+      "epoch": 9.44223701201947,
+      "grad_norm": 0.2421875,
+      "learning_rate": 4.46250124168074e-05,
+      "loss": 0.4987,
+      "step": 190110
+    },
+    {
+      "epoch": 9.442733684315089,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 4.458527863315785e-05,
+      "loss": 0.4833,
+      "step": 190120
+    },
+    {
+      "epoch": 9.443230356610709,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.45455448495083e-05,
+      "loss": 0.4911,
+      "step": 190130
+    },
+    {
+      "epoch": 9.443727028906327,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 4.450581106585875e-05,
+      "loss": 0.4735,
+      "step": 190140
+    },
+    {
+      "epoch": 9.444223701201947,
+      "grad_norm": 0.19140625,
+      "learning_rate": 4.44660772822092e-05,
+      "loss": 0.4937,
+      "step": 190150
+    },
+    {
+      "epoch": 9.444720373497566,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.442634349855965e-05,
+      "loss": 0.47,
+      "step": 190160
+    },
+    {
+      "epoch": 9.445217045793186,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.43866097149101e-05,
+      "loss": 0.4796,
+      "step": 190170
+    },
+    {
+      "epoch": 9.445713718088806,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.4346875931260554e-05,
+      "loss": 0.4912,
+      "step": 190180
+    },
+    {
+      "epoch": 9.446210390384424,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 4.4307142147611005e-05,
+      "loss": 0.5073,
+      "step": 190190
+    },
+    {
+      "epoch": 9.446707062680044,
+      "grad_norm": 0.205078125,
+      "learning_rate": 4.426740836396146e-05,
+      "loss": 0.4368,
+      "step": 190200
+    },
+    {
+      "epoch": 9.447203734975663,
+      "grad_norm": 0.234375,
+      "learning_rate": 4.4227674580311914e-05,
+      "loss": 0.5063,
+      "step": 190210
+    },
+    {
+      "epoch": 9.447700407271283,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.4187940796662365e-05,
+      "loss": 0.4876,
+      "step": 190220
+    },
+    {
+      "epoch": 9.448197079566901,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.4148207013012816e-05,
+      "loss": 0.4991,
+      "step": 190230
+    },
+    {
+      "epoch": 9.448693751862521,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.410847322936327e-05,
+      "loss": 0.475,
+      "step": 190240
+    },
+    {
+      "epoch": 9.449190424158141,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 4.406873944571372e-05,
+      "loss": 0.5007,
+      "step": 190250
+    },
+    {
+      "epoch": 9.44968709645376,
+      "grad_norm": 0.208984375,
+      "learning_rate": 4.402900566206417e-05,
+      "loss": 0.4957,
+      "step": 190260
+    },
+    {
+      "epoch": 9.45018376874938,
+      "grad_norm": 0.205078125,
+      "learning_rate": 4.398927187841462e-05,
+      "loss": 0.474,
+      "step": 190270
+    },
+    {
+      "epoch": 9.450680441044998,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 4.394953809476508e-05,
+      "loss": 0.4743,
+      "step": 190280
+    },
+    {
+      "epoch": 9.451177113340618,
+      "grad_norm": 0.2109375,
+      "learning_rate": 4.390980431111553e-05,
+      "loss": 0.4712,
+      "step": 190290
+    },
+    {
+      "epoch": 9.451673785636237,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.387007052746598e-05,
+      "loss": 0.458,
+      "step": 190300
+    },
+    {
+      "epoch": 9.452170457931857,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.383033674381643e-05,
+      "loss": 0.4802,
+      "step": 190310
+    },
+    {
+      "epoch": 9.452667130227477,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 4.379060296016688e-05,
+      "loss": 0.4896,
+      "step": 190320
+    },
+    {
+      "epoch": 9.453163802523095,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 4.375086917651733e-05,
+      "loss": 0.4979,
+      "step": 190330
+    },
+    {
+      "epoch": 9.453660474818715,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 4.3711135392867784e-05,
+      "loss": 0.5046,
+      "step": 190340
+    },
+    {
+      "epoch": 9.454157147114334,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.3671401609218235e-05,
+      "loss": 0.4758,
+      "step": 190350
+    },
+    {
+      "epoch": 9.454653819409954,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 4.363166782556869e-05,
+      "loss": 0.4676,
+      "step": 190360
+    },
+    {
+      "epoch": 9.455150491705572,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 4.3591934041919144e-05,
+      "loss": 0.4821,
+      "step": 190370
+    },
+    {
+      "epoch": 9.455647164001192,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.3552200258269595e-05,
+      "loss": 0.5047,
+      "step": 190380
+    },
+    {
+      "epoch": 9.456143836296812,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 4.351246647462005e-05,
+      "loss": 0.5165,
+      "step": 190390
+    },
+    {
+      "epoch": 9.45664050859243,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 4.3472732690970504e-05,
+      "loss": 0.4672,
+      "step": 190400
+    },
+    {
+      "epoch": 9.45713718088805,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.3432998907320955e-05,
+      "loss": 0.472,
+      "step": 190410
+    },
+    {
+      "epoch": 9.457633853183669,
+      "grad_norm": 0.208984375,
+      "learning_rate": 4.3393265123671406e-05,
+      "loss": 0.4751,
+      "step": 190420
+    },
+    {
+      "epoch": 9.45813052547929,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.335353134002186e-05,
+      "loss": 0.4832,
+      "step": 190430
+    },
+    {
+      "epoch": 9.458627197774907,
+      "grad_norm": 0.228515625,
+      "learning_rate": 4.331379755637231e-05,
+      "loss": 0.5197,
+      "step": 190440
+    },
+    {
+      "epoch": 9.459123870070528,
+      "grad_norm": 0.21484375,
+      "learning_rate": 4.327406377272276e-05,
+      "loss": 0.5084,
+      "step": 190450
+    },
+    {
+      "epoch": 9.459620542366146,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.3234329989073217e-05,
+      "loss": 0.4717,
+      "step": 190460
+    },
+    {
+      "epoch": 9.460117214661766,
+      "grad_norm": 0.203125,
+      "learning_rate": 4.319459620542367e-05,
+      "loss": 0.5032,
+      "step": 190470
+    },
+    {
+      "epoch": 9.460613886957386,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 4.315486242177412e-05,
+      "loss": 0.4943,
+      "step": 190480
+    },
+    {
+      "epoch": 9.461110559253004,
+      "grad_norm": 0.240234375,
+      "learning_rate": 4.311512863812457e-05,
+      "loss": 0.4644,
+      "step": 190490
+    },
+    {
+      "epoch": 9.461607231548625,
+      "grad_norm": 0.203125,
+      "learning_rate": 4.307539485447502e-05,
+      "loss": 0.4937,
+      "step": 190500
+    },
+    {
+      "epoch": 9.462103903844243,
+      "grad_norm": 0.232421875,
+      "learning_rate": 4.303566107082547e-05,
+      "loss": 0.5034,
+      "step": 190510
+    },
+    {
+      "epoch": 9.462600576139863,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 4.299592728717592e-05,
+      "loss": 0.4807,
+      "step": 190520
+    },
+    {
+      "epoch": 9.463097248435481,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.2956193503526374e-05,
+      "loss": 0.4897,
+      "step": 190530
+    },
+    {
+      "epoch": 9.463593920731102,
+      "grad_norm": 0.197265625,
+      "learning_rate": 4.291645971987683e-05,
+      "loss": 0.4738,
+      "step": 190540
+    },
+    {
+      "epoch": 9.464090593026722,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 4.287672593622728e-05,
+      "loss": 0.5169,
+      "step": 190550
+    },
+    {
+      "epoch": 9.46458726532234,
+      "grad_norm": 0.205078125,
+      "learning_rate": 4.2836992152577734e-05,
+      "loss": 0.5095,
+      "step": 190560
+    },
+    {
+      "epoch": 9.46508393761796,
+      "grad_norm": 0.201171875,
+      "learning_rate": 4.2797258368928185e-05,
+      "loss": 0.4726,
+      "step": 190570
+    },
+    {
+      "epoch": 9.465580609913578,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 4.2757524585278636e-05,
+      "loss": 0.4548,
+      "step": 190580
+    },
+    {
+      "epoch": 9.466077282209199,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.271779080162909e-05,
+      "loss": 0.481,
+      "step": 190590
+    },
+    {
+      "epoch": 9.466573954504817,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 4.267805701797954e-05,
+      "loss": 0.5057,
+      "step": 190600
+    },
+    {
+      "epoch": 9.467070626800437,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.263832323432999e-05,
+      "loss": 0.4813,
+      "step": 190610
+    },
+    {
+      "epoch": 9.467567299096057,
+      "grad_norm": 0.201171875,
+      "learning_rate": 4.259858945068044e-05,
+      "loss": 0.5279,
+      "step": 190620
+    },
+    {
+      "epoch": 9.468063971391675,
+      "grad_norm": 0.193359375,
+      "learning_rate": 4.25588556670309e-05,
+      "loss": 0.4943,
+      "step": 190630
+    },
+    {
+      "epoch": 9.468560643687296,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.251912188338135e-05,
+      "loss": 0.4927,
+      "step": 190640
+    },
+    {
+      "epoch": 9.469057315982914,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 4.24793880997318e-05,
+      "loss": 0.4796,
+      "step": 190650
+    },
+    {
+      "epoch": 9.469553988278534,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 4.243965431608225e-05,
+      "loss": 0.4778,
+      "step": 190660
+    },
+    {
+      "epoch": 9.470050660574152,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 4.23999205324327e-05,
+      "loss": 0.5134,
+      "step": 190670
+    },
+    {
+      "epoch": 9.470547332869772,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 4.236018674878315e-05,
+      "loss": 0.4747,
+      "step": 190680
+    },
+    {
+      "epoch": 9.471044005165393,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 4.2320452965133604e-05,
+      "loss": 0.5005,
+      "step": 190690
+    },
+    {
+      "epoch": 9.471540677461011,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 4.2280719181484055e-05,
+      "loss": 0.4956,
+      "step": 190700
+    },
+    {
+      "epoch": 9.472037349756631,
+      "grad_norm": 0.228515625,
+      "learning_rate": 4.224098539783451e-05,
+      "loss": 0.5076,
+      "step": 190710
+    },
+    {
+      "epoch": 9.47253402205225,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.2201251614184964e-05,
+      "loss": 0.4921,
+      "step": 190720
+    },
+    {
+      "epoch": 9.47303069434787,
+      "grad_norm": 0.208984375,
+      "learning_rate": 4.2161517830535415e-05,
+      "loss": 0.4581,
+      "step": 190730
+    },
+    {
+      "epoch": 9.473527366643488,
+      "grad_norm": 0.201171875,
+      "learning_rate": 4.2121784046885866e-05,
+      "loss": 0.4879,
+      "step": 190740
+    },
+    {
+      "epoch": 9.474024038939108,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 4.2082050263236317e-05,
+      "loss": 0.4743,
+      "step": 190750
+    },
+    {
+      "epoch": 9.474520711234728,
+      "grad_norm": 0.220703125,
+      "learning_rate": 4.204231647958677e-05,
+      "loss": 0.4943,
+      "step": 190760
+    },
+    {
+      "epoch": 9.475017383530346,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 4.200258269593722e-05,
+      "loss": 0.4593,
+      "step": 190770
+    },
+    {
+      "epoch": 9.475514055825967,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.196284891228767e-05,
+      "loss": 0.4725,
+      "step": 190780
+    },
+    {
+      "epoch": 9.476010728121585,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 4.192311512863812e-05,
+      "loss": 0.4773,
+      "step": 190790
+    },
+    {
+      "epoch": 9.476507400417205,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.188338134498858e-05,
+      "loss": 0.4791,
+      "step": 190800
+    },
+    {
+      "epoch": 9.477004072712823,
+      "grad_norm": 0.208984375,
+      "learning_rate": 4.184364756133903e-05,
+      "loss": 0.503,
+      "step": 190810
+    },
+    {
+      "epoch": 9.477500745008443,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 4.180391377768949e-05,
+      "loss": 0.5181,
+      "step": 190820
+    },
+    {
+      "epoch": 9.477997417304064,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 4.176417999403994e-05,
+      "loss": 0.465,
+      "step": 190830
+    },
+    {
+      "epoch": 9.478494089599682,
+      "grad_norm": 0.2109375,
+      "learning_rate": 4.172444621039039e-05,
+      "loss": 0.511,
+      "step": 190840
+    },
+    {
+      "epoch": 9.478990761895302,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 4.168471242674084e-05,
+      "loss": 0.4744,
+      "step": 190850
+    },
+    {
+      "epoch": 9.47948743419092,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.164497864309129e-05,
+      "loss": 0.4672,
+      "step": 190860
+    },
+    {
+      "epoch": 9.47998410648654,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 4.160524485944174e-05,
+      "loss": 0.472,
+      "step": 190870
+    },
+    {
+      "epoch": 9.480480778782159,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.1565511075792194e-05,
+      "loss": 0.4951,
+      "step": 190880
+    },
+    {
+      "epoch": 9.480977451077779,
+      "grad_norm": 0.23046875,
+      "learning_rate": 4.152577729214265e-05,
+      "loss": 0.5035,
+      "step": 190890
+    },
+    {
+      "epoch": 9.481474123373399,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 4.14860435084931e-05,
+      "loss": 0.4604,
+      "step": 190900
+    },
+    {
+      "epoch": 9.481970795669017,
+      "grad_norm": 0.19140625,
+      "learning_rate": 4.144630972484355e-05,
+      "loss": 0.5188,
+      "step": 190910
+    },
+    {
+      "epoch": 9.482467467964637,
+      "grad_norm": 0.212890625,
+      "learning_rate": 4.1406575941194004e-05,
+      "loss": 0.4791,
+      "step": 190920
+    },
+    {
+      "epoch": 9.482964140260256,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 4.1366842157544455e-05,
+      "loss": 0.4911,
+      "step": 190930
+    },
+    {
+      "epoch": 9.483460812555876,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 4.1327108373894906e-05,
+      "loss": 0.4816,
+      "step": 190940
+    },
+    {
+      "epoch": 9.483957484851494,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 4.128737459024536e-05,
+      "loss": 0.4844,
+      "step": 190950
+    },
+    {
+      "epoch": 9.484454157147114,
+      "grad_norm": 0.21484375,
+      "learning_rate": 4.124764080659581e-05,
+      "loss": 0.5126,
+      "step": 190960
+    },
+    {
+      "epoch": 9.484950829442734,
+      "grad_norm": 0.205078125,
+      "learning_rate": 4.120790702294626e-05,
+      "loss": 0.5173,
+      "step": 190970
+    },
+    {
+      "epoch": 9.485447501738353,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 4.116817323929672e-05,
+      "loss": 0.4946,
+      "step": 190980
+    },
+    {
+      "epoch": 9.485944174033973,
+      "grad_norm": 0.2265625,
+      "learning_rate": 4.112843945564717e-05,
+      "loss": 0.5297,
+      "step": 190990
+    },
+    {
+      "epoch": 9.486440846329591,
+      "grad_norm": 0.201171875,
+      "learning_rate": 4.108870567199762e-05,
+      "loss": 0.4899,
+      "step": 191000
+    },
+    {
+      "epoch": 9.486937518625211,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 4.104897188834807e-05,
+      "loss": 0.5052,
+      "step": 191010
+    },
+    {
+      "epoch": 9.48743419092083,
+      "grad_norm": 0.21484375,
+      "learning_rate": 4.100923810469852e-05,
+      "loss": 0.5071,
+      "step": 191020
+    },
+    {
+      "epoch": 9.48793086321645,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 4.096950432104897e-05,
+      "loss": 0.4538,
+      "step": 191030
+    },
+    {
+      "epoch": 9.48842753551207,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 4.0929770537399423e-05,
+      "loss": 0.4967,
+      "step": 191040
+    },
+    {
+      "epoch": 9.488924207807688,
+      "grad_norm": 0.1875,
+      "learning_rate": 4.0890036753749875e-05,
+      "loss": 0.5068,
+      "step": 191050
+    },
+    {
+      "epoch": 9.489420880103308,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 4.085030297010033e-05,
+      "loss": 0.4787,
+      "step": 191060
+    },
+    {
+      "epoch": 9.489917552398927,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 4.081056918645078e-05,
+      "loss": 0.4781,
+      "step": 191070
+    },
+    {
+      "epoch": 9.490414224694547,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 4.0770835402801234e-05,
+      "loss": 0.5005,
+      "step": 191080
+    },
+    {
+      "epoch": 9.490910896990165,
+      "grad_norm": 0.205078125,
+      "learning_rate": 4.0731101619151685e-05,
+      "loss": 0.4583,
+      "step": 191090
+    },
+    {
+      "epoch": 9.491407569285785,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 4.0691367835502136e-05,
+      "loss": 0.4993,
+      "step": 191100
+    },
+    {
+      "epoch": 9.491904241581405,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 4.065163405185259e-05,
+      "loss": 0.4824,
+      "step": 191110
+    },
+    {
+      "epoch": 9.492400913877024,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.061190026820304e-05,
+      "loss": 0.4716,
+      "step": 191120
+    },
+    {
+      "epoch": 9.492897586172644,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 4.057216648455349e-05,
+      "loss": 0.4972,
+      "step": 191130
+    },
+    {
+      "epoch": 9.493394258468262,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 4.053243270090395e-05,
+      "loss": 0.5021,
+      "step": 191140
+    },
+    {
+      "epoch": 9.493890930763882,
+      "grad_norm": 0.205078125,
+      "learning_rate": 4.04926989172544e-05,
+      "loss": 0.5069,
+      "step": 191150
+    },
+    {
+      "epoch": 9.4943876030595,
+      "grad_norm": 0.234375,
+      "learning_rate": 4.045296513360485e-05,
+      "loss": 0.5028,
+      "step": 191160
+    },
+    {
+      "epoch": 9.49488427535512,
+      "grad_norm": 0.212890625,
+      "learning_rate": 4.04132313499553e-05,
+      "loss": 0.4892,
+      "step": 191170
+    },
+    {
+      "epoch": 9.495380947650741,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 4.037349756630575e-05,
+      "loss": 0.5041,
+      "step": 191180
+    },
+    {
+      "epoch": 9.49587761994636,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 4.03337637826562e-05,
+      "loss": 0.461,
+      "step": 191190
+    },
+    {
+      "epoch": 9.49637429224198,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 4.0294029999006653e-05,
+      "loss": 0.5065,
+      "step": 191200
+    },
+    {
+      "epoch": 9.496870964537598,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.0254296215357105e-05,
+      "loss": 0.4549,
+      "step": 191210
+    },
+    {
+      "epoch": 9.497367636833218,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 4.0214562431707556e-05,
+      "loss": 0.5129,
+      "step": 191220
+    },
+    {
+      "epoch": 9.497864309128836,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.017482864805801e-05,
+      "loss": 0.4957,
+      "step": 191230
+    },
+    {
+      "epoch": 9.498360981424456,
+      "grad_norm": 0.18359375,
+      "learning_rate": 4.013509486440847e-05,
+      "loss": 0.4678,
+      "step": 191240
+    },
+    {
+      "epoch": 9.498857653720076,
+      "grad_norm": 0.203125,
+      "learning_rate": 4.009536108075892e-05,
+      "loss": 0.4919,
+      "step": 191250
+    },
+    {
+      "epoch": 9.499354326015695,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 4.005562729710937e-05,
+      "loss": 0.473,
+      "step": 191260
+    },
+    {
+      "epoch": 9.499850998311315,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 4.0015893513459824e-05,
+      "loss": 0.4701,
+      "step": 191270
+    },
+    {
+      "epoch": 9.500347670606933,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 3.9976159729810275e-05,
+      "loss": 0.5078,
+      "step": 191280
+    },
+    {
+      "epoch": 9.500844342902553,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.9936425946160726e-05,
+      "loss": 0.4614,
+      "step": 191290
+    },
+    {
+      "epoch": 9.501341015198172,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 3.989669216251118e-05,
+      "loss": 0.5163,
+      "step": 191300
+    },
+    {
+      "epoch": 9.501837687493792,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 3.985695837886163e-05,
+      "loss": 0.4596,
+      "step": 191310
+    },
+    {
+      "epoch": 9.502334359789412,
+      "grad_norm": 0.2109375,
+      "learning_rate": 3.981722459521208e-05,
+      "loss": 0.4776,
+      "step": 191320
+    },
+    {
+      "epoch": 9.50283103208503,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 3.977749081156254e-05,
+      "loss": 0.4652,
+      "step": 191330
+    },
+    {
+      "epoch": 9.50332770438065,
+      "grad_norm": 0.216796875,
+      "learning_rate": 3.973775702791299e-05,
+      "loss": 0.4922,
+      "step": 191340
+    },
+    {
+      "epoch": 9.503824376676269,
+      "grad_norm": 0.203125,
+      "learning_rate": 3.969802324426344e-05,
+      "loss": 0.4952,
+      "step": 191350
+    },
+    {
+      "epoch": 9.504321048971889,
+      "grad_norm": 0.201171875,
+      "learning_rate": 3.965828946061389e-05,
+      "loss": 0.4622,
+      "step": 191360
+    },
+    {
+      "epoch": 9.504817721267507,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 3.961855567696434e-05,
+      "loss": 0.4596,
+      "step": 191370
+    },
+    {
+      "epoch": 9.505314393563127,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.957882189331479e-05,
+      "loss": 0.4928,
+      "step": 191380
+    },
+    {
+      "epoch": 9.505811065858746,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.953908810966524e-05,
+      "loss": 0.4832,
+      "step": 191390
+    },
+    {
+      "epoch": 9.506307738154366,
+      "grad_norm": 0.197265625,
+      "learning_rate": 3.9499354326015694e-05,
+      "loss": 0.4834,
+      "step": 191400
+    },
+    {
+      "epoch": 9.506804410449986,
+      "grad_norm": 0.197265625,
+      "learning_rate": 3.945962054236615e-05,
+      "loss": 0.4693,
+      "step": 191410
+    },
+    {
+      "epoch": 9.507301082745604,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.94198867587166e-05,
+      "loss": 0.4934,
+      "step": 191420
+    },
+    {
+      "epoch": 9.507797755041224,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.9380152975067054e-05,
+      "loss": 0.4893,
+      "step": 191430
+    },
+    {
+      "epoch": 9.508294427336843,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.9340419191417505e-05,
+      "loss": 0.4647,
+      "step": 191440
+    },
+    {
+      "epoch": 9.508791099632463,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.9300685407767956e-05,
+      "loss": 0.4931,
+      "step": 191450
+    },
+    {
+      "epoch": 9.509287771928081,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 3.926095162411841e-05,
+      "loss": 0.4797,
+      "step": 191460
+    },
+    {
+      "epoch": 9.509784444223701,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.922121784046886e-05,
+      "loss": 0.4645,
+      "step": 191470
+    },
+    {
+      "epoch": 9.510281116519321,
+      "grad_norm": 0.1953125,
+      "learning_rate": 3.918148405681931e-05,
+      "loss": 0.4774,
+      "step": 191480
+    },
+    {
+      "epoch": 9.51077778881494,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 3.914175027316977e-05,
+      "loss": 0.5144,
+      "step": 191490
+    },
+    {
+      "epoch": 9.51127446111056,
+      "grad_norm": 0.216796875,
+      "learning_rate": 3.910201648952022e-05,
+      "loss": 0.5031,
+      "step": 191500
+    },
+    {
+      "epoch": 9.511771133406178,
+      "grad_norm": 0.23046875,
+      "learning_rate": 3.906228270587067e-05,
+      "loss": 0.4978,
+      "step": 191510
+    },
+    {
+      "epoch": 9.512267805701798,
+      "grad_norm": 0.203125,
+      "learning_rate": 3.902254892222112e-05,
+      "loss": 0.4938,
+      "step": 191520
+    },
+    {
+      "epoch": 9.512764477997417,
+      "grad_norm": 0.197265625,
+      "learning_rate": 3.898281513857157e-05,
+      "loss": 0.4878,
+      "step": 191530
+    },
+    {
+      "epoch": 9.513261150293037,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.894308135492202e-05,
+      "loss": 0.4879,
+      "step": 191540
+    },
+    {
+      "epoch": 9.513757822588657,
+      "grad_norm": 0.21484375,
+      "learning_rate": 3.890334757127247e-05,
+      "loss": 0.5107,
+      "step": 191550
+    },
+    {
+      "epoch": 9.514254494884275,
+      "grad_norm": 0.2109375,
+      "learning_rate": 3.8863613787622924e-05,
+      "loss": 0.457,
+      "step": 191560
+    },
+    {
+      "epoch": 9.514751167179895,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.8823880003973375e-05,
+      "loss": 0.4798,
+      "step": 191570
+    },
+    {
+      "epoch": 9.515247839475514,
+      "grad_norm": 0.197265625,
+      "learning_rate": 3.878414622032383e-05,
+      "loss": 0.4988,
+      "step": 191580
+    },
+    {
+      "epoch": 9.515744511771134,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.8744412436674284e-05,
+      "loss": 0.4823,
+      "step": 191590
+    },
+    {
+      "epoch": 9.516241184066752,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 3.8704678653024735e-05,
+      "loss": 0.4892,
+      "step": 191600
+    },
+    {
+      "epoch": 9.516737856362372,
+      "grad_norm": 0.189453125,
+      "learning_rate": 3.8664944869375186e-05,
+      "loss": 0.4959,
+      "step": 191610
+    },
+    {
+      "epoch": 9.517234528657992,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.862521108572564e-05,
+      "loss": 0.479,
+      "step": 191620
+    },
+    {
+      "epoch": 9.51773120095361,
+      "grad_norm": 0.197265625,
+      "learning_rate": 3.858547730207609e-05,
+      "loss": 0.4896,
+      "step": 191630
+    },
+    {
+      "epoch": 9.51822787324923,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 3.854574351842654e-05,
+      "loss": 0.4593,
+      "step": 191640
+    },
+    {
+      "epoch": 9.518724545544849,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 3.850600973477699e-05,
+      "loss": 0.5126,
+      "step": 191650
+    },
+    {
+      "epoch": 9.51922121784047,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 3.846627595112745e-05,
+      "loss": 0.513,
+      "step": 191660
+    },
+    {
+      "epoch": 9.519717890136087,
+      "grad_norm": 0.205078125,
+      "learning_rate": 3.84265421674779e-05,
+      "loss": 0.4827,
+      "step": 191670
+    },
+    {
+      "epoch": 9.520214562431708,
+      "grad_norm": 0.201171875,
+      "learning_rate": 3.838680838382836e-05,
+      "loss": 0.467,
+      "step": 191680
+    },
+    {
+      "epoch": 9.520711234727328,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.834707460017881e-05,
+      "loss": 0.4703,
+      "step": 191690
+    },
+    {
+      "epoch": 9.521207907022946,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.830734081652926e-05,
+      "loss": 0.4589,
+      "step": 191700
+    },
+    {
+      "epoch": 9.521704579318566,
+      "grad_norm": 0.197265625,
+      "learning_rate": 3.826760703287971e-05,
+      "loss": 0.4747,
+      "step": 191710
+    },
+    {
+      "epoch": 9.522201251614185,
+      "grad_norm": 0.201171875,
+      "learning_rate": 3.822787324923016e-05,
+      "loss": 0.4633,
+      "step": 191720
+    },
+    {
+      "epoch": 9.522697923909805,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.818813946558061e-05,
+      "loss": 0.49,
+      "step": 191730
+    },
+    {
+      "epoch": 9.523194596205423,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 3.814840568193106e-05,
+      "loss": 0.5207,
+      "step": 191740
+    },
+    {
+      "epoch": 9.523691268501043,
+      "grad_norm": 0.1953125,
+      "learning_rate": 3.8108671898281514e-05,
+      "loss": 0.4922,
+      "step": 191750
+    },
+    {
+      "epoch": 9.524187940796661,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.806893811463197e-05,
+      "loss": 0.4768,
+      "step": 191760
+    },
+    {
+      "epoch": 9.524684613092282,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.802920433098242e-05,
+      "loss": 0.5086,
+      "step": 191770
+    },
+    {
+      "epoch": 9.525181285387902,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 3.7989470547332874e-05,
+      "loss": 0.4887,
+      "step": 191780
+    },
+    {
+      "epoch": 9.52567795768352,
+      "grad_norm": 0.201171875,
+      "learning_rate": 3.7949736763683325e-05,
+      "loss": 0.4955,
+      "step": 191790
+    },
+    {
+      "epoch": 9.52617462997914,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 3.7910002980033776e-05,
+      "loss": 0.4866,
+      "step": 191800
+    },
+    {
+      "epoch": 9.526671302274758,
+      "grad_norm": 0.21484375,
+      "learning_rate": 3.787026919638423e-05,
+      "loss": 0.4779,
+      "step": 191810
+    },
+    {
+      "epoch": 9.527167974570379,
+      "grad_norm": 0.232421875,
+      "learning_rate": 3.783053541273468e-05,
+      "loss": 0.4713,
+      "step": 191820
+    },
+    {
+      "epoch": 9.527664646865997,
+      "grad_norm": 0.20703125,
+      "learning_rate": 3.779080162908513e-05,
+      "loss": 0.5013,
+      "step": 191830
+    },
+    {
+      "epoch": 9.528161319161617,
+      "grad_norm": 0.193359375,
+      "learning_rate": 3.775106784543559e-05,
+      "loss": 0.5004,
+      "step": 191840
+    },
+    {
+      "epoch": 9.528657991457237,
+      "grad_norm": 0.236328125,
+      "learning_rate": 3.771133406178604e-05,
+      "loss": 0.5178,
+      "step": 191850
+    },
+    {
+      "epoch": 9.529154663752855,
+      "grad_norm": 0.1953125,
+      "learning_rate": 3.767160027813649e-05,
+      "loss": 0.4951,
+      "step": 191860
+    },
+    {
+      "epoch": 9.529651336048476,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 3.763186649448694e-05,
+      "loss": 0.4878,
+      "step": 191870
+    },
+    {
+      "epoch": 9.530148008344094,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 3.759213271083739e-05,
+      "loss": 0.489,
+      "step": 191880
+    },
+    {
+      "epoch": 9.530644680639714,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.755239892718784e-05,
+      "loss": 0.4764,
+      "step": 191890
+    },
+    {
+      "epoch": 9.531141352935332,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 3.751266514353829e-05,
+      "loss": 0.4773,
+      "step": 191900
+    },
+    {
+      "epoch": 9.531638025230952,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 3.7472931359888744e-05,
+      "loss": 0.4708,
+      "step": 191910
+    },
+    {
+      "epoch": 9.532134697526573,
+      "grad_norm": 0.20703125,
+      "learning_rate": 3.7433197576239195e-05,
+      "loss": 0.5081,
+      "step": 191920
+    },
+    {
+      "epoch": 9.532631369822191,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 3.739346379258965e-05,
+      "loss": 0.4885,
+      "step": 191930
+    },
+    {
+      "epoch": 9.533128042117811,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 3.7353730008940104e-05,
+      "loss": 0.4689,
+      "step": 191940
+    },
+    {
+      "epoch": 9.53362471441343,
+      "grad_norm": 0.203125,
+      "learning_rate": 3.7313996225290555e-05,
+      "loss": 0.4842,
+      "step": 191950
+    },
+    {
+      "epoch": 9.53412138670905,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.7274262441641006e-05,
+      "loss": 0.4605,
+      "step": 191960
+    },
+    {
+      "epoch": 9.534618059004668,
+      "grad_norm": 0.220703125,
+      "learning_rate": 3.723452865799146e-05,
+      "loss": 0.4799,
+      "step": 191970
+    },
+    {
+      "epoch": 9.535114731300288,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 3.719479487434191e-05,
+      "loss": 0.5146,
+      "step": 191980
+    },
+    {
+      "epoch": 9.535611403595908,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 3.715506109069236e-05,
+      "loss": 0.5379,
+      "step": 191990
+    },
+    {
+      "epoch": 9.536108075891526,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.711532730704281e-05,
+      "loss": 0.5096,
+      "step": 192000
+    },
+    {
+      "epoch": 9.536604748187147,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 3.707559352339327e-05,
+      "loss": 0.5056,
+      "step": 192010
+    },
+    {
+      "epoch": 9.537101420482765,
+      "grad_norm": 0.23046875,
+      "learning_rate": 3.703585973974372e-05,
+      "loss": 0.5013,
+      "step": 192020
+    },
+    {
+      "epoch": 9.537598092778385,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.699612595609417e-05,
+      "loss": 0.4694,
+      "step": 192030
+    },
+    {
+      "epoch": 9.538094765074003,
+      "grad_norm": 0.203125,
+      "learning_rate": 3.695639217244462e-05,
+      "loss": 0.4828,
+      "step": 192040
+    },
+    {
+      "epoch": 9.538591437369623,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.691665838879507e-05,
+      "loss": 0.5081,
+      "step": 192050
+    },
+    {
+      "epoch": 9.539088109665244,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 3.687692460514552e-05,
+      "loss": 0.5002,
+      "step": 192060
+    },
+    {
+      "epoch": 9.539584781960862,
+      "grad_norm": 0.208984375,
+      "learning_rate": 3.6837190821495974e-05,
+      "loss": 0.4973,
+      "step": 192070
+    },
+    {
+      "epoch": 9.540081454256482,
+      "grad_norm": 0.220703125,
+      "learning_rate": 3.679745703784643e-05,
+      "loss": 0.4881,
+      "step": 192080
+    },
+    {
+      "epoch": 9.5405781265521,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 3.675772325419688e-05,
+      "loss": 0.5215,
+      "step": 192090
+    },
+    {
+      "epoch": 9.54107479884772,
+      "grad_norm": 0.203125,
+      "learning_rate": 3.6717989470547334e-05,
+      "loss": 0.4756,
+      "step": 192100
+    },
+    {
+      "epoch": 9.541571471143339,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.667825568689779e-05,
+      "loss": 0.4757,
+      "step": 192110
+    },
+    {
+      "epoch": 9.542068143438959,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.663852190324824e-05,
+      "loss": 0.4859,
+      "step": 192120
+    },
+    {
+      "epoch": 9.542564815734579,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.6598788119598694e-05,
+      "loss": 0.4845,
+      "step": 192130
+    },
+    {
+      "epoch": 9.543061488030197,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.6559054335949145e-05,
+      "loss": 0.4964,
+      "step": 192140
+    },
+    {
+      "epoch": 9.543558160325817,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 3.6519320552299596e-05,
+      "loss": 0.4743,
+      "step": 192150
+    },
+    {
+      "epoch": 9.544054832621436,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.647958676865005e-05,
+      "loss": 0.4867,
+      "step": 192160
+    },
+    {
+      "epoch": 9.544551504917056,
+      "grad_norm": 0.201171875,
+      "learning_rate": 3.64398529850005e-05,
+      "loss": 0.49,
+      "step": 192170
+    },
+    {
+      "epoch": 9.545048177212674,
+      "grad_norm": 0.201171875,
+      "learning_rate": 3.640011920135095e-05,
+      "loss": 0.5057,
+      "step": 192180
+    },
+    {
+      "epoch": 9.545544849508294,
+      "grad_norm": 0.220703125,
+      "learning_rate": 3.636038541770141e-05,
+      "loss": 0.48,
+      "step": 192190
+    },
+    {
+      "epoch": 9.546041521803915,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.632065163405186e-05,
+      "loss": 0.4865,
+      "step": 192200
+    },
+    {
+      "epoch": 9.546538194099533,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.628091785040231e-05,
+      "loss": 0.5045,
+      "step": 192210
+    },
+    {
+      "epoch": 9.547034866395153,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.624118406675276e-05,
+      "loss": 0.5066,
+      "step": 192220
+    },
+    {
+      "epoch": 9.547531538690771,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 3.620145028310321e-05,
+      "loss": 0.5052,
+      "step": 192230
+    },
+    {
+      "epoch": 9.548028210986391,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.616171649945366e-05,
+      "loss": 0.5034,
+      "step": 192240
+    },
+    {
+      "epoch": 9.54852488328201,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 3.612198271580411e-05,
+      "loss": 0.5156,
+      "step": 192250
+    },
+    {
+      "epoch": 9.54902155557763,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.6082248932154564e-05,
+      "loss": 0.4992,
+      "step": 192260
+    },
+    {
+      "epoch": 9.54951822787325,
+      "grad_norm": 0.21484375,
+      "learning_rate": 3.6042515148505015e-05,
+      "loss": 0.5048,
+      "step": 192270
+    },
+    {
+      "epoch": 9.550014900168868,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.600278136485547e-05,
+      "loss": 0.4819,
+      "step": 192280
+    },
+    {
+      "epoch": 9.550511572464488,
+      "grad_norm": 0.2578125,
+      "learning_rate": 3.5963047581205924e-05,
+      "loss": 0.4945,
+      "step": 192290
+    },
+    {
+      "epoch": 9.551008244760107,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 3.5923313797556375e-05,
+      "loss": 0.4933,
+      "step": 192300
+    },
+    {
+      "epoch": 9.551504917055727,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 3.5883580013906826e-05,
+      "loss": 0.4834,
+      "step": 192310
+    },
+    {
+      "epoch": 9.552001589351345,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.584384623025728e-05,
+      "loss": 0.5017,
+      "step": 192320
+    },
+    {
+      "epoch": 9.552498261646965,
+      "grad_norm": 0.203125,
+      "learning_rate": 3.580411244660773e-05,
+      "loss": 0.4988,
+      "step": 192330
+    },
+    {
+      "epoch": 9.552994933942585,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 3.576437866295818e-05,
+      "loss": 0.4934,
+      "step": 192340
+    },
+    {
+      "epoch": 9.553491606238204,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.572464487930863e-05,
+      "loss": 0.5099,
+      "step": 192350
+    },
+    {
+      "epoch": 9.553988278533824,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 3.568491109565909e-05,
+      "loss": 0.5079,
+      "step": 192360
+    },
+    {
+      "epoch": 9.554484950829442,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.564517731200954e-05,
+      "loss": 0.4867,
+      "step": 192370
+    },
+    {
+      "epoch": 9.554981623125062,
+      "grad_norm": 0.20703125,
+      "learning_rate": 3.560544352835999e-05,
+      "loss": 0.4577,
+      "step": 192380
+    },
+    {
+      "epoch": 9.55547829542068,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.556570974471044e-05,
+      "loss": 0.4575,
+      "step": 192390
+    },
+    {
+      "epoch": 9.5559749677163,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 3.552597596106089e-05,
+      "loss": 0.4862,
+      "step": 192400
+    },
+    {
+      "epoch": 9.556471640011921,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 3.548624217741134e-05,
+      "loss": 0.4689,
+      "step": 192410
+    },
+    {
+      "epoch": 9.55696831230754,
+      "grad_norm": 0.2109375,
+      "learning_rate": 3.5446508393761794e-05,
+      "loss": 0.4932,
+      "step": 192420
+    },
+    {
+      "epoch": 9.55746498460316,
+      "grad_norm": 0.205078125,
+      "learning_rate": 3.5406774610112245e-05,
+      "loss": 0.4783,
+      "step": 192430
+    },
+    {
+      "epoch": 9.557961656898778,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.53670408264627e-05,
+      "loss": 0.4805,
+      "step": 192440
+    },
+    {
+      "epoch": 9.558458329194398,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.5327307042813154e-05,
+      "loss": 0.5025,
+      "step": 192450
+    },
+    {
+      "epoch": 9.558955001490016,
+      "grad_norm": 0.189453125,
+      "learning_rate": 3.5287573259163605e-05,
+      "loss": 0.4831,
+      "step": 192460
+    },
+    {
+      "epoch": 9.559451673785636,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.5247839475514056e-05,
+      "loss": 0.4607,
+      "step": 192470
+    },
+    {
+      "epoch": 9.559948346081256,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.520810569186451e-05,
+      "loss": 0.4851,
+      "step": 192480
+    },
+    {
+      "epoch": 9.560445018376875,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.516837190821496e-05,
+      "loss": 0.4971,
+      "step": 192490
+    },
+    {
+      "epoch": 9.560941690672495,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 3.5128638124565416e-05,
+      "loss": 0.4889,
+      "step": 192500
+    },
+    {
+      "epoch": 9.561438362968113,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.508890434091587e-05,
+      "loss": 0.4994,
+      "step": 192510
+    },
+    {
+      "epoch": 9.561935035263733,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.504917055726632e-05,
+      "loss": 0.4843,
+      "step": 192520
+    },
+    {
+      "epoch": 9.562431707559352,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.500943677361677e-05,
+      "loss": 0.4535,
+      "step": 192530
+    },
+    {
+      "epoch": 9.562928379854972,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.4969702989967227e-05,
+      "loss": 0.469,
+      "step": 192540
+    },
+    {
+      "epoch": 9.563425052150592,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.492996920631768e-05,
+      "loss": 0.4811,
+      "step": 192550
+    },
+    {
+      "epoch": 9.56392172444621,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 3.489023542266813e-05,
+      "loss": 0.4826,
+      "step": 192560
+    },
+    {
+      "epoch": 9.56441839674183,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.485050163901858e-05,
+      "loss": 0.4918,
+      "step": 192570
+    },
+    {
+      "epoch": 9.564915069037449,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.481076785536903e-05,
+      "loss": 0.4913,
+      "step": 192580
+    },
+    {
+      "epoch": 9.565411741333069,
+      "grad_norm": 0.19921875,
+      "learning_rate": 3.477103407171948e-05,
+      "loss": 0.4949,
+      "step": 192590
+    },
+    {
+      "epoch": 9.565908413628687,
+      "grad_norm": 0.201171875,
+      "learning_rate": 3.473130028806993e-05,
+      "loss": 0.5122,
+      "step": 192600
+    },
+    {
+      "epoch": 9.566405085924307,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 3.4691566504420384e-05,
+      "loss": 0.4734,
+      "step": 192610
+    },
+    {
+      "epoch": 9.566901758219927,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 3.4651832720770835e-05,
+      "loss": 0.4896,
+      "step": 192620
+    },
+    {
+      "epoch": 9.567398430515546,
+      "grad_norm": 0.1953125,
+      "learning_rate": 3.461209893712129e-05,
+      "loss": 0.5158,
+      "step": 192630
+    },
+    {
+      "epoch": 9.567895102811166,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 3.4572365153471744e-05,
+      "loss": 0.5142,
+      "step": 192640
+    },
+    {
+      "epoch": 9.568391775106784,
+      "grad_norm": 0.220703125,
+      "learning_rate": 3.4532631369822195e-05,
+      "loss": 0.4838,
+      "step": 192650
+    },
+    {
+      "epoch": 9.568888447402404,
+      "grad_norm": 0.205078125,
+      "learning_rate": 3.4492897586172646e-05,
+      "loss": 0.4967,
+      "step": 192660
+    },
+    {
+      "epoch": 9.569385119698023,
+      "grad_norm": 0.23046875,
+      "learning_rate": 3.44531638025231e-05,
+      "loss": 0.4832,
+      "step": 192670
+    },
+    {
+      "epoch": 9.569881791993643,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 3.441343001887355e-05,
+      "loss": 0.4955,
+      "step": 192680
+    },
+    {
+      "epoch": 9.570378464289263,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.4373696235224e-05,
+      "loss": 0.4461,
+      "step": 192690
+    },
+    {
+      "epoch": 9.570875136584881,
+      "grad_norm": 0.234375,
+      "learning_rate": 3.433396245157445e-05,
+      "loss": 0.4708,
+      "step": 192700
+    },
+    {
+      "epoch": 9.571371808880501,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.429422866792491e-05,
+      "loss": 0.468,
+      "step": 192710
+    },
+    {
+      "epoch": 9.57186848117612,
+      "grad_norm": 0.201171875,
+      "learning_rate": 3.425449488427536e-05,
+      "loss": 0.5008,
+      "step": 192720
+    },
+    {
+      "epoch": 9.57236515347174,
+      "grad_norm": 0.197265625,
+      "learning_rate": 3.421476110062581e-05,
+      "loss": 0.4814,
+      "step": 192730
+    },
+    {
+      "epoch": 9.572861825767358,
+      "grad_norm": 0.2109375,
+      "learning_rate": 3.417502731697626e-05,
+      "loss": 0.4937,
+      "step": 192740
+    },
+    {
+      "epoch": 9.573358498062978,
+      "grad_norm": 0.2109375,
+      "learning_rate": 3.413529353332671e-05,
+      "loss": 0.4934,
+      "step": 192750
+    },
+    {
+      "epoch": 9.573855170358597,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 3.409555974967716e-05,
+      "loss": 0.4747,
+      "step": 192760
+    },
+    {
+      "epoch": 9.574351842654217,
+      "grad_norm": 0.2578125,
+      "learning_rate": 3.4055825966027614e-05,
+      "loss": 0.5158,
+      "step": 192770
+    },
+    {
+      "epoch": 9.574848514949837,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.4016092182378065e-05,
+      "loss": 0.4607,
+      "step": 192780
+    },
+    {
+      "epoch": 9.575345187245455,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 3.397635839872852e-05,
+      "loss": 0.507,
+      "step": 192790
+    },
+    {
+      "epoch": 9.575841859541075,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 3.3936624615078974e-05,
+      "loss": 0.5017,
+      "step": 192800
+    },
+    {
+      "epoch": 9.576338531836694,
+      "grad_norm": 0.205078125,
+      "learning_rate": 3.3896890831429425e-05,
+      "loss": 0.5241,
+      "step": 192810
+    },
+    {
+      "epoch": 9.576835204132314,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.3857157047779876e-05,
+      "loss": 0.4668,
+      "step": 192820
+    },
+    {
+      "epoch": 9.577331876427932,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.381742326413033e-05,
+      "loss": 0.5134,
+      "step": 192830
+    },
+    {
+      "epoch": 9.577828548723552,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.377768948048078e-05,
+      "loss": 0.4805,
+      "step": 192840
+    },
+    {
+      "epoch": 9.578325221019172,
+      "grad_norm": 0.216796875,
+      "learning_rate": 3.373795569683123e-05,
+      "loss": 0.4943,
+      "step": 192850
+    },
+    {
+      "epoch": 9.57882189331479,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 3.369822191318168e-05,
+      "loss": 0.4839,
+      "step": 192860
+    },
+    {
+      "epoch": 9.57931856561041,
+      "grad_norm": 0.201171875,
+      "learning_rate": 3.365848812953213e-05,
+      "loss": 0.4885,
+      "step": 192870
+    },
+    {
+      "epoch": 9.579815237906029,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.361875434588259e-05,
+      "loss": 0.4661,
+      "step": 192880
+    },
+    {
+      "epoch": 9.58031191020165,
+      "grad_norm": 0.201171875,
+      "learning_rate": 3.357902056223304e-05,
+      "loss": 0.4859,
+      "step": 192890
+    },
+    {
+      "epoch": 9.580808582497268,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 3.353928677858349e-05,
+      "loss": 0.4815,
+      "step": 192900
+    },
+    {
+      "epoch": 9.581305254792888,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 3.349955299493394e-05,
+      "loss": 0.5095,
+      "step": 192910
+    },
+    {
+      "epoch": 9.581801927088508,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 3.34598192112844e-05,
+      "loss": 0.4854,
+      "step": 192920
+    },
+    {
+      "epoch": 9.582298599384126,
+      "grad_norm": 0.236328125,
+      "learning_rate": 3.342008542763485e-05,
+      "loss": 0.5088,
+      "step": 192930
+    },
+    {
+      "epoch": 9.582795271679746,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.33803516439853e-05,
+      "loss": 0.4978,
+      "step": 192940
+    },
+    {
+      "epoch": 9.583291943975365,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.334061786033575e-05,
+      "loss": 0.5023,
+      "step": 192950
+    },
+    {
+      "epoch": 9.583788616270985,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.3300884076686204e-05,
+      "loss": 0.4632,
+      "step": 192960
+    },
+    {
+      "epoch": 9.584285288566603,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.326115029303666e-05,
+      "loss": 0.4744,
+      "step": 192970
+    },
+    {
+      "epoch": 9.584781960862223,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.322141650938711e-05,
+      "loss": 0.4646,
+      "step": 192980
+    },
+    {
+      "epoch": 9.585278633157843,
+      "grad_norm": 0.23046875,
+      "learning_rate": 3.3181682725737563e-05,
+      "loss": 0.4881,
+      "step": 192990
+    },
+    {
+      "epoch": 9.585775305453462,
+      "grad_norm": 0.23828125,
+      "learning_rate": 3.3141948942088014e-05,
+      "loss": 0.4947,
+      "step": 193000
+    },
+    {
+      "epoch": 9.586271977749082,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 3.3102215158438466e-05,
+      "loss": 0.4865,
+      "step": 193010
+    },
+    {
+      "epoch": 9.5867686500447,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 3.3062481374788917e-05,
+      "loss": 0.4791,
+      "step": 193020
+    },
+    {
+      "epoch": 9.58726532234032,
+      "grad_norm": 0.224609375,
+      "learning_rate": 3.302274759113937e-05,
+      "loss": 0.4772,
+      "step": 193030
+    },
+    {
+      "epoch": 9.587761994635938,
+      "grad_norm": 0.26171875,
+      "learning_rate": 3.298301380748982e-05,
+      "loss": 0.5063,
+      "step": 193040
+    },
+    {
+      "epoch": 9.588258666931559,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 3.294328002384027e-05,
+      "loss": 0.4819,
+      "step": 193050
+    },
+    {
+      "epoch": 9.588755339227179,
+      "grad_norm": 0.1953125,
+      "learning_rate": 3.290354624019073e-05,
+      "loss": 0.5097,
+      "step": 193060
+    },
+    {
+      "epoch": 9.589252011522797,
+      "grad_norm": 0.21484375,
+      "learning_rate": 3.286381245654118e-05,
+      "loss": 0.4994,
+      "step": 193070
+    },
+    {
+      "epoch": 9.589748683818417,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.282407867289163e-05,
+      "loss": 0.4718,
+      "step": 193080
+    },
+    {
+      "epoch": 9.590245356114036,
+      "grad_norm": 0.24609375,
+      "learning_rate": 3.278434488924208e-05,
+      "loss": 0.5238,
+      "step": 193090
+    },
+    {
+      "epoch": 9.590742028409656,
+      "grad_norm": 0.205078125,
+      "learning_rate": 3.274461110559253e-05,
+      "loss": 0.4581,
+      "step": 193100
+    },
+    {
+      "epoch": 9.591238700705274,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 3.270487732194298e-05,
+      "loss": 0.454,
+      "step": 193110
+    },
+    {
+      "epoch": 9.591735373000894,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 3.2665143538293434e-05,
+      "loss": 0.4849,
+      "step": 193120
+    },
+    {
+      "epoch": 9.592232045296514,
+      "grad_norm": 0.197265625,
+      "learning_rate": 3.2625409754643885e-05,
+      "loss": 0.5066,
+      "step": 193130
+    },
+    {
+      "epoch": 9.592728717592133,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.258567597099434e-05,
+      "loss": 0.4867,
+      "step": 193140
+    },
+    {
+      "epoch": 9.593225389887753,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.2545942187344793e-05,
+      "loss": 0.4706,
+      "step": 193150
+    },
+    {
+      "epoch": 9.593722062183371,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 3.2506208403695244e-05,
+      "loss": 0.4917,
+      "step": 193160
+    },
+    {
+      "epoch": 9.594218734478991,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.2466474620045696e-05,
+      "loss": 0.4563,
+      "step": 193170
+    },
+    {
+      "epoch": 9.59471540677461,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.2426740836396147e-05,
+      "loss": 0.4949,
+      "step": 193180
+    },
+    {
+      "epoch": 9.59521207907023,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.23870070527466e-05,
+      "loss": 0.5061,
+      "step": 193190
+    },
+    {
+      "epoch": 9.595708751365848,
+      "grad_norm": 0.205078125,
+      "learning_rate": 3.234727326909705e-05,
+      "loss": 0.4821,
+      "step": 193200
+    },
+    {
+      "epoch": 9.596205423661468,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.23075394854475e-05,
+      "loss": 0.4701,
+      "step": 193210
+    },
+    {
+      "epoch": 9.596702095957088,
+      "grad_norm": 0.224609375,
+      "learning_rate": 3.226780570179796e-05,
+      "loss": 0.5114,
+      "step": 193220
+    },
+    {
+      "epoch": 9.597198768252706,
+      "grad_norm": 0.205078125,
+      "learning_rate": 3.222807191814841e-05,
+      "loss": 0.4598,
+      "step": 193230
+    },
+    {
+      "epoch": 9.597695440548327,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 3.218833813449886e-05,
+      "loss": 0.4873,
+      "step": 193240
+    },
+    {
+      "epoch": 9.598192112843945,
+      "grad_norm": 0.21484375,
+      "learning_rate": 3.214860435084931e-05,
+      "loss": 0.4576,
+      "step": 193250
+    },
+    {
+      "epoch": 9.598688785139565,
+      "grad_norm": 0.236328125,
+      "learning_rate": 3.210887056719976e-05,
+      "loss": 0.5117,
+      "step": 193260
+    },
+    {
+      "epoch": 9.599185457435183,
+      "grad_norm": 0.20703125,
+      "learning_rate": 3.206913678355021e-05,
+      "loss": 0.4732,
+      "step": 193270
+    },
+    {
+      "epoch": 9.599682129730803,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 3.2029402999900664e-05,
+      "loss": 0.4554,
+      "step": 193280
+    },
+    {
+      "epoch": 9.600178802026424,
+      "grad_norm": 0.2265625,
+      "learning_rate": 3.1989669216251115e-05,
+      "loss": 0.4936,
+      "step": 193290
+    },
+    {
+      "epoch": 9.600675474322042,
+      "grad_norm": 0.197265625,
+      "learning_rate": 3.1949935432601566e-05,
+      "loss": 0.4698,
+      "step": 193300
+    },
+    {
+      "epoch": 9.601172146617662,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 3.1910201648952023e-05,
+      "loss": 0.4765,
+      "step": 193310
+    },
+    {
+      "epoch": 9.60166881891328,
+      "grad_norm": 0.2109375,
+      "learning_rate": 3.1870467865302474e-05,
+      "loss": 0.487,
+      "step": 193320
+    },
+    {
+      "epoch": 9.6021654912089,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.1830734081652925e-05,
+      "loss": 0.501,
+      "step": 193330
+    },
+    {
+      "epoch": 9.602662163504519,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 3.179100029800338e-05,
+      "loss": 0.4816,
+      "step": 193340
+    },
+    {
+      "epoch": 9.603158835800139,
+      "grad_norm": 0.23046875,
+      "learning_rate": 3.1751266514353834e-05,
+      "loss": 0.4811,
+      "step": 193350
+    },
+    {
+      "epoch": 9.603655508095759,
+      "grad_norm": 0.19921875,
+      "learning_rate": 3.1711532730704285e-05,
+      "loss": 0.4869,
+      "step": 193360
+    },
+    {
+      "epoch": 9.604152180391377,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 3.1671798947054736e-05,
+      "loss": 0.4844,
+      "step": 193370
+    },
+    {
+      "epoch": 9.604648852686998,
+      "grad_norm": 0.228515625,
+      "learning_rate": 3.163206516340519e-05,
+      "loss": 0.5069,
+      "step": 193380
+    },
+    {
+      "epoch": 9.605145524982616,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 3.159233137975564e-05,
+      "loss": 0.5005,
+      "step": 193390
+    },
+    {
+      "epoch": 9.605642197278236,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 3.155259759610609e-05,
+      "loss": 0.5013,
+      "step": 193400
+    },
+    {
+      "epoch": 9.606138869573854,
+      "grad_norm": 0.2109375,
+      "learning_rate": 3.151286381245655e-05,
+      "loss": 0.5013,
+      "step": 193410
+    },
+    {
+      "epoch": 9.606635541869474,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 3.1473130028807e-05,
+      "loss": 0.505,
+      "step": 193420
+    },
+    {
+      "epoch": 9.607132214165095,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.143339624515745e-05,
+      "loss": 0.4849,
+      "step": 193430
+    },
+    {
+      "epoch": 9.607628886460713,
+      "grad_norm": 0.20703125,
+      "learning_rate": 3.13936624615079e-05,
+      "loss": 0.5046,
+      "step": 193440
+    },
+    {
+      "epoch": 9.608125558756333,
+      "grad_norm": 0.203125,
+      "learning_rate": 3.135392867785835e-05,
+      "loss": 0.4999,
+      "step": 193450
+    },
+    {
+      "epoch": 9.608622231051951,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.13141948942088e-05,
+      "loss": 0.4982,
+      "step": 193460
+    },
+    {
+      "epoch": 9.609118903347571,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.1274461110559253e-05,
+      "loss": 0.5025,
+      "step": 193470
+    },
+    {
+      "epoch": 9.60961557564319,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 3.1234727326909704e-05,
+      "loss": 0.4829,
+      "step": 193480
+    },
+    {
+      "epoch": 9.61011224793881,
+      "grad_norm": 0.234375,
+      "learning_rate": 3.119499354326016e-05,
+      "loss": 0.4738,
+      "step": 193490
+    },
+    {
+      "epoch": 9.61060892023443,
+      "grad_norm": 0.197265625,
+      "learning_rate": 3.115525975961061e-05,
+      "loss": 0.4664,
+      "step": 193500
+    },
+    {
+      "epoch": 9.611105592530048,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 3.1115525975961064e-05,
+      "loss": 0.4688,
+      "step": 193510
+    },
+    {
+      "epoch": 9.611602264825668,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 3.1075792192311515e-05,
+      "loss": 0.4797,
+      "step": 193520
+    },
+    {
+      "epoch": 9.612098937121287,
+      "grad_norm": 0.23828125,
+      "learning_rate": 3.1036058408661966e-05,
+      "loss": 0.4989,
+      "step": 193530
+    },
+    {
+      "epoch": 9.612595609416907,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 3.099632462501242e-05,
+      "loss": 0.5014,
+      "step": 193540
+    },
+    {
+      "epoch": 9.613092281712525,
+      "grad_norm": 0.2109375,
+      "learning_rate": 3.095659084136287e-05,
+      "loss": 0.4888,
+      "step": 193550
+    },
+    {
+      "epoch": 9.613588954008145,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.091685705771332e-05,
+      "loss": 0.5036,
+      "step": 193560
+    },
+    {
+      "epoch": 9.614085626303766,
+      "grad_norm": 0.201171875,
+      "learning_rate": 3.087712327406378e-05,
+      "loss": 0.4386,
+      "step": 193570
+    },
+    {
+      "epoch": 9.614582298599384,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.083738949041423e-05,
+      "loss": 0.5126,
+      "step": 193580
+    },
+    {
+      "epoch": 9.615078970895004,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 3.079765570676468e-05,
+      "loss": 0.4903,
+      "step": 193590
+    },
+    {
+      "epoch": 9.615575643190622,
+      "grad_norm": 0.22265625,
+      "learning_rate": 3.075792192311513e-05,
+      "loss": 0.4977,
+      "step": 193600
+    },
+    {
+      "epoch": 9.616072315486242,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.071818813946558e-05,
+      "loss": 0.479,
+      "step": 193610
+    },
+    {
+      "epoch": 9.61656898778186,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.067845435581603e-05,
+      "loss": 0.5186,
+      "step": 193620
+    },
+    {
+      "epoch": 9.61706566007748,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 3.0638720572166483e-05,
+      "loss": 0.52,
+      "step": 193630
+    },
+    {
+      "epoch": 9.617562332373101,
+      "grad_norm": 0.208984375,
+      "learning_rate": 3.0598986788516934e-05,
+      "loss": 0.4909,
+      "step": 193640
+    },
+    {
+      "epoch": 9.61805900466872,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.0559253004867385e-05,
+      "loss": 0.4971,
+      "step": 193650
+    },
+    {
+      "epoch": 9.61855567696434,
+      "grad_norm": 0.224609375,
+      "learning_rate": 3.051951922121784e-05,
+      "loss": 0.492,
+      "step": 193660
+    },
+    {
+      "epoch": 9.619052349259958,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.047978543756829e-05,
+      "loss": 0.4944,
+      "step": 193670
+    },
+    {
+      "epoch": 9.619549021555578,
+      "grad_norm": 0.275390625,
+      "learning_rate": 3.0440051653918745e-05,
+      "loss": 0.5009,
+      "step": 193680
+    },
+    {
+      "epoch": 9.620045693851196,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 3.0400317870269196e-05,
+      "loss": 0.5048,
+      "step": 193690
+    },
+    {
+      "epoch": 9.620542366146816,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.0360584086619647e-05,
+      "loss": 0.47,
+      "step": 193700
+    },
+    {
+      "epoch": 9.621039038442436,
+      "grad_norm": 0.189453125,
+      "learning_rate": 3.03208503029701e-05,
+      "loss": 0.5024,
+      "step": 193710
+    },
+    {
+      "epoch": 9.621535710738055,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.0281116519320553e-05,
+      "loss": 0.4578,
+      "step": 193720
+    },
+    {
+      "epoch": 9.622032383033675,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.0241382735671004e-05,
+      "loss": 0.5051,
+      "step": 193730
+    },
+    {
+      "epoch": 9.622529055329293,
+      "grad_norm": 0.2109375,
+      "learning_rate": 3.0201648952021455e-05,
+      "loss": 0.4967,
+      "step": 193740
+    },
+    {
+      "epoch": 9.623025727624913,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.0161915168371906e-05,
+      "loss": 0.4644,
+      "step": 193750
+    },
+    {
+      "epoch": 9.623522399920532,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 3.0122181384722364e-05,
+      "loss": 0.468,
+      "step": 193760
+    },
+    {
+      "epoch": 9.624019072216152,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 3.0082447601072815e-05,
+      "loss": 0.4674,
+      "step": 193770
+    },
+    {
+      "epoch": 9.624515744511772,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.004271381742327e-05,
+      "loss": 0.4686,
+      "step": 193780
+    },
+    {
+      "epoch": 9.62501241680739,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.000298003377372e-05,
+      "loss": 0.5239,
+      "step": 193790
+    },
+    {
+      "epoch": 9.62550908910301,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.996324625012417e-05,
+      "loss": 0.5107,
+      "step": 193800
+    },
+    {
+      "epoch": 9.626005761398629,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 2.9923512466474622e-05,
+      "loss": 0.5074,
+      "step": 193810
+    },
+    {
+      "epoch": 9.626502433694249,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.9883778682825077e-05,
+      "loss": 0.4984,
+      "step": 193820
+    },
+    {
+      "epoch": 9.626999105989867,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.9844044899175528e-05,
+      "loss": 0.4873,
+      "step": 193830
+    },
+    {
+      "epoch": 9.627495778285487,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.980431111552598e-05,
+      "loss": 0.5125,
+      "step": 193840
+    },
+    {
+      "epoch": 9.627992450581107,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.976457733187643e-05,
+      "loss": 0.4712,
+      "step": 193850
+    },
+    {
+      "epoch": 9.628489122876726,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.9724843548226884e-05,
+      "loss": 0.4807,
+      "step": 193860
+    },
+    {
+      "epoch": 9.628985795172346,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 2.9685109764577335e-05,
+      "loss": 0.4664,
+      "step": 193870
+    },
+    {
+      "epoch": 9.629482467467964,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.9645375980927786e-05,
+      "loss": 0.515,
+      "step": 193880
+    },
+    {
+      "epoch": 9.629979139763584,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.9605642197278237e-05,
+      "loss": 0.4993,
+      "step": 193890
+    },
+    {
+      "epoch": 9.630475812059203,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 2.956590841362869e-05,
+      "loss": 0.5217,
+      "step": 193900
+    },
+    {
+      "epoch": 9.630972484354823,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 2.9526174629979143e-05,
+      "loss": 0.4946,
+      "step": 193910
+    },
+    {
+      "epoch": 9.631469156650443,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.9486440846329594e-05,
+      "loss": 0.4654,
+      "step": 193920
+    },
+    {
+      "epoch": 9.631965828946061,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.9446707062680045e-05,
+      "loss": 0.4723,
+      "step": 193930
+    },
+    {
+      "epoch": 9.632462501241681,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.94069732790305e-05,
+      "loss": 0.5216,
+      "step": 193940
+    },
+    {
+      "epoch": 9.6329591735373,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.936723949538095e-05,
+      "loss": 0.4662,
+      "step": 193950
+    },
+    {
+      "epoch": 9.63345584583292,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.93275057117314e-05,
+      "loss": 0.4802,
+      "step": 193960
+    },
+    {
+      "epoch": 9.633952518128538,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.9287771928081852e-05,
+      "loss": 0.4854,
+      "step": 193970
+    },
+    {
+      "epoch": 9.634449190424158,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 2.9248038144432303e-05,
+      "loss": 0.5109,
+      "step": 193980
+    },
+    {
+      "epoch": 9.634945862719778,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.9208304360782758e-05,
+      "loss": 0.4891,
+      "step": 193990
+    },
+    {
+      "epoch": 9.635442535015397,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.916857057713321e-05,
+      "loss": 0.4895,
+      "step": 194000
+    },
+    {
+      "epoch": 9.635939207311017,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 2.912883679348366e-05,
+      "loss": 0.5063,
+      "step": 194010
+    },
+    {
+      "epoch": 9.636435879606635,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 2.908910300983411e-05,
+      "loss": 0.473,
+      "step": 194020
+    },
+    {
+      "epoch": 9.636932551902255,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.9049369226184565e-05,
+      "loss": 0.5293,
+      "step": 194030
+    },
+    {
+      "epoch": 9.637429224197874,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.9009635442535016e-05,
+      "loss": 0.478,
+      "step": 194040
+    },
+    {
+      "epoch": 9.637925896493494,
+      "grad_norm": 0.236328125,
+      "learning_rate": 2.8969901658885467e-05,
+      "loss": 0.4794,
+      "step": 194050
+    },
+    {
+      "epoch": 9.638422568789114,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.8930167875235918e-05,
+      "loss": 0.5039,
+      "step": 194060
+    },
+    {
+      "epoch": 9.638919241084732,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.8890434091586373e-05,
+      "loss": 0.4825,
+      "step": 194070
+    },
+    {
+      "epoch": 9.639415913380352,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 2.8850700307936824e-05,
+      "loss": 0.4791,
+      "step": 194080
+    },
+    {
+      "epoch": 9.63991258567597,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.8810966524287275e-05,
+      "loss": 0.4545,
+      "step": 194090
+    },
+    {
+      "epoch": 9.64040925797159,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.8771232740637726e-05,
+      "loss": 0.5037,
+      "step": 194100
+    },
+    {
+      "epoch": 9.640905930267209,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.873149895698818e-05,
+      "loss": 0.4948,
+      "step": 194110
+    },
+    {
+      "epoch": 9.64140260256283,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.869176517333863e-05,
+      "loss": 0.4682,
+      "step": 194120
+    },
+    {
+      "epoch": 9.64189927485845,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.8652031389689082e-05,
+      "loss": 0.4699,
+      "step": 194130
+    },
+    {
+      "epoch": 9.642395947154068,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.8612297606039533e-05,
+      "loss": 0.5437,
+      "step": 194140
+    },
+    {
+      "epoch": 9.642892619449688,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.8572563822389988e-05,
+      "loss": 0.4852,
+      "step": 194150
+    },
+    {
+      "epoch": 9.643389291745306,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.853283003874044e-05,
+      "loss": 0.5102,
+      "step": 194160
+    },
+    {
+      "epoch": 9.643885964040926,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.849309625509089e-05,
+      "loss": 0.5137,
+      "step": 194170
+    },
+    {
+      "epoch": 9.644382636336545,
+      "grad_norm": 0.228515625,
+      "learning_rate": 2.845336247144134e-05,
+      "loss": 0.4675,
+      "step": 194180
+    },
+    {
+      "epoch": 9.644879308632165,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.84136286877918e-05,
+      "loss": 0.4924,
+      "step": 194190
+    },
+    {
+      "epoch": 9.645375980927783,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.837389490414225e-05,
+      "loss": 0.5043,
+      "step": 194200
+    },
+    {
+      "epoch": 9.645872653223403,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.8334161120492704e-05,
+      "loss": 0.4471,
+      "step": 194210
+    },
+    {
+      "epoch": 9.646369325519023,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 2.8294427336843155e-05,
+      "loss": 0.4832,
+      "step": 194220
+    },
+    {
+      "epoch": 9.646865997814642,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.8254693553193606e-05,
+      "loss": 0.487,
+      "step": 194230
+    },
+    {
+      "epoch": 9.647362670110262,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.8214959769544057e-05,
+      "loss": 0.4722,
+      "step": 194240
+    },
+    {
+      "epoch": 9.64785934240588,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.817522598589451e-05,
+      "loss": 0.4637,
+      "step": 194250
+    },
+    {
+      "epoch": 9.6483560147015,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.8135492202244962e-05,
+      "loss": 0.4972,
+      "step": 194260
+    },
+    {
+      "epoch": 9.648852686997119,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 2.8095758418595413e-05,
+      "loss": 0.4992,
+      "step": 194270
+    },
+    {
+      "epoch": 9.649349359292739,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.8056024634945865e-05,
+      "loss": 0.4637,
+      "step": 194280
+    },
+    {
+      "epoch": 9.649846031588359,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.801629085129632e-05,
+      "loss": 0.4976,
+      "step": 194290
+    },
+    {
+      "epoch": 9.650342703883977,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.797655706764677e-05,
+      "loss": 0.4965,
+      "step": 194300
+    },
+    {
+      "epoch": 9.650839376179597,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.793682328399722e-05,
+      "loss": 0.4809,
+      "step": 194310
+    },
+    {
+      "epoch": 9.651336048475216,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.7897089500347672e-05,
+      "loss": 0.5083,
+      "step": 194320
+    },
+    {
+      "epoch": 9.651832720770836,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.7857355716698123e-05,
+      "loss": 0.4955,
+      "step": 194330
+    },
+    {
+      "epoch": 9.652329393066454,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 2.7817621933048577e-05,
+      "loss": 0.4847,
+      "step": 194340
+    },
+    {
+      "epoch": 9.652826065362074,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.777788814939903e-05,
+      "loss": 0.4941,
+      "step": 194350
+    },
+    {
+      "epoch": 9.653322737657694,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.773815436574948e-05,
+      "loss": 0.5079,
+      "step": 194360
+    },
+    {
+      "epoch": 9.653819409953313,
+      "grad_norm": 0.23828125,
+      "learning_rate": 2.769842058209993e-05,
+      "loss": 0.5025,
+      "step": 194370
+    },
+    {
+      "epoch": 9.654316082248933,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.7658686798450385e-05,
+      "loss": 0.4983,
+      "step": 194380
+    },
+    {
+      "epoch": 9.654812754544551,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 2.7618953014800836e-05,
+      "loss": 0.4958,
+      "step": 194390
+    },
+    {
+      "epoch": 9.655309426840171,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.7579219231151287e-05,
+      "loss": 0.4882,
+      "step": 194400
+    },
+    {
+      "epoch": 9.65580609913579,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 2.7539485447501738e-05,
+      "loss": 0.4729,
+      "step": 194410
+    },
+    {
+      "epoch": 9.65630277143141,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 2.7499751663852192e-05,
+      "loss": 0.4555,
+      "step": 194420
+    },
+    {
+      "epoch": 9.65679944372703,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 2.7460017880202643e-05,
+      "loss": 0.4827,
+      "step": 194430
+    },
+    {
+      "epoch": 9.657296116022648,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.7420284096553094e-05,
+      "loss": 0.5014,
+      "step": 194440
+    },
+    {
+      "epoch": 9.657792788318268,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.7380550312903546e-05,
+      "loss": 0.4615,
+      "step": 194450
+    },
+    {
+      "epoch": 9.658289460613886,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 2.7340816529254e-05,
+      "loss": 0.4873,
+      "step": 194460
+    },
+    {
+      "epoch": 9.658786132909507,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.730108274560445e-05,
+      "loss": 0.4867,
+      "step": 194470
+    },
+    {
+      "epoch": 9.659282805205125,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.7261348961954902e-05,
+      "loss": 0.5168,
+      "step": 194480
+    },
+    {
+      "epoch": 9.659779477500745,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.7221615178305353e-05,
+      "loss": 0.4792,
+      "step": 194490
+    },
+    {
+      "epoch": 9.660276149796365,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 2.7181881394655807e-05,
+      "loss": 0.4787,
+      "step": 194500
+    },
+    {
+      "epoch": 9.660772822091984,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.714214761100626e-05,
+      "loss": 0.5178,
+      "step": 194510
+    },
+    {
+      "epoch": 9.661269494387604,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 2.710241382735671e-05,
+      "loss": 0.4603,
+      "step": 194520
+    },
+    {
+      "epoch": 9.661766166683222,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.706268004370716e-05,
+      "loss": 0.5243,
+      "step": 194530
+    },
+    {
+      "epoch": 9.662262838978842,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.7022946260057615e-05,
+      "loss": 0.4726,
+      "step": 194540
+    },
+    {
+      "epoch": 9.66275951127446,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.6983212476408066e-05,
+      "loss": 0.488,
+      "step": 194550
+    },
+    {
+      "epoch": 9.66325618357008,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 2.6943478692758517e-05,
+      "loss": 0.5003,
+      "step": 194560
+    },
+    {
+      "epoch": 9.663752855865699,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.6903744909108968e-05,
+      "loss": 0.4723,
+      "step": 194570
+    },
+    {
+      "epoch": 9.664249528161319,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.686401112545942e-05,
+      "loss": 0.4933,
+      "step": 194580
+    },
+    {
+      "epoch": 9.664746200456939,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.6824277341809873e-05,
+      "loss": 0.4836,
+      "step": 194590
+    },
+    {
+      "epoch": 9.665242872752557,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.6784543558160324e-05,
+      "loss": 0.4715,
+      "step": 194600
+    },
+    {
+      "epoch": 9.665739545048178,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.6744809774510782e-05,
+      "loss": 0.4682,
+      "step": 194610
+    },
+    {
+      "epoch": 9.666236217343796,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.6705075990861233e-05,
+      "loss": 0.5242,
+      "step": 194620
+    },
+    {
+      "epoch": 9.666732889639416,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 2.6665342207211684e-05,
+      "loss": 0.5015,
+      "step": 194630
+    },
+    {
+      "epoch": 9.667229561935034,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.662560842356214e-05,
+      "loss": 0.504,
+      "step": 194640
+    },
+    {
+      "epoch": 9.667726234230654,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 2.658587463991259e-05,
+      "loss": 0.467,
+      "step": 194650
+    },
+    {
+      "epoch": 9.668222906526275,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.654614085626304e-05,
+      "loss": 0.4887,
+      "step": 194660
+    },
+    {
+      "epoch": 9.668719578821893,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.6506407072613492e-05,
+      "loss": 0.4907,
+      "step": 194670
+    },
+    {
+      "epoch": 9.669216251117513,
+      "grad_norm": 0.234375,
+      "learning_rate": 2.6466673288963946e-05,
+      "loss": 0.5036,
+      "step": 194680
+    },
+    {
+      "epoch": 9.669712923413131,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.6426939505314397e-05,
+      "loss": 0.4953,
+      "step": 194690
+    },
+    {
+      "epoch": 9.670209595708751,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.6387205721664848e-05,
+      "loss": 0.5137,
+      "step": 194700
+    },
+    {
+      "epoch": 9.67070626800437,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.63474719380153e-05,
+      "loss": 0.4861,
+      "step": 194710
+    },
+    {
+      "epoch": 9.67120294029999,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.630773815436575e-05,
+      "loss": 0.4877,
+      "step": 194720
+    },
+    {
+      "epoch": 9.67169961259561,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.6268004370716205e-05,
+      "loss": 0.486,
+      "step": 194730
+    },
+    {
+      "epoch": 9.672196284891228,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.6228270587066656e-05,
+      "loss": 0.4809,
+      "step": 194740
+    },
+    {
+      "epoch": 9.672692957186849,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.6188536803417107e-05,
+      "loss": 0.4748,
+      "step": 194750
+    },
+    {
+      "epoch": 9.673189629482467,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.6148803019767558e-05,
+      "loss": 0.4607,
+      "step": 194760
+    },
+    {
+      "epoch": 9.673686301778087,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.6109069236118012e-05,
+      "loss": 0.4913,
+      "step": 194770
+    },
+    {
+      "epoch": 9.674182974073705,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.6069335452468463e-05,
+      "loss": 0.4711,
+      "step": 194780
+    },
+    {
+      "epoch": 9.674679646369325,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.6029601668818914e-05,
+      "loss": 0.4707,
+      "step": 194790
+    },
+    {
+      "epoch": 9.675176318664946,
+      "grad_norm": 0.25,
+      "learning_rate": 2.5989867885169365e-05,
+      "loss": 0.48,
+      "step": 194800
+    },
+    {
+      "epoch": 9.675672990960564,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.595013410151982e-05,
+      "loss": 0.4536,
+      "step": 194810
+    },
+    {
+      "epoch": 9.676169663256184,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.591040031787027e-05,
+      "loss": 0.4831,
+      "step": 194820
+    },
+    {
+      "epoch": 9.676666335551802,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.5870666534220722e-05,
+      "loss": 0.4773,
+      "step": 194830
+    },
+    {
+      "epoch": 9.677163007847422,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.5830932750571173e-05,
+      "loss": 0.502,
+      "step": 194840
+    },
+    {
+      "epoch": 9.67765968014304,
+      "grad_norm": 0.23046875,
+      "learning_rate": 2.5791198966921627e-05,
+      "loss": 0.4914,
+      "step": 194850
+    },
+    {
+      "epoch": 9.678156352438661,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.5751465183272078e-05,
+      "loss": 0.4692,
+      "step": 194860
+    },
+    {
+      "epoch": 9.678653024734281,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.571173139962253e-05,
+      "loss": 0.476,
+      "step": 194870
+    },
+    {
+      "epoch": 9.6791496970299,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.567199761597298e-05,
+      "loss": 0.4852,
+      "step": 194880
+    },
+    {
+      "epoch": 9.67964636932552,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.5632263832323435e-05,
+      "loss": 0.5134,
+      "step": 194890
+    },
+    {
+      "epoch": 9.680143041621138,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.5592530048673886e-05,
+      "loss": 0.4981,
+      "step": 194900
+    },
+    {
+      "epoch": 9.680639713916758,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.5552796265024337e-05,
+      "loss": 0.4656,
+      "step": 194910
+    },
+    {
+      "epoch": 9.681136386212376,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.5513062481374788e-05,
+      "loss": 0.4665,
+      "step": 194920
+    },
+    {
+      "epoch": 9.681633058507996,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.5473328697725242e-05,
+      "loss": 0.5203,
+      "step": 194930
+    },
+    {
+      "epoch": 9.682129730803616,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.5433594914075693e-05,
+      "loss": 0.526,
+      "step": 194940
+    },
+    {
+      "epoch": 9.682626403099235,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.5393861130426144e-05,
+      "loss": 0.5292,
+      "step": 194950
+    },
+    {
+      "epoch": 9.683123075394855,
+      "grad_norm": 0.23828125,
+      "learning_rate": 2.5354127346776595e-05,
+      "loss": 0.5179,
+      "step": 194960
+    },
+    {
+      "epoch": 9.683619747690473,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.5314393563127046e-05,
+      "loss": 0.4797,
+      "step": 194970
+    },
+    {
+      "epoch": 9.684116419986093,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.52746597794775e-05,
+      "loss": 0.4866,
+      "step": 194980
+    },
+    {
+      "epoch": 9.684613092281712,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.5234925995827952e-05,
+      "loss": 0.5057,
+      "step": 194990
+    },
+    {
+      "epoch": 9.685109764577332,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 2.5195192212178403e-05,
+      "loss": 0.5067,
+      "step": 195000
+    },
+    {
+      "epoch": 9.685606436872952,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 2.5155458428528854e-05,
+      "loss": 0.4849,
+      "step": 195010
+    },
+    {
+      "epoch": 9.68610310916857,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 2.5115724644879308e-05,
+      "loss": 0.4757,
+      "step": 195020
+    },
+    {
+      "epoch": 9.68659978146419,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.5075990861229766e-05,
+      "loss": 0.4971,
+      "step": 195030
+    },
+    {
+      "epoch": 9.687096453759809,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.5036257077580217e-05,
+      "loss": 0.5103,
+      "step": 195040
+    },
+    {
+      "epoch": 9.687593126055429,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.4996523293930665e-05,
+      "loss": 0.4784,
+      "step": 195050
+    },
+    {
+      "epoch": 9.688089798351047,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.4956789510281116e-05,
+      "loss": 0.501,
+      "step": 195060
+    },
+    {
+      "epoch": 9.688586470646667,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 2.491705572663157e-05,
+      "loss": 0.46,
+      "step": 195070
+    },
+    {
+      "epoch": 9.689083142942287,
+      "grad_norm": 0.2109375,
+      "learning_rate": 2.487732194298202e-05,
+      "loss": 0.5057,
+      "step": 195080
+    },
+    {
+      "epoch": 9.689579815237906,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 2.4837588159332472e-05,
+      "loss": 0.5015,
+      "step": 195090
+    },
+    {
+      "epoch": 9.690076487533526,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.4797854375682923e-05,
+      "loss": 0.4884,
+      "step": 195100
+    },
+    {
+      "epoch": 9.690573159829144,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.4758120592033378e-05,
+      "loss": 0.5,
+      "step": 195110
+    },
+    {
+      "epoch": 9.691069832124764,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.471838680838383e-05,
+      "loss": 0.496,
+      "step": 195120
+    },
+    {
+      "epoch": 9.691566504420383,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.4678653024734283e-05,
+      "loss": 0.5178,
+      "step": 195130
+    },
+    {
+      "epoch": 9.692063176716003,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.4638919241084734e-05,
+      "loss": 0.4895,
+      "step": 195140
+    },
+    {
+      "epoch": 9.692559849011623,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.4599185457435185e-05,
+      "loss": 0.4871,
+      "step": 195150
+    },
+    {
+      "epoch": 9.693056521307241,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.455945167378564e-05,
+      "loss": 0.5003,
+      "step": 195160
+    },
+    {
+      "epoch": 9.693553193602861,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.451971789013609e-05,
+      "loss": 0.4816,
+      "step": 195170
+    },
+    {
+      "epoch": 9.69404986589848,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.447998410648654e-05,
+      "loss": 0.4768,
+      "step": 195180
+    },
+    {
+      "epoch": 9.6945465381941,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.4440250322836993e-05,
+      "loss": 0.4494,
+      "step": 195190
+    },
+    {
+      "epoch": 9.695043210489718,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.4400516539187447e-05,
+      "loss": 0.4537,
+      "step": 195200
+    },
+    {
+      "epoch": 9.695539882785338,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.4360782755537898e-05,
+      "loss": 0.4809,
+      "step": 195210
+    },
+    {
+      "epoch": 9.696036555080958,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 2.432104897188835e-05,
+      "loss": 0.46,
+      "step": 195220
+    },
+    {
+      "epoch": 9.696533227376577,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.42813151882388e-05,
+      "loss": 0.4753,
+      "step": 195230
+    },
+    {
+      "epoch": 9.697029899672197,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.4241581404589255e-05,
+      "loss": 0.4959,
+      "step": 195240
+    },
+    {
+      "epoch": 9.697526571967815,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.4201847620939706e-05,
+      "loss": 0.4751,
+      "step": 195250
+    },
+    {
+      "epoch": 9.698023244263435,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.4162113837290157e-05,
+      "loss": 0.5048,
+      "step": 195260
+    },
+    {
+      "epoch": 9.698519916559054,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.4122380053640608e-05,
+      "loss": 0.4776,
+      "step": 195270
+    },
+    {
+      "epoch": 9.699016588854674,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.4082646269991062e-05,
+      "loss": 0.4882,
+      "step": 195280
+    },
+    {
+      "epoch": 9.699513261150294,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.4042912486341513e-05,
+      "loss": 0.4751,
+      "step": 195290
+    },
+    {
+      "epoch": 9.700009933445912,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.4003178702691964e-05,
+      "loss": 0.4944,
+      "step": 195300
+    },
+    {
+      "epoch": 9.700506605741532,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 2.3963444919042415e-05,
+      "loss": 0.5295,
+      "step": 195310
+    },
+    {
+      "epoch": 9.70100327803715,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.3923711135392866e-05,
+      "loss": 0.5129,
+      "step": 195320
+    },
+    {
+      "epoch": 9.70149995033277,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.388397735174332e-05,
+      "loss": 0.481,
+      "step": 195330
+    },
+    {
+      "epoch": 9.70199662262839,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.3844243568093775e-05,
+      "loss": 0.4714,
+      "step": 195340
+    },
+    {
+      "epoch": 9.70249329492401,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.3804509784444226e-05,
+      "loss": 0.5126,
+      "step": 195350
+    },
+    {
+      "epoch": 9.70298996721963,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.3764776000794677e-05,
+      "loss": 0.4717,
+      "step": 195360
+    },
+    {
+      "epoch": 9.703486639515248,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.3725042217145128e-05,
+      "loss": 0.5067,
+      "step": 195370
+    },
+    {
+      "epoch": 9.703983311810868,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.3685308433495582e-05,
+      "loss": 0.4728,
+      "step": 195380
+    },
+    {
+      "epoch": 9.704479984106486,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 2.3645574649846034e-05,
+      "loss": 0.4708,
+      "step": 195390
+    },
+    {
+      "epoch": 9.704976656402106,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.3605840866196485e-05,
+      "loss": 0.4978,
+      "step": 195400
+    },
+    {
+      "epoch": 9.705473328697725,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.3566107082546936e-05,
+      "loss": 0.491,
+      "step": 195410
+    },
+    {
+      "epoch": 9.705970000993345,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.352637329889739e-05,
+      "loss": 0.5006,
+      "step": 195420
+    },
+    {
+      "epoch": 9.706466673288965,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.348663951524784e-05,
+      "loss": 0.4791,
+      "step": 195430
+    },
+    {
+      "epoch": 9.706963345584583,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.3446905731598292e-05,
+      "loss": 0.5047,
+      "step": 195440
+    },
+    {
+      "epoch": 9.707460017880203,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.3407171947948743e-05,
+      "loss": 0.4874,
+      "step": 195450
+    },
+    {
+      "epoch": 9.707956690175822,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.3367438164299197e-05,
+      "loss": 0.4617,
+      "step": 195460
+    },
+    {
+      "epoch": 9.708453362471442,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.332770438064965e-05,
+      "loss": 0.4805,
+      "step": 195470
+    },
+    {
+      "epoch": 9.70895003476706,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.32879705970001e-05,
+      "loss": 0.5012,
+      "step": 195480
+    },
+    {
+      "epoch": 9.70944670706268,
+      "grad_norm": 0.19140625,
+      "learning_rate": 2.324823681335055e-05,
+      "loss": 0.4576,
+      "step": 195490
+    },
+    {
+      "epoch": 9.7099433793583,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.3208503029701005e-05,
+      "loss": 0.4666,
+      "step": 195500
+    },
+    {
+      "epoch": 9.710440051653919,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.3168769246051456e-05,
+      "loss": 0.4796,
+      "step": 195510
+    },
+    {
+      "epoch": 9.710936723949539,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 2.3129035462401907e-05,
+      "loss": 0.5225,
+      "step": 195520
+    },
+    {
+      "epoch": 9.711433396245157,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.3089301678752358e-05,
+      "loss": 0.5004,
+      "step": 195530
+    },
+    {
+      "epoch": 9.711930068540777,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.3049567895102812e-05,
+      "loss": 0.4764,
+      "step": 195540
+    },
+    {
+      "epoch": 9.712426740836396,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.3009834111453267e-05,
+      "loss": 0.496,
+      "step": 195550
+    },
+    {
+      "epoch": 9.712923413132016,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.2970100327803718e-05,
+      "loss": 0.4956,
+      "step": 195560
+    },
+    {
+      "epoch": 9.713420085427634,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.293036654415417e-05,
+      "loss": 0.4903,
+      "step": 195570
+    },
+    {
+      "epoch": 9.713916757723254,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.289063276050462e-05,
+      "loss": 0.4937,
+      "step": 195580
+    },
+    {
+      "epoch": 9.714413430018874,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.2850898976855074e-05,
+      "loss": 0.5052,
+      "step": 195590
+    },
+    {
+      "epoch": 9.714910102314493,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.2811165193205525e-05,
+      "loss": 0.514,
+      "step": 195600
+    },
+    {
+      "epoch": 9.715406774610113,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 2.2771431409555976e-05,
+      "loss": 0.5074,
+      "step": 195610
+    },
+    {
+      "epoch": 9.715903446905731,
+      "grad_norm": 0.23046875,
+      "learning_rate": 2.2731697625906427e-05,
+      "loss": 0.4891,
+      "step": 195620
+    },
+    {
+      "epoch": 9.716400119201351,
+      "grad_norm": 0.240234375,
+      "learning_rate": 2.2691963842256882e-05,
+      "loss": 0.4955,
+      "step": 195630
+    },
+    {
+      "epoch": 9.71689679149697,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 2.2652230058607333e-05,
+      "loss": 0.5105,
+      "step": 195640
+    },
+    {
+      "epoch": 9.71739346379259,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.2612496274957784e-05,
+      "loss": 0.4582,
+      "step": 195650
+    },
+    {
+      "epoch": 9.71789013608821,
+      "grad_norm": 0.19140625,
+      "learning_rate": 2.2572762491308235e-05,
+      "loss": 0.4924,
+      "step": 195660
+    },
+    {
+      "epoch": 9.718386808383828,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.2533028707658686e-05,
+      "loss": 0.4607,
+      "step": 195670
+    },
+    {
+      "epoch": 9.718883480679448,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.249329492400914e-05,
+      "loss": 0.5146,
+      "step": 195680
+    },
+    {
+      "epoch": 9.719380152975067,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.245356114035959e-05,
+      "loss": 0.4757,
+      "step": 195690
+    },
+    {
+      "epoch": 9.719876825270687,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 2.2413827356710042e-05,
+      "loss": 0.469,
+      "step": 195700
+    },
+    {
+      "epoch": 9.720373497566305,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.2374093573060493e-05,
+      "loss": 0.4821,
+      "step": 195710
+    },
+    {
+      "epoch": 9.720870169861925,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.2334359789410948e-05,
+      "loss": 0.5017,
+      "step": 195720
+    },
+    {
+      "epoch": 9.721366842157545,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.22946260057614e-05,
+      "loss": 0.5143,
+      "step": 195730
+    },
+    {
+      "epoch": 9.721863514453164,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.225489222211185e-05,
+      "loss": 0.5029,
+      "step": 195740
+    },
+    {
+      "epoch": 9.722360186748784,
+      "grad_norm": 0.2109375,
+      "learning_rate": 2.22151584384623e-05,
+      "loss": 0.4939,
+      "step": 195750
+    },
+    {
+      "epoch": 9.722856859044402,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.2175424654812755e-05,
+      "loss": 0.4829,
+      "step": 195760
+    },
+    {
+      "epoch": 9.723353531340022,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.213569087116321e-05,
+      "loss": 0.4811,
+      "step": 195770
+    },
+    {
+      "epoch": 9.72385020363564,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 2.209595708751366e-05,
+      "loss": 0.4619,
+      "step": 195780
+    },
+    {
+      "epoch": 9.72434687593126,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.2056223303864112e-05,
+      "loss": 0.4848,
+      "step": 195790
+    },
+    {
+      "epoch": 9.72484354822688,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.2016489520214563e-05,
+      "loss": 0.4892,
+      "step": 195800
+    },
+    {
+      "epoch": 9.725340220522499,
+      "grad_norm": 0.2109375,
+      "learning_rate": 2.1976755736565017e-05,
+      "loss": 0.4849,
+      "step": 195810
+    },
+    {
+      "epoch": 9.72583689281812,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.193702195291547e-05,
+      "loss": 0.4736,
+      "step": 195820
+    },
+    {
+      "epoch": 9.726333565113737,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.189728816926592e-05,
+      "loss": 0.5,
+      "step": 195830
+    },
+    {
+      "epoch": 9.726830237409358,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.185755438561637e-05,
+      "loss": 0.5078,
+      "step": 195840
+    },
+    {
+      "epoch": 9.727326909704976,
+      "grad_norm": 0.279296875,
+      "learning_rate": 2.1817820601966825e-05,
+      "loss": 0.4833,
+      "step": 195850
+    },
+    {
+      "epoch": 9.727823582000596,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.1778086818317276e-05,
+      "loss": 0.5043,
+      "step": 195860
+    },
+    {
+      "epoch": 9.728320254296216,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.1738353034667727e-05,
+      "loss": 0.4861,
+      "step": 195870
+    },
+    {
+      "epoch": 9.728816926591835,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.1698619251018178e-05,
+      "loss": 0.4772,
+      "step": 195880
+    },
+    {
+      "epoch": 9.729313598887455,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.1658885467368632e-05,
+      "loss": 0.5058,
+      "step": 195890
+    },
+    {
+      "epoch": 9.729810271183073,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.1619151683719083e-05,
+      "loss": 0.507,
+      "step": 195900
+    },
+    {
+      "epoch": 9.730306943478693,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 2.1579417900069534e-05,
+      "loss": 0.4756,
+      "step": 195910
+    },
+    {
+      "epoch": 9.730803615774311,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.1539684116419985e-05,
+      "loss": 0.4806,
+      "step": 195920
+    },
+    {
+      "epoch": 9.731300288069932,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.149995033277044e-05,
+      "loss": 0.4856,
+      "step": 195930
+    },
+    {
+      "epoch": 9.731796960365552,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.146021654912089e-05,
+      "loss": 0.4852,
+      "step": 195940
+    },
+    {
+      "epoch": 9.73229363266117,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.1420482765471342e-05,
+      "loss": 0.5089,
+      "step": 195950
+    },
+    {
+      "epoch": 9.73279030495679,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.1380748981821793e-05,
+      "loss": 0.4961,
+      "step": 195960
+    },
+    {
+      "epoch": 9.733286977252408,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.1341015198172244e-05,
+      "loss": 0.4767,
+      "step": 195970
+    },
+    {
+      "epoch": 9.733783649548029,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.1301281414522702e-05,
+      "loss": 0.4586,
+      "step": 195980
+    },
+    {
+      "epoch": 9.734280321843647,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.1261547630873153e-05,
+      "loss": 0.4875,
+      "step": 195990
+    },
+    {
+      "epoch": 9.734776994139267,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.1221813847223604e-05,
+      "loss": 0.5347,
+      "step": 196000
+    },
+    {
+      "epoch": 9.735273666434885,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.1182080063574055e-05,
+      "loss": 0.4966,
+      "step": 196010
+    },
+    {
+      "epoch": 9.735770338730505,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.114234627992451e-05,
+      "loss": 0.4585,
+      "step": 196020
+    },
+    {
+      "epoch": 9.736267011026126,
+      "grad_norm": 0.25,
+      "learning_rate": 2.110261249627496e-05,
+      "loss": 0.4704,
+      "step": 196030
+    },
+    {
+      "epoch": 9.736763683321744,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.106287871262541e-05,
+      "loss": 0.4908,
+      "step": 196040
+    },
+    {
+      "epoch": 9.737260355617364,
+      "grad_norm": 0.2109375,
+      "learning_rate": 2.1023144928975862e-05,
+      "loss": 0.4916,
+      "step": 196050
+    },
+    {
+      "epoch": 9.737757027912982,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.0983411145326313e-05,
+      "loss": 0.5012,
+      "step": 196060
+    },
+    {
+      "epoch": 9.738253700208602,
+      "grad_norm": 0.224609375,
+      "learning_rate": 2.0943677361676768e-05,
+      "loss": 0.4638,
+      "step": 196070
+    },
+    {
+      "epoch": 9.73875037250422,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.090394357802722e-05,
+      "loss": 0.4841,
+      "step": 196080
+    },
+    {
+      "epoch": 9.739247044799841,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.086420979437767e-05,
+      "loss": 0.4794,
+      "step": 196090
+    },
+    {
+      "epoch": 9.739743717095461,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.082447601072812e-05,
+      "loss": 0.5137,
+      "step": 196100
+    },
+    {
+      "epoch": 9.74024038939108,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.0784742227078575e-05,
+      "loss": 0.4988,
+      "step": 196110
+    },
+    {
+      "epoch": 9.7407370616867,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.0745008443429026e-05,
+      "loss": 0.4938,
+      "step": 196120
+    },
+    {
+      "epoch": 9.741233733982318,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.0705274659779477e-05,
+      "loss": 0.4864,
+      "step": 196130
+    },
+    {
+      "epoch": 9.741730406277938,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.0665540876129928e-05,
+      "loss": 0.4936,
+      "step": 196140
+    },
+    {
+      "epoch": 9.742227078573556,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.0625807092480383e-05,
+      "loss": 0.4587,
+      "step": 196150
+    },
+    {
+      "epoch": 9.742723750869176,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.0586073308830834e-05,
+      "loss": 0.4832,
+      "step": 196160
+    },
+    {
+      "epoch": 9.743220423164797,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.0546339525181285e-05,
+      "loss": 0.5016,
+      "step": 196170
+    },
+    {
+      "epoch": 9.743717095460415,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.0506605741531736e-05,
+      "loss": 0.4711,
+      "step": 196180
+    },
+    {
+      "epoch": 9.744213767756035,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.046687195788219e-05,
+      "loss": 0.48,
+      "step": 196190
+    },
+    {
+      "epoch": 9.744710440051653,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 2.0427138174232645e-05,
+      "loss": 0.4777,
+      "step": 196200
+    },
+    {
+      "epoch": 9.745207112347273,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.0387404390583096e-05,
+      "loss": 0.4614,
+      "step": 196210
+    },
+    {
+      "epoch": 9.745703784642892,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.0347670606933547e-05,
+      "loss": 0.5297,
+      "step": 196220
+    },
+    {
+      "epoch": 9.746200456938512,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.0307936823283998e-05,
+      "loss": 0.5064,
+      "step": 196230
+    },
+    {
+      "epoch": 9.746697129234132,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.0268203039634452e-05,
+      "loss": 0.4925,
+      "step": 196240
+    },
+    {
+      "epoch": 9.74719380152975,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.0228469255984903e-05,
+      "loss": 0.4921,
+      "step": 196250
+    },
+    {
+      "epoch": 9.74769047382537,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.0188735472335354e-05,
+      "loss": 0.4602,
+      "step": 196260
+    },
+    {
+      "epoch": 9.748187146120989,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.0149001688685805e-05,
+      "loss": 0.4966,
+      "step": 196270
+    },
+    {
+      "epoch": 9.748683818416609,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 2.010926790503626e-05,
+      "loss": 0.4766,
+      "step": 196280
+    },
+    {
+      "epoch": 9.749180490712227,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.006953412138671e-05,
+      "loss": 0.4918,
+      "step": 196290
+    },
+    {
+      "epoch": 9.749677163007847,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.002980033773716e-05,
+      "loss": 0.5133,
+      "step": 196300
+    },
+    {
+      "epoch": 9.750173835303467,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.9990066554087613e-05,
+      "loss": 0.4892,
+      "step": 196310
+    },
+    {
+      "epoch": 9.750670507599086,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.9950332770438067e-05,
+      "loss": 0.4952,
+      "step": 196320
+    },
+    {
+      "epoch": 9.751167179894706,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.9910598986788518e-05,
+      "loss": 0.5125,
+      "step": 196330
+    },
+    {
+      "epoch": 9.751663852190324,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.987086520313897e-05,
+      "loss": 0.5006,
+      "step": 196340
+    },
+    {
+      "epoch": 9.752160524485944,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.983113141948942e-05,
+      "loss": 0.5008,
+      "step": 196350
+    },
+    {
+      "epoch": 9.752657196781563,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.979139763583987e-05,
+      "loss": 0.4845,
+      "step": 196360
+    },
+    {
+      "epoch": 9.753153869077183,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.9751663852190326e-05,
+      "loss": 0.4881,
+      "step": 196370
+    },
+    {
+      "epoch": 9.753650541372803,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.9711930068540777e-05,
+      "loss": 0.5026,
+      "step": 196380
+    },
+    {
+      "epoch": 9.754147213668421,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.9672196284891228e-05,
+      "loss": 0.4911,
+      "step": 196390
+    },
+    {
+      "epoch": 9.754643885964041,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.9632462501241682e-05,
+      "loss": 0.467,
+      "step": 196400
+    },
+    {
+      "epoch": 9.75514055825966,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.9592728717592133e-05,
+      "loss": 0.47,
+      "step": 196410
+    },
+    {
+      "epoch": 9.75563723055528,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.9552994933942588e-05,
+      "loss": 0.4833,
+      "step": 196420
+    },
+    {
+      "epoch": 9.756133902850898,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.951326115029304e-05,
+      "loss": 0.4716,
+      "step": 196430
+    },
+    {
+      "epoch": 9.756630575146518,
+      "grad_norm": 0.234375,
+      "learning_rate": 1.947352736664349e-05,
+      "loss": 0.5068,
+      "step": 196440
+    },
+    {
+      "epoch": 9.757127247442138,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.943379358299394e-05,
+      "loss": 0.4756,
+      "step": 196450
+    },
+    {
+      "epoch": 9.757623919737757,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.9394059799344395e-05,
+      "loss": 0.4787,
+      "step": 196460
+    },
+    {
+      "epoch": 9.758120592033377,
+      "grad_norm": 0.19921875,
+      "learning_rate": 1.9354326015694846e-05,
+      "loss": 0.5143,
+      "step": 196470
+    },
+    {
+      "epoch": 9.758617264328995,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.9314592232045297e-05,
+      "loss": 0.4796,
+      "step": 196480
+    },
+    {
+      "epoch": 9.759113936624615,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.9274858448395748e-05,
+      "loss": 0.4623,
+      "step": 196490
+    },
+    {
+      "epoch": 9.759610608920234,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.9235124664746203e-05,
+      "loss": 0.5225,
+      "step": 196500
+    },
+    {
+      "epoch": 9.760107281215854,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.9195390881096654e-05,
+      "loss": 0.5526,
+      "step": 196510
+    },
+    {
+      "epoch": 9.760603953511474,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.9155657097447105e-05,
+      "loss": 0.4987,
+      "step": 196520
+    },
+    {
+      "epoch": 9.761100625807092,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.9115923313797556e-05,
+      "loss": 0.4952,
+      "step": 196530
+    },
+    {
+      "epoch": 9.761597298102712,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.907618953014801e-05,
+      "loss": 0.5011,
+      "step": 196540
+    },
+    {
+      "epoch": 9.76209397039833,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.903645574649846e-05,
+      "loss": 0.4748,
+      "step": 196550
+    },
+    {
+      "epoch": 9.76259064269395,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.8996721962848912e-05,
+      "loss": 0.5095,
+      "step": 196560
+    },
+    {
+      "epoch": 9.76308731498957,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.8956988179199363e-05,
+      "loss": 0.5107,
+      "step": 196570
+    },
+    {
+      "epoch": 9.76358398728519,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.8917254395549818e-05,
+      "loss": 0.5026,
+      "step": 196580
+    },
+    {
+      "epoch": 9.76408065958081,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.887752061190027e-05,
+      "loss": 0.512,
+      "step": 196590
+    },
+    {
+      "epoch": 9.764577331876428,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.883778682825072e-05,
+      "loss": 0.4859,
+      "step": 196600
+    },
+    {
+      "epoch": 9.765074004172048,
+      "grad_norm": 0.19921875,
+      "learning_rate": 1.8798053044601174e-05,
+      "loss": 0.4648,
+      "step": 196610
+    },
+    {
+      "epoch": 9.765570676467666,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.8758319260951625e-05,
+      "loss": 0.4673,
+      "step": 196620
+    },
+    {
+      "epoch": 9.766067348763286,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.871858547730208e-05,
+      "loss": 0.4805,
+      "step": 196630
+    },
+    {
+      "epoch": 9.766564021058905,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.867885169365253e-05,
+      "loss": 0.4962,
+      "step": 196640
+    },
+    {
+      "epoch": 9.767060693354525,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.863911791000298e-05,
+      "loss": 0.4537,
+      "step": 196650
+    },
+    {
+      "epoch": 9.767557365650145,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.8599384126353433e-05,
+      "loss": 0.484,
+      "step": 196660
+    },
+    {
+      "epoch": 9.768054037945763,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.8559650342703887e-05,
+      "loss": 0.4855,
+      "step": 196670
+    },
+    {
+      "epoch": 9.768550710241383,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.8519916559054338e-05,
+      "loss": 0.4983,
+      "step": 196680
+    },
+    {
+      "epoch": 9.769047382537002,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.848018277540479e-05,
+      "loss": 0.4894,
+      "step": 196690
+    },
+    {
+      "epoch": 9.769544054832622,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.844044899175524e-05,
+      "loss": 0.4924,
+      "step": 196700
+    },
+    {
+      "epoch": 9.77004072712824,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.840071520810569e-05,
+      "loss": 0.496,
+      "step": 196710
+    },
+    {
+      "epoch": 9.77053739942386,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.8360981424456145e-05,
+      "loss": 0.4968,
+      "step": 196720
+    },
+    {
+      "epoch": 9.77103407171948,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.8321247640806596e-05,
+      "loss": 0.4638,
+      "step": 196730
+    },
+    {
+      "epoch": 9.771530744015099,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 1.8281513857157048e-05,
+      "loss": 0.5139,
+      "step": 196740
+    },
+    {
+      "epoch": 9.772027416310719,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.82417800735075e-05,
+      "loss": 0.4681,
+      "step": 196750
+    },
+    {
+      "epoch": 9.772524088606337,
+      "grad_norm": 0.232421875,
+      "learning_rate": 1.8202046289857953e-05,
+      "loss": 0.48,
+      "step": 196760
+    },
+    {
+      "epoch": 9.773020760901957,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.8162312506208404e-05,
+      "loss": 0.4901,
+      "step": 196770
+    },
+    {
+      "epoch": 9.773517433197576,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.8122578722558855e-05,
+      "loss": 0.5155,
+      "step": 196780
+    },
+    {
+      "epoch": 9.774014105493196,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 1.8082844938909306e-05,
+      "loss": 0.4819,
+      "step": 196790
+    },
+    {
+      "epoch": 9.774510777788816,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.804311115525976e-05,
+      "loss": 0.4388,
+      "step": 196800
+    },
+    {
+      "epoch": 9.775007450084434,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.800337737161021e-05,
+      "loss": 0.5005,
+      "step": 196810
+    },
+    {
+      "epoch": 9.775504122380054,
+      "grad_norm": 0.2421875,
+      "learning_rate": 1.7963643587960666e-05,
+      "loss": 0.4617,
+      "step": 196820
+    },
+    {
+      "epoch": 9.776000794675673,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.7923909804311117e-05,
+      "loss": 0.468,
+      "step": 196830
+    },
+    {
+      "epoch": 9.776497466971293,
+      "grad_norm": 0.201171875,
+      "learning_rate": 1.7884176020661568e-05,
+      "loss": 0.5165,
+      "step": 196840
+    },
+    {
+      "epoch": 9.776994139266911,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.7844442237012022e-05,
+      "loss": 0.4809,
+      "step": 196850
+    },
+    {
+      "epoch": 9.777490811562531,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 1.7804708453362473e-05,
+      "loss": 0.4636,
+      "step": 196860
+    },
+    {
+      "epoch": 9.777987483858151,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.7764974669712924e-05,
+      "loss": 0.482,
+      "step": 196870
+    },
+    {
+      "epoch": 9.77848415615377,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.7725240886063375e-05,
+      "loss": 0.4925,
+      "step": 196880
+    },
+    {
+      "epoch": 9.77898082844939,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.768550710241383e-05,
+      "loss": 0.4965,
+      "step": 196890
+    },
+    {
+      "epoch": 9.779477500745008,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.764577331876428e-05,
+      "loss": 0.4706,
+      "step": 196900
+    },
+    {
+      "epoch": 9.779974173040628,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.7606039535114732e-05,
+      "loss": 0.492,
+      "step": 196910
+    },
+    {
+      "epoch": 9.780470845336247,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 1.7566305751465183e-05,
+      "loss": 0.4736,
+      "step": 196920
+    },
+    {
+      "epoch": 9.780967517631867,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.7526571967815637e-05,
+      "loss": 0.5131,
+      "step": 196930
+    },
+    {
+      "epoch": 9.781464189927487,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 1.748683818416609e-05,
+      "loss": 0.464,
+      "step": 196940
+    },
+    {
+      "epoch": 9.781960862223105,
+      "grad_norm": 0.19921875,
+      "learning_rate": 1.744710440051654e-05,
+      "loss": 0.4476,
+      "step": 196950
+    },
+    {
+      "epoch": 9.782457534518725,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.740737061686699e-05,
+      "loss": 0.496,
+      "step": 196960
+    },
+    {
+      "epoch": 9.782954206814344,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.7367636833217445e-05,
+      "loss": 0.472,
+      "step": 196970
+    },
+    {
+      "epoch": 9.783450879109964,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.7327903049567896e-05,
+      "loss": 0.4679,
+      "step": 196980
+    },
+    {
+      "epoch": 9.783947551405582,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.7288169265918347e-05,
+      "loss": 0.4861,
+      "step": 196990
+    },
+    {
+      "epoch": 9.784444223701202,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.7248435482268798e-05,
+      "loss": 0.4758,
+      "step": 197000
+    },
+    {
+      "epoch": 9.78494089599682,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.720870169861925e-05,
+      "loss": 0.4768,
+      "step": 197010
+    },
+    {
+      "epoch": 9.78543756829244,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.7168967914969703e-05,
+      "loss": 0.4772,
+      "step": 197020
+    },
+    {
+      "epoch": 9.78593424058806,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.7129234131320158e-05,
+      "loss": 0.4638,
+      "step": 197030
+    },
+    {
+      "epoch": 9.786430912883679,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.708950034767061e-05,
+      "loss": 0.4954,
+      "step": 197040
+    },
+    {
+      "epoch": 9.7869275851793,
+      "grad_norm": 0.22265625,
+      "learning_rate": 1.704976656402106e-05,
+      "loss": 0.4877,
+      "step": 197050
+    },
+    {
+      "epoch": 9.787424257474918,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.7010032780371514e-05,
+      "loss": 0.4782,
+      "step": 197060
+    },
+    {
+      "epoch": 9.787920929770538,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.6970298996721965e-05,
+      "loss": 0.4893,
+      "step": 197070
+    },
+    {
+      "epoch": 9.788417602066156,
+      "grad_norm": 0.234375,
+      "learning_rate": 1.6930565213072416e-05,
+      "loss": 0.4931,
+      "step": 197080
+    },
+    {
+      "epoch": 9.788914274361776,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.6890831429422867e-05,
+      "loss": 0.4815,
+      "step": 197090
+    },
+    {
+      "epoch": 9.789410946657396,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.685109764577332e-05,
+      "loss": 0.4825,
+      "step": 197100
+    },
+    {
+      "epoch": 9.789907618953015,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.6811363862123773e-05,
+      "loss": 0.4978,
+      "step": 197110
+    },
+    {
+      "epoch": 9.790404291248635,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.6771630078474224e-05,
+      "loss": 0.4736,
+      "step": 197120
+    },
+    {
+      "epoch": 9.790900963544253,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.6731896294824675e-05,
+      "loss": 0.4855,
+      "step": 197130
+    },
+    {
+      "epoch": 9.791397635839873,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.6692162511175126e-05,
+      "loss": 0.4818,
+      "step": 197140
+    },
+    {
+      "epoch": 9.791894308135491,
+      "grad_norm": 0.244140625,
+      "learning_rate": 1.665242872752558e-05,
+      "loss": 0.5244,
+      "step": 197150
+    },
+    {
+      "epoch": 9.792390980431112,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.661269494387603e-05,
+      "loss": 0.4878,
+      "step": 197160
+    },
+    {
+      "epoch": 9.792887652726732,
+      "grad_norm": 0.232421875,
+      "learning_rate": 1.6572961160226482e-05,
+      "loss": 0.5117,
+      "step": 197170
+    },
+    {
+      "epoch": 9.79338432502235,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.6533227376576933e-05,
+      "loss": 0.4566,
+      "step": 197180
+    },
+    {
+      "epoch": 9.79388099731797,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.6493493592927388e-05,
+      "loss": 0.468,
+      "step": 197190
+    },
+    {
+      "epoch": 9.794377669613588,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.645375980927784e-05,
+      "loss": 0.4971,
+      "step": 197200
+    },
+    {
+      "epoch": 9.794874341909209,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.641402602562829e-05,
+      "loss": 0.4705,
+      "step": 197210
+    },
+    {
+      "epoch": 9.795371014204827,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.637429224197874e-05,
+      "loss": 0.466,
+      "step": 197220
+    },
+    {
+      "epoch": 9.795867686500447,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.6334558458329195e-05,
+      "loss": 0.5081,
+      "step": 197230
+    },
+    {
+      "epoch": 9.796364358796067,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 1.629482467467965e-05,
+      "loss": 0.5119,
+      "step": 197240
+    },
+    {
+      "epoch": 9.796861031091685,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.62550908910301e-05,
+      "loss": 0.5001,
+      "step": 197250
+    },
+    {
+      "epoch": 9.797357703387306,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.6215357107380552e-05,
+      "loss": 0.5009,
+      "step": 197260
+    },
+    {
+      "epoch": 9.797854375682924,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.6175623323731003e-05,
+      "loss": 0.4913,
+      "step": 197270
+    },
+    {
+      "epoch": 9.798351047978544,
+      "grad_norm": 0.244140625,
+      "learning_rate": 1.6135889540081457e-05,
+      "loss": 0.5293,
+      "step": 197280
+    },
+    {
+      "epoch": 9.798847720274162,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.6096155756431908e-05,
+      "loss": 0.4953,
+      "step": 197290
+    },
+    {
+      "epoch": 9.799344392569783,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.605642197278236e-05,
+      "loss": 0.4845,
+      "step": 197300
+    },
+    {
+      "epoch": 9.799841064865403,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.601668818913281e-05,
+      "loss": 0.5252,
+      "step": 197310
+    },
+    {
+      "epoch": 9.800337737161021,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.5976954405483265e-05,
+      "loss": 0.4906,
+      "step": 197320
+    },
+    {
+      "epoch": 9.800834409456641,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.5937220621833716e-05,
+      "loss": 0.4903,
+      "step": 197330
+    },
+    {
+      "epoch": 9.80133108175226,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 1.5897486838184167e-05,
+      "loss": 0.4786,
+      "step": 197340
+    },
+    {
+      "epoch": 9.80182775404788,
+      "grad_norm": 0.240234375,
+      "learning_rate": 1.5857753054534618e-05,
+      "loss": 0.4979,
+      "step": 197350
+    },
+    {
+      "epoch": 9.802324426343498,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.5818019270885072e-05,
+      "loss": 0.4802,
+      "step": 197360
+    },
+    {
+      "epoch": 9.802821098639118,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.5778285487235523e-05,
+      "loss": 0.4789,
+      "step": 197370
+    },
+    {
+      "epoch": 9.803317770934736,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.5738551703585974e-05,
+      "loss": 0.4494,
+      "step": 197380
+    },
+    {
+      "epoch": 9.803814443230356,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.5698817919936425e-05,
+      "loss": 0.4801,
+      "step": 197390
+    },
+    {
+      "epoch": 9.804311115525977,
+      "grad_norm": 0.240234375,
+      "learning_rate": 1.5659084136286876e-05,
+      "loss": 0.5141,
+      "step": 197400
+    },
+    {
+      "epoch": 9.804807787821595,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.561935035263733e-05,
+      "loss": 0.5059,
+      "step": 197410
+    },
+    {
+      "epoch": 9.805304460117215,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.5579616568987782e-05,
+      "loss": 0.4943,
+      "step": 197420
+    },
+    {
+      "epoch": 9.805801132412833,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.5539882785338233e-05,
+      "loss": 0.4801,
+      "step": 197430
+    },
+    {
+      "epoch": 9.806297804708453,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.5500149001688684e-05,
+      "loss": 0.4785,
+      "step": 197440
+    },
+    {
+      "epoch": 9.806794477004072,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 1.5460415218039138e-05,
+      "loss": 0.49,
+      "step": 197450
+    },
+    {
+      "epoch": 9.807291149299692,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.5420681434389593e-05,
+      "loss": 0.4955,
+      "step": 197460
+    },
+    {
+      "epoch": 9.807787821595312,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.5380947650740044e-05,
+      "loss": 0.4877,
+      "step": 197470
+    },
+    {
+      "epoch": 9.80828449389093,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.5341213867090495e-05,
+      "loss": 0.4834,
+      "step": 197480
+    },
+    {
+      "epoch": 9.80878116618655,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.5301480083440946e-05,
+      "loss": 0.4689,
+      "step": 197490
+    },
+    {
+      "epoch": 9.809277838482169,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.52617462997914e-05,
+      "loss": 0.4608,
+      "step": 197500
+    },
+    {
+      "epoch": 9.809774510777789,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.5222012516141851e-05,
+      "loss": 0.528,
+      "step": 197510
+    },
+    {
+      "epoch": 9.810271183073407,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.5182278732492302e-05,
+      "loss": 0.5089,
+      "step": 197520
+    },
+    {
+      "epoch": 9.810767855369027,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.5142544948842755e-05,
+      "loss": 0.487,
+      "step": 197530
+    },
+    {
+      "epoch": 9.811264527664648,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.5102811165193206e-05,
+      "loss": 0.4944,
+      "step": 197540
+    },
+    {
+      "epoch": 9.811761199960266,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.5063077381543659e-05,
+      "loss": 0.51,
+      "step": 197550
+    },
+    {
+      "epoch": 9.812257872255886,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 1.502334359789411e-05,
+      "loss": 0.486,
+      "step": 197560
+    },
+    {
+      "epoch": 9.812754544551504,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.4983609814244562e-05,
+      "loss": 0.501,
+      "step": 197570
+    },
+    {
+      "epoch": 9.813251216847124,
+      "grad_norm": 0.2421875,
+      "learning_rate": 1.4943876030595013e-05,
+      "loss": 0.4839,
+      "step": 197580
+    },
+    {
+      "epoch": 9.813747889142743,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.4904142246945466e-05,
+      "loss": 0.4591,
+      "step": 197590
+    },
+    {
+      "epoch": 9.814244561438363,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.4864408463295917e-05,
+      "loss": 0.484,
+      "step": 197600
+    },
+    {
+      "epoch": 9.814741233733983,
+      "grad_norm": 0.22265625,
+      "learning_rate": 1.482467467964637e-05,
+      "loss": 0.5014,
+      "step": 197610
+    },
+    {
+      "epoch": 9.815237906029601,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.4784940895996821e-05,
+      "loss": 0.454,
+      "step": 197620
+    },
+    {
+      "epoch": 9.815734578325221,
+      "grad_norm": 0.228515625,
+      "learning_rate": 1.4745207112347274e-05,
+      "loss": 0.526,
+      "step": 197630
+    },
+    {
+      "epoch": 9.81623125062084,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.4705473328697725e-05,
+      "loss": 0.495,
+      "step": 197640
+    },
+    {
+      "epoch": 9.81672792291646,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.4665739545048177e-05,
+      "loss": 0.4979,
+      "step": 197650
+    },
+    {
+      "epoch": 9.817224595212078,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.4626005761398632e-05,
+      "loss": 0.5173,
+      "step": 197660
+    },
+    {
+      "epoch": 9.817721267507698,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.4586271977749083e-05,
+      "loss": 0.5001,
+      "step": 197670
+    },
+    {
+      "epoch": 9.818217939803318,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.4546538194099536e-05,
+      "loss": 0.4927,
+      "step": 197680
+    },
+    {
+      "epoch": 9.818714612098937,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.4506804410449987e-05,
+      "loss": 0.4713,
+      "step": 197690
+    },
+    {
+      "epoch": 9.819211284394557,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.446707062680044e-05,
+      "loss": 0.4837,
+      "step": 197700
+    },
+    {
+      "epoch": 9.819707956690175,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.442733684315089e-05,
+      "loss": 0.4783,
+      "step": 197710
+    },
+    {
+      "epoch": 9.820204628985795,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.4387603059501343e-05,
+      "loss": 0.5004,
+      "step": 197720
+    },
+    {
+      "epoch": 9.820701301281414,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.4347869275851794e-05,
+      "loss": 0.516,
+      "step": 197730
+    },
+    {
+      "epoch": 9.821197973577034,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.4308135492202247e-05,
+      "loss": 0.4959,
+      "step": 197740
+    },
+    {
+      "epoch": 9.821694645872654,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.4268401708552698e-05,
+      "loss": 0.5084,
+      "step": 197750
+    },
+    {
+      "epoch": 9.822191318168272,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.4228667924903149e-05,
+      "loss": 0.4627,
+      "step": 197760
+    },
+    {
+      "epoch": 9.822687990463892,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.4188934141253602e-05,
+      "loss": 0.4914,
+      "step": 197770
+    },
+    {
+      "epoch": 9.82318466275951,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.4149200357604053e-05,
+      "loss": 0.4794,
+      "step": 197780
+    },
+    {
+      "epoch": 9.82368133505513,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 1.4109466573954505e-05,
+      "loss": 0.4824,
+      "step": 197790
+    },
+    {
+      "epoch": 9.82417800735075,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.4069732790304956e-05,
+      "loss": 0.4791,
+      "step": 197800
+    },
+    {
+      "epoch": 9.82467467964637,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.4029999006655409e-05,
+      "loss": 0.4911,
+      "step": 197810
+    },
+    {
+      "epoch": 9.82517135194199,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.399026522300586e-05,
+      "loss": 0.4759,
+      "step": 197820
+    },
+    {
+      "epoch": 9.825668024237608,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.3950531439356313e-05,
+      "loss": 0.481,
+      "step": 197830
+    },
+    {
+      "epoch": 9.826164696533228,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.3910797655706764e-05,
+      "loss": 0.5043,
+      "step": 197840
+    },
+    {
+      "epoch": 9.826661368828846,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.3871063872057217e-05,
+      "loss": 0.4735,
+      "step": 197850
+    },
+    {
+      "epoch": 9.827158041124466,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.3831330088407668e-05,
+      "loss": 0.5065,
+      "step": 197860
+    },
+    {
+      "epoch": 9.827654713420085,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.3791596304758122e-05,
+      "loss": 0.5063,
+      "step": 197870
+    },
+    {
+      "epoch": 9.828151385715705,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.3751862521108575e-05,
+      "loss": 0.4593,
+      "step": 197880
+    },
+    {
+      "epoch": 9.828648058011325,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 1.3712128737459026e-05,
+      "loss": 0.4911,
+      "step": 197890
+    },
+    {
+      "epoch": 9.829144730306943,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.3672394953809478e-05,
+      "loss": 0.4695,
+      "step": 197900
+    },
+    {
+      "epoch": 9.829641402602563,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.363266117015993e-05,
+      "loss": 0.4862,
+      "step": 197910
+    },
+    {
+      "epoch": 9.830138074898182,
+      "grad_norm": 0.234375,
+      "learning_rate": 1.3592927386510382e-05,
+      "loss": 0.464,
+      "step": 197920
+    },
+    {
+      "epoch": 9.830634747193802,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.3553193602860833e-05,
+      "loss": 0.4768,
+      "step": 197930
+    },
+    {
+      "epoch": 9.83113141948942,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.3513459819211286e-05,
+      "loss": 0.4961,
+      "step": 197940
+    },
+    {
+      "epoch": 9.83162809178504,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.3473726035561737e-05,
+      "loss": 0.4947,
+      "step": 197950
+    },
+    {
+      "epoch": 9.83212476408066,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.343399225191219e-05,
+      "loss": 0.4792,
+      "step": 197960
+    },
+    {
+      "epoch": 9.832621436376279,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.339425846826264e-05,
+      "loss": 0.4931,
+      "step": 197970
+    },
+    {
+      "epoch": 9.833118108671899,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.3354524684613093e-05,
+      "loss": 0.497,
+      "step": 197980
+    },
+    {
+      "epoch": 9.833614780967517,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.3314790900963544e-05,
+      "loss": 0.4727,
+      "step": 197990
+    },
+    {
+      "epoch": 9.834111453263137,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.3275057117313997e-05,
+      "loss": 0.4828,
+      "step": 198000
+    },
+    {
+      "epoch": 9.834608125558756,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.3235323333664448e-05,
+      "loss": 0.4798,
+      "step": 198010
+    },
+    {
+      "epoch": 9.835104797854376,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.3195589550014901e-05,
+      "loss": 0.4992,
+      "step": 198020
+    },
+    {
+      "epoch": 9.835601470149996,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.3155855766365352e-05,
+      "loss": 0.4932,
+      "step": 198030
+    },
+    {
+      "epoch": 9.836098142445614,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.3116121982715805e-05,
+      "loss": 0.4665,
+      "step": 198040
+    },
+    {
+      "epoch": 9.836594814741234,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.3076388199066256e-05,
+      "loss": 0.4977,
+      "step": 198050
+    },
+    {
+      "epoch": 9.837091487036853,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.3036654415416707e-05,
+      "loss": 0.4823,
+      "step": 198060
+    },
+    {
+      "epoch": 9.837588159332473,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.299692063176716e-05,
+      "loss": 0.5245,
+      "step": 198070
+    },
+    {
+      "epoch": 9.838084831628091,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.2957186848117614e-05,
+      "loss": 0.5101,
+      "step": 198080
+    },
+    {
+      "epoch": 9.838581503923711,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.2917453064468067e-05,
+      "loss": 0.5216,
+      "step": 198090
+    },
+    {
+      "epoch": 9.839078176219331,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.2877719280818518e-05,
+      "loss": 0.4817,
+      "step": 198100
+    },
+    {
+      "epoch": 9.83957484851495,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.2837985497168969e-05,
+      "loss": 0.4895,
+      "step": 198110
+    },
+    {
+      "epoch": 9.84007152081057,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.2798251713519421e-05,
+      "loss": 0.5008,
+      "step": 198120
+    },
+    {
+      "epoch": 9.840568193106188,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.2758517929869872e-05,
+      "loss": 0.4854,
+      "step": 198130
+    },
+    {
+      "epoch": 9.841064865401808,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.2718784146220325e-05,
+      "loss": 0.4617,
+      "step": 198140
+    },
+    {
+      "epoch": 9.841561537697427,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.2679050362570776e-05,
+      "loss": 0.5255,
+      "step": 198150
+    },
+    {
+      "epoch": 9.842058209993047,
+      "grad_norm": 0.22265625,
+      "learning_rate": 1.2639316578921229e-05,
+      "loss": 0.4817,
+      "step": 198160
+    },
+    {
+      "epoch": 9.842554882288667,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.259958279527168e-05,
+      "loss": 0.4856,
+      "step": 198170
+    },
+    {
+      "epoch": 9.843051554584285,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.2559849011622133e-05,
+      "loss": 0.4613,
+      "step": 198180
+    },
+    {
+      "epoch": 9.843548226879905,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.2520115227972584e-05,
+      "loss": 0.5174,
+      "step": 198190
+    },
+    {
+      "epoch": 9.844044899175524,
+      "grad_norm": 0.228515625,
+      "learning_rate": 1.2480381444323036e-05,
+      "loss": 0.5079,
+      "step": 198200
+    },
+    {
+      "epoch": 9.844541571471144,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.2440647660673487e-05,
+      "loss": 0.4683,
+      "step": 198210
+    },
+    {
+      "epoch": 9.845038243766762,
+      "grad_norm": 0.2421875,
+      "learning_rate": 1.240091387702394e-05,
+      "loss": 0.5125,
+      "step": 198220
+    },
+    {
+      "epoch": 9.845534916062382,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.2361180093374391e-05,
+      "loss": 0.4804,
+      "step": 198230
+    },
+    {
+      "epoch": 9.846031588358002,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.2321446309724846e-05,
+      "loss": 0.477,
+      "step": 198240
+    },
+    {
+      "epoch": 9.84652826065362,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.2281712526075297e-05,
+      "loss": 0.4946,
+      "step": 198250
+    },
+    {
+      "epoch": 9.84702493294924,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.224197874242575e-05,
+      "loss": 0.4718,
+      "step": 198260
+    },
+    {
+      "epoch": 9.847521605244859,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.22022449587762e-05,
+      "loss": 0.4889,
+      "step": 198270
+    },
+    {
+      "epoch": 9.84801827754048,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.2162511175126651e-05,
+      "loss": 0.4428,
+      "step": 198280
+    },
+    {
+      "epoch": 9.848514949836098,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.2122777391477104e-05,
+      "loss": 0.4802,
+      "step": 198290
+    },
+    {
+      "epoch": 9.849011622131718,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.2083043607827555e-05,
+      "loss": 0.478,
+      "step": 198300
+    },
+    {
+      "epoch": 9.849508294427338,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 1.2043309824178008e-05,
+      "loss": 0.4929,
+      "step": 198310
+    },
+    {
+      "epoch": 9.850004966722956,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.2003576040528459e-05,
+      "loss": 0.4814,
+      "step": 198320
+    },
+    {
+      "epoch": 9.850501639018576,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.1963842256878912e-05,
+      "loss": 0.498,
+      "step": 198330
+    },
+    {
+      "epoch": 9.850998311314195,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.1924108473229364e-05,
+      "loss": 0.4992,
+      "step": 198340
+    },
+    {
+      "epoch": 9.851494983609815,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.1884374689579817e-05,
+      "loss": 0.5012,
+      "step": 198350
+    },
+    {
+      "epoch": 9.851991655905433,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.1844640905930268e-05,
+      "loss": 0.4566,
+      "step": 198360
+    },
+    {
+      "epoch": 9.852488328201053,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.180490712228072e-05,
+      "loss": 0.4975,
+      "step": 198370
+    },
+    {
+      "epoch": 9.852985000496671,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.1765173338631172e-05,
+      "loss": 0.4822,
+      "step": 198380
+    },
+    {
+      "epoch": 9.853481672792292,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.1725439554981624e-05,
+      "loss": 0.4713,
+      "step": 198390
+    },
+    {
+      "epoch": 9.853978345087912,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.1685705771332076e-05,
+      "loss": 0.5041,
+      "step": 198400
+    },
+    {
+      "epoch": 9.85447501738353,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.1645971987682528e-05,
+      "loss": 0.4846,
+      "step": 198410
+    },
+    {
+      "epoch": 9.85497168967915,
+      "grad_norm": 0.201171875,
+      "learning_rate": 1.160623820403298e-05,
+      "loss": 0.4568,
+      "step": 198420
+    },
+    {
+      "epoch": 9.855468361974768,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.156650442038343e-05,
+      "loss": 0.5135,
+      "step": 198430
+    },
+    {
+      "epoch": 9.855965034270389,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.1526770636733883e-05,
+      "loss": 0.5184,
+      "step": 198440
+    },
+    {
+      "epoch": 9.856461706566007,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.1487036853084336e-05,
+      "loss": 0.4928,
+      "step": 198450
+    },
+    {
+      "epoch": 9.856958378861627,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.1447303069434788e-05,
+      "loss": 0.4681,
+      "step": 198460
+    },
+    {
+      "epoch": 9.857455051157247,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.140756928578524e-05,
+      "loss": 0.469,
+      "step": 198470
+    },
+    {
+      "epoch": 9.857951723452866,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.1367835502135692e-05,
+      "loss": 0.4632,
+      "step": 198480
+    },
+    {
+      "epoch": 9.858448395748486,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.1328101718486143e-05,
+      "loss": 0.4931,
+      "step": 198490
+    },
+    {
+      "epoch": 9.858945068044104,
+      "grad_norm": 0.232421875,
+      "learning_rate": 1.1288367934836596e-05,
+      "loss": 0.4856,
+      "step": 198500
+    },
+    {
+      "epoch": 9.859441740339724,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.1248634151187047e-05,
+      "loss": 0.4895,
+      "step": 198510
+    },
+    {
+      "epoch": 9.859938412635342,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.12089003675375e-05,
+      "loss": 0.4701,
+      "step": 198520
+    },
+    {
+      "epoch": 9.860435084930963,
+      "grad_norm": 0.240234375,
+      "learning_rate": 1.116916658388795e-05,
+      "loss": 0.4943,
+      "step": 198530
+    },
+    {
+      "epoch": 9.860931757226583,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 1.1129432800238403e-05,
+      "loss": 0.482,
+      "step": 198540
+    },
+    {
+      "epoch": 9.861428429522201,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.1089699016588854e-05,
+      "loss": 0.4901,
+      "step": 198550
+    },
+    {
+      "epoch": 9.861925101817821,
+      "grad_norm": 0.244140625,
+      "learning_rate": 1.1049965232939307e-05,
+      "loss": 0.4929,
+      "step": 198560
+    },
+    {
+      "epoch": 9.86242177411344,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.101023144928976e-05,
+      "loss": 0.5104,
+      "step": 198570
+    },
+    {
+      "epoch": 9.86291844640906,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.0970497665640211e-05,
+      "loss": 0.4708,
+      "step": 198580
+    },
+    {
+      "epoch": 9.863415118704678,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.0930763881990664e-05,
+      "loss": 0.4916,
+      "step": 198590
+    },
+    {
+      "epoch": 9.863911791000298,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.0891030098341115e-05,
+      "loss": 0.4666,
+      "step": 198600
+    },
+    {
+      "epoch": 9.864408463295918,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.0851296314691567e-05,
+      "loss": 0.5095,
+      "step": 198610
+    },
+    {
+      "epoch": 9.864905135591536,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.0811562531042018e-05,
+      "loss": 0.47,
+      "step": 198620
+    },
+    {
+      "epoch": 9.865401807887157,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.0771828747392471e-05,
+      "loss": 0.4777,
+      "step": 198630
+    },
+    {
+      "epoch": 9.865898480182775,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.0732094963742922e-05,
+      "loss": 0.4784,
+      "step": 198640
+    },
+    {
+      "epoch": 9.866395152478395,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.0692361180093375e-05,
+      "loss": 0.4772,
+      "step": 198650
+    },
+    {
+      "epoch": 9.866891824774013,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.0652627396443828e-05,
+      "loss": 0.4804,
+      "step": 198660
+    },
+    {
+      "epoch": 9.867388497069634,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.0612893612794279e-05,
+      "loss": 0.4588,
+      "step": 198670
+    },
+    {
+      "epoch": 9.867885169365254,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.0573159829144731e-05,
+      "loss": 0.4805,
+      "step": 198680
+    },
+    {
+      "epoch": 9.868381841660872,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.0533426045495182e-05,
+      "loss": 0.4908,
+      "step": 198690
+    },
+    {
+      "epoch": 9.868878513956492,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.0493692261845635e-05,
+      "loss": 0.4779,
+      "step": 198700
+    },
+    {
+      "epoch": 9.86937518625211,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 1.0453958478196086e-05,
+      "loss": 0.4911,
+      "step": 198710
+    },
+    {
+      "epoch": 9.86987185854773,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.0414224694546539e-05,
+      "loss": 0.49,
+      "step": 198720
+    },
+    {
+      "epoch": 9.870368530843349,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.037449091089699e-05,
+      "loss": 0.467,
+      "step": 198730
+    },
+    {
+      "epoch": 9.870865203138969,
+      "grad_norm": 0.244140625,
+      "learning_rate": 1.0334757127247443e-05,
+      "loss": 0.5067,
+      "step": 198740
+    },
+    {
+      "epoch": 9.871361875434587,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.0295023343597894e-05,
+      "loss": 0.4727,
+      "step": 198750
+    },
+    {
+      "epoch": 9.871858547730207,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.0255289559948346e-05,
+      "loss": 0.468,
+      "step": 198760
+    },
+    {
+      "epoch": 9.872355220025828,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.0215555776298799e-05,
+      "loss": 0.488,
+      "step": 198770
+    },
+    {
+      "epoch": 9.872851892321446,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.017582199264925e-05,
+      "loss": 0.4788,
+      "step": 198780
+    },
+    {
+      "epoch": 9.873348564617066,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.0136088208999703e-05,
+      "loss": 0.5071,
+      "step": 198790
+    },
+    {
+      "epoch": 9.873845236912684,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.0096354425350154e-05,
+      "loss": 0.4792,
+      "step": 198800
+    },
+    {
+      "epoch": 9.874341909208304,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.0056620641700607e-05,
+      "loss": 0.5005,
+      "step": 198810
+    },
+    {
+      "epoch": 9.874838581503923,
+      "grad_norm": 0.24609375,
+      "learning_rate": 1.0016886858051058e-05,
+      "loss": 0.5227,
+      "step": 198820
+    },
+    {
+      "epoch": 9.875335253799543,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.97715307440151e-06,
+      "loss": 0.4724,
+      "step": 198830
+    },
+    {
+      "epoch": 9.875831926095163,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.937419290751961e-06,
+      "loss": 0.4943,
+      "step": 198840
+    },
+    {
+      "epoch": 9.876328598390781,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.897685507102414e-06,
+      "loss": 0.5062,
+      "step": 198850
+    },
+    {
+      "epoch": 9.876825270686401,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.857951723452865e-06,
+      "loss": 0.5059,
+      "step": 198860
+    },
+    {
+      "epoch": 9.87732194298202,
+      "grad_norm": 0.23828125,
+      "learning_rate": 9.81821793980332e-06,
+      "loss": 0.5141,
+      "step": 198870
+    },
+    {
+      "epoch": 9.87781861527764,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.77848415615377e-06,
+      "loss": 0.5031,
+      "step": 198880
+    },
+    {
+      "epoch": 9.878315287573258,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 9.738750372504223e-06,
+      "loss": 0.4937,
+      "step": 198890
+    },
+    {
+      "epoch": 9.878811959868878,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.699016588854674e-06,
+      "loss": 0.4737,
+      "step": 198900
+    },
+    {
+      "epoch": 9.879308632164499,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.659282805205127e-06,
+      "loss": 0.475,
+      "step": 198910
+    },
+    {
+      "epoch": 9.879805304460117,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.619549021555578e-06,
+      "loss": 0.5084,
+      "step": 198920
+    },
+    {
+      "epoch": 9.880301976755737,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.57981523790603e-06,
+      "loss": 0.5012,
+      "step": 198930
+    },
+    {
+      "epoch": 9.880798649051355,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.540081454256482e-06,
+      "loss": 0.483,
+      "step": 198940
+    },
+    {
+      "epoch": 9.881295321346975,
+      "grad_norm": 0.236328125,
+      "learning_rate": 9.500347670606933e-06,
+      "loss": 0.5229,
+      "step": 198950
+    },
+    {
+      "epoch": 9.881791993642594,
+      "grad_norm": 0.2421875,
+      "learning_rate": 9.460613886957386e-06,
+      "loss": 0.524,
+      "step": 198960
+    },
+    {
+      "epoch": 9.882288665938214,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.420880103307837e-06,
+      "loss": 0.4776,
+      "step": 198970
+    },
+    {
+      "epoch": 9.882785338233834,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.381146319658291e-06,
+      "loss": 0.4885,
+      "step": 198980
+    },
+    {
+      "epoch": 9.883282010529452,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.341412536008742e-06,
+      "loss": 0.4874,
+      "step": 198990
+    },
+    {
+      "epoch": 9.883778682825072,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.301678752359195e-06,
+      "loss": 0.5069,
+      "step": 199000
+    },
+    {
+      "epoch": 9.88427535512069,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.261944968709646e-06,
+      "loss": 0.4734,
+      "step": 199010
+    },
+    {
+      "epoch": 9.884772027416311,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.222211185060098e-06,
+      "loss": 0.4917,
+      "step": 199020
+    },
+    {
+      "epoch": 9.88526869971193,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.18247740141055e-06,
+      "loss": 0.4996,
+      "step": 199030
+    },
+    {
+      "epoch": 9.88576537200755,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.142743617761002e-06,
+      "loss": 0.5089,
+      "step": 199040
+    },
+    {
+      "epoch": 9.88626204430317,
+      "grad_norm": 0.212890625,
+      "learning_rate": 9.103009834111453e-06,
+      "loss": 0.5113,
+      "step": 199050
+    },
+    {
+      "epoch": 9.886758716598788,
+      "grad_norm": 0.205078125,
+      "learning_rate": 9.063276050461906e-06,
+      "loss": 0.483,
+      "step": 199060
+    },
+    {
+      "epoch": 9.887255388894408,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.023542266812357e-06,
+      "loss": 0.5336,
+      "step": 199070
+    },
+    {
+      "epoch": 9.887752061190026,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.98380848316281e-06,
+      "loss": 0.4434,
+      "step": 199080
+    },
+    {
+      "epoch": 9.888248733485646,
+      "grad_norm": 0.19921875,
+      "learning_rate": 8.944074699513262e-06,
+      "loss": 0.494,
+      "step": 199090
+    },
+    {
+      "epoch": 9.888745405781265,
+      "grad_norm": 0.22265625,
+      "learning_rate": 8.904340915863713e-06,
+      "loss": 0.5094,
+      "step": 199100
+    },
+    {
+      "epoch": 9.889242078076885,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.864607132214166e-06,
+      "loss": 0.4903,
+      "step": 199110
+    },
+    {
+      "epoch": 9.889738750372505,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.824873348564617e-06,
+      "loss": 0.4492,
+      "step": 199120
+    },
+    {
+      "epoch": 9.890235422668123,
+      "grad_norm": 0.228515625,
+      "learning_rate": 8.78513956491507e-06,
+      "loss": 0.5074,
+      "step": 199130
+    },
+    {
+      "epoch": 9.890732094963743,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 8.745405781265521e-06,
+      "loss": 0.4771,
+      "step": 199140
+    },
+    {
+      "epoch": 9.891228767259362,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 8.705671997615974e-06,
+      "loss": 0.5127,
+      "step": 199150
+    },
+    {
+      "epoch": 9.891725439554982,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 8.665938213966425e-06,
+      "loss": 0.4915,
+      "step": 199160
+    },
+    {
+      "epoch": 9.8922221118506,
+      "grad_norm": 0.216796875,
+      "learning_rate": 8.626204430316877e-06,
+      "loss": 0.4868,
+      "step": 199170
+    },
+    {
+      "epoch": 9.89271878414622,
+      "grad_norm": 0.279296875,
+      "learning_rate": 8.586470646667328e-06,
+      "loss": 0.5011,
+      "step": 199180
+    },
+    {
+      "epoch": 9.89321545644184,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8.546736863017781e-06,
+      "loss": 0.4708,
+      "step": 199190
+    },
+    {
+      "epoch": 9.893712128737459,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 8.507003079368234e-06,
+      "loss": 0.4672,
+      "step": 199200
+    },
+    {
+      "epoch": 9.894208801033079,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.467269295718685e-06,
+      "loss": 0.496,
+      "step": 199210
+    },
+    {
+      "epoch": 9.894705473328697,
+      "grad_norm": 0.25390625,
+      "learning_rate": 8.427535512069138e-06,
+      "loss": 0.4965,
+      "step": 199220
+    },
+    {
+      "epoch": 9.895202145624317,
+      "grad_norm": 0.22265625,
+      "learning_rate": 8.387801728419589e-06,
+      "loss": 0.4911,
+      "step": 199230
+    },
+    {
+      "epoch": 9.895698817919936,
+      "grad_norm": 0.201171875,
+      "learning_rate": 8.348067944770041e-06,
+      "loss": 0.4907,
+      "step": 199240
+    },
+    {
+      "epoch": 9.896195490215556,
+      "grad_norm": 0.236328125,
+      "learning_rate": 8.308334161120492e-06,
+      "loss": 0.5264,
+      "step": 199250
+    },
+    {
+      "epoch": 9.896692162511176,
+      "grad_norm": 0.220703125,
+      "learning_rate": 8.268600377470945e-06,
+      "loss": 0.4876,
+      "step": 199260
+    },
+    {
+      "epoch": 9.897188834806794,
+      "grad_norm": 0.197265625,
+      "learning_rate": 8.228866593821396e-06,
+      "loss": 0.5253,
+      "step": 199270
+    },
+    {
+      "epoch": 9.897685507102414,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 8.189132810171849e-06,
+      "loss": 0.4679,
+      "step": 199280
+    },
+    {
+      "epoch": 9.898182179398033,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 8.149399026522302e-06,
+      "loss": 0.512,
+      "step": 199290
+    },
+    {
+      "epoch": 9.898678851693653,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 8.109665242872753e-06,
+      "loss": 0.4922,
+      "step": 199300
+    },
+    {
+      "epoch": 9.899175523989271,
+      "grad_norm": 0.216796875,
+      "learning_rate": 8.069931459223205e-06,
+      "loss": 0.4891,
+      "step": 199310
+    },
+    {
+      "epoch": 9.899672196284891,
+      "grad_norm": 0.291015625,
+      "learning_rate": 8.030197675573656e-06,
+      "loss": 0.5064,
+      "step": 199320
+    },
+    {
+      "epoch": 9.900168868580511,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.990463891924109e-06,
+      "loss": 0.4944,
+      "step": 199330
+    },
+    {
+      "epoch": 9.90066554087613,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.95073010827456e-06,
+      "loss": 0.4719,
+      "step": 199340
+    },
+    {
+      "epoch": 9.90116221317175,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 7.910996324625013e-06,
+      "loss": 0.4804,
+      "step": 199350
+    },
+    {
+      "epoch": 9.901658885467368,
+      "grad_norm": 0.197265625,
+      "learning_rate": 7.871262540975464e-06,
+      "loss": 0.5012,
+      "step": 199360
+    },
+    {
+      "epoch": 9.902155557762988,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 7.831528757325917e-06,
+      "loss": 0.5368,
+      "step": 199370
+    },
+    {
+      "epoch": 9.902652230058607,
+      "grad_norm": 0.212890625,
+      "learning_rate": 7.791794973676368e-06,
+      "loss": 0.5028,
+      "step": 199380
+    },
+    {
+      "epoch": 9.903148902354227,
+      "grad_norm": 0.19921875,
+      "learning_rate": 7.75206119002682e-06,
+      "loss": 0.4807,
+      "step": 199390
+    },
+    {
+      "epoch": 9.903645574649847,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 7.712327406377273e-06,
+      "loss": 0.4862,
+      "step": 199400
+    },
+    {
+      "epoch": 9.904142246945465,
+      "grad_norm": 0.21484375,
+      "learning_rate": 7.672593622727726e-06,
+      "loss": 0.5135,
+      "step": 199410
+    },
+    {
+      "epoch": 9.904638919241085,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 7.632859839078177e-06,
+      "loss": 0.4843,
+      "step": 199420
+    },
+    {
+      "epoch": 9.905135591536704,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.593126055428629e-06,
+      "loss": 0.5021,
+      "step": 199430
+    },
+    {
+      "epoch": 9.905632263832324,
+      "grad_norm": 0.216796875,
+      "learning_rate": 7.5533922717790806e-06,
+      "loss": 0.4686,
+      "step": 199440
+    },
+    {
+      "epoch": 9.906128936127942,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 7.5136584881295324e-06,
+      "loss": 0.4893,
+      "step": 199450
+    },
+    {
+      "epoch": 9.906625608423562,
+      "grad_norm": 0.216796875,
+      "learning_rate": 7.473924704479984e-06,
+      "loss": 0.4977,
+      "step": 199460
+    },
+    {
+      "epoch": 9.907122280719182,
+      "grad_norm": 0.23046875,
+      "learning_rate": 7.434190920830436e-06,
+      "loss": 0.517,
+      "step": 199470
+    },
+    {
+      "epoch": 9.9076189530148,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 7.394457137180888e-06,
+      "loss": 0.4733,
+      "step": 199480
+    },
+    {
+      "epoch": 9.90811562531042,
+      "grad_norm": 0.21484375,
+      "learning_rate": 7.35472335353134e-06,
+      "loss": 0.4651,
+      "step": 199490
+    },
+    {
+      "epoch": 9.90861229760604,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 7.314989569881793e-06,
+      "loss": 0.4776,
+      "step": 199500
+    },
+    {
+      "epoch": 9.90910896990166,
+      "grad_norm": 0.2265625,
+      "learning_rate": 7.2752557862322445e-06,
+      "loss": 0.4696,
+      "step": 199510
+    },
+    {
+      "epoch": 9.909605642197278,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 7.235522002582696e-06,
+      "loss": 0.4766,
+      "step": 199520
+    },
+    {
+      "epoch": 9.910102314492898,
+      "grad_norm": 0.20703125,
+      "learning_rate": 7.195788218933148e-06,
+      "loss": 0.4903,
+      "step": 199530
+    },
+    {
+      "epoch": 9.910598986788518,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 7.1560544352836e-06,
+      "loss": 0.4591,
+      "step": 199540
+    },
+    {
+      "epoch": 9.911095659084136,
+      "grad_norm": 0.22265625,
+      "learning_rate": 7.116320651634052e-06,
+      "loss": 0.4726,
+      "step": 199550
+    },
+    {
+      "epoch": 9.911592331379756,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.076586867984504e-06,
+      "loss": 0.5133,
+      "step": 199560
+    },
+    {
+      "epoch": 9.912089003675375,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 7.036853084334956e-06,
+      "loss": 0.4847,
+      "step": 199570
+    },
+    {
+      "epoch": 9.912585675970995,
+      "grad_norm": 0.216796875,
+      "learning_rate": 6.997119300685408e-06,
+      "loss": 0.5002,
+      "step": 199580
+    },
+    {
+      "epoch": 9.913082348266613,
+      "grad_norm": 0.220703125,
+      "learning_rate": 6.9573855170358595e-06,
+      "loss": 0.4941,
+      "step": 199590
+    },
+    {
+      "epoch": 9.913579020562233,
+      "grad_norm": 0.220703125,
+      "learning_rate": 6.917651733386311e-06,
+      "loss": 0.5123,
+      "step": 199600
+    },
+    {
+      "epoch": 9.914075692857853,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 6.877917949736765e-06,
+      "loss": 0.5071,
+      "step": 199610
+    },
+    {
+      "epoch": 9.914572365153472,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 6.838184166087217e-06,
+      "loss": 0.4905,
+      "step": 199620
+    },
+    {
+      "epoch": 9.915069037449092,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 6.798450382437669e-06,
+      "loss": 0.4938,
+      "step": 199630
+    },
+    {
+      "epoch": 9.91556570974471,
+      "grad_norm": 0.232421875,
+      "learning_rate": 6.75871659878812e-06,
+      "loss": 0.4943,
+      "step": 199640
+    },
+    {
+      "epoch": 9.91606238204033,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 6.718982815138572e-06,
+      "loss": 0.4932,
+      "step": 199650
+    },
+    {
+      "epoch": 9.916559054335949,
+      "grad_norm": 0.29296875,
+      "learning_rate": 6.6792490314890235e-06,
+      "loss": 0.4677,
+      "step": 199660
+    },
+    {
+      "epoch": 9.917055726631569,
+      "grad_norm": 0.21875,
+      "learning_rate": 6.639515247839475e-06,
+      "loss": 0.4805,
+      "step": 199670
+    },
+    {
+      "epoch": 9.917552398927189,
+      "grad_norm": 0.25390625,
+      "learning_rate": 6.599781464189927e-06,
+      "loss": 0.5073,
+      "step": 199680
+    },
+    {
+      "epoch": 9.918049071222807,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 6.560047680540379e-06,
+      "loss": 0.4645,
+      "step": 199690
+    },
+    {
+      "epoch": 9.918545743518427,
+      "grad_norm": 0.203125,
+      "learning_rate": 6.520313896890831e-06,
+      "loss": 0.458,
+      "step": 199700
+    },
+    {
+      "epoch": 9.919042415814046,
+      "grad_norm": 0.20703125,
+      "learning_rate": 6.4805801132412845e-06,
+      "loss": 0.4685,
+      "step": 199710
+    },
+    {
+      "epoch": 9.919539088109666,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 6.440846329591736e-06,
+      "loss": 0.4949,
+      "step": 199720
+    },
+    {
+      "epoch": 9.920035760405284,
+      "grad_norm": 0.203125,
+      "learning_rate": 6.401112545942188e-06,
+      "loss": 0.4867,
+      "step": 199730
+    },
+    {
+      "epoch": 9.920532432700904,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 6.36137876229264e-06,
+      "loss": 0.4741,
+      "step": 199740
+    },
+    {
+      "epoch": 9.921029104996522,
+      "grad_norm": 0.271484375,
+      "learning_rate": 6.321644978643092e-06,
+      "loss": 0.4931,
+      "step": 199750
+    },
+    {
+      "epoch": 9.921525777292143,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 6.281911194993544e-06,
+      "loss": 0.4854,
+      "step": 199760
+    },
+    {
+      "epoch": 9.922022449587763,
+      "grad_norm": 0.23046875,
+      "learning_rate": 6.242177411343996e-06,
+      "loss": 0.4866,
+      "step": 199770
+    },
+    {
+      "epoch": 9.922519121883381,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 6.202443627694448e-06,
+      "loss": 0.4936,
+      "step": 199780
+    },
+    {
+      "epoch": 9.923015794179001,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 6.1627098440448995e-06,
+      "loss": 0.4695,
+      "step": 199790
+    },
+    {
+      "epoch": 9.92351246647462,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.122976060395351e-06,
+      "loss": 0.5042,
+      "step": 199800
+    },
+    {
+      "epoch": 9.92400913877024,
+      "grad_norm": 0.208984375,
+      "learning_rate": 6.083242276745803e-06,
+      "loss": 0.4995,
+      "step": 199810
+    },
+    {
+      "epoch": 9.924505811065858,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 6.043508493096255e-06,
+      "loss": 0.4638,
+      "step": 199820
+    },
+    {
+      "epoch": 9.925002483361478,
+      "grad_norm": 0.224609375,
+      "learning_rate": 6.003774709446707e-06,
+      "loss": 0.4836,
+      "step": 199830
+    },
+    {
+      "epoch": 9.925499155657098,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.964040925797159e-06,
+      "loss": 0.4783,
+      "step": 199840
+    },
+    {
+      "epoch": 9.925995827952717,
+      "grad_norm": 0.185546875,
+      "learning_rate": 5.924307142147612e-06,
+      "loss": 0.4765,
+      "step": 199850
+    },
+    {
+      "epoch": 9.926492500248337,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.8845733584980635e-06,
+      "loss": 0.5236,
+      "step": 199860
+    },
+    {
+      "epoch": 9.926989172543955,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 5.844839574848515e-06,
+      "loss": 0.5028,
+      "step": 199870
+    },
+    {
+      "epoch": 9.927485844839575,
+      "grad_norm": 0.208984375,
+      "learning_rate": 5.805105791198967e-06,
+      "loss": 0.4942,
+      "step": 199880
+    },
+    {
+      "epoch": 9.927982517135193,
+      "grad_norm": 0.21875,
+      "learning_rate": 5.765372007549419e-06,
+      "loss": 0.5034,
+      "step": 199890
+    },
+    {
+      "epoch": 9.928479189430814,
+      "grad_norm": 0.205078125,
+      "learning_rate": 5.725638223899871e-06,
+      "loss": 0.4904,
+      "step": 199900
+    },
+    {
+      "epoch": 9.928975861726434,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.685904440250323e-06,
+      "loss": 0.5059,
+      "step": 199910
+    },
+    {
+      "epoch": 9.929472534022052,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 5.646170656600775e-06,
+      "loss": 0.4754,
+      "step": 199920
+    },
+    {
+      "epoch": 9.929969206317672,
+      "grad_norm": 0.212890625,
+      "learning_rate": 5.606436872951227e-06,
+      "loss": 0.5001,
+      "step": 199930
+    },
+    {
+      "epoch": 9.93046587861329,
+      "grad_norm": 0.205078125,
+      "learning_rate": 5.5667030893016785e-06,
+      "loss": 0.4922,
+      "step": 199940
+    },
+    {
+      "epoch": 9.93096255090891,
+      "grad_norm": 0.21875,
+      "learning_rate": 5.526969305652131e-06,
+      "loss": 0.4963,
+      "step": 199950
+    },
+    {
+      "epoch": 9.931459223204529,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 5.487235522002583e-06,
+      "loss": 0.5201,
+      "step": 199960
+    },
+    {
+      "epoch": 9.931955895500149,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 5.447501738353035e-06,
+      "loss": 0.4807,
+      "step": 199970
+    },
+    {
+      "epoch": 9.93245256779577,
+      "grad_norm": 0.208984375,
+      "learning_rate": 5.407767954703487e-06,
+      "loss": 0.4466,
+      "step": 199980
+    },
+    {
+      "epoch": 9.932949240091387,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 5.368034171053939e-06,
+      "loss": 0.4951,
+      "step": 199990
+    },
+    {
+      "epoch": 9.933445912387008,
+      "grad_norm": 0.201171875,
+      "learning_rate": 5.3283003874043914e-06,
+      "loss": 0.4759,
+      "step": 200000
+    },
+    {
+      "epoch": 9.933942584682626,
+      "grad_norm": 0.20703125,
+      "learning_rate": 5.288566603754843e-06,
+      "loss": 0.4694,
+      "step": 200010
+    },
+    {
+      "epoch": 9.934439256978246,
+      "grad_norm": 0.189453125,
+      "learning_rate": 5.248832820105295e-06,
+      "loss": 0.4585,
+      "step": 200020
+    },
+    {
+      "epoch": 9.934935929273864,
+      "grad_norm": 0.216796875,
+      "learning_rate": 5.209099036455747e-06,
+      "loss": 0.4969,
+      "step": 200030
+    },
+    {
+      "epoch": 9.935432601569484,
+      "grad_norm": 0.201171875,
+      "learning_rate": 5.169365252806199e-06,
+      "loss": 0.4757,
+      "step": 200040
+    },
+    {
+      "epoch": 9.935929273865105,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 5.12963146915665e-06,
+      "loss": 0.4838,
+      "step": 200050
+    },
+    {
+      "epoch": 9.936425946160723,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 5.089897685507103e-06,
+      "loss": 0.4847,
+      "step": 200060
+    },
+    {
+      "epoch": 9.936922618456343,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.0501639018575545e-06,
+      "loss": 0.4847,
+      "step": 200070
+    },
+    {
+      "epoch": 9.937419290751961,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 5.010430118208006e-06,
+      "loss": 0.4708,
+      "step": 200080
+    },
+    {
+      "epoch": 9.937915963047582,
+      "grad_norm": 0.18359375,
+      "learning_rate": 4.970696334558458e-06,
+      "loss": 0.4802,
+      "step": 200090
+    },
+    {
+      "epoch": 9.9384126353432,
+      "grad_norm": 0.197265625,
+      "learning_rate": 4.93096255090891e-06,
+      "loss": 0.4859,
+      "step": 200100
+    },
+    {
+      "epoch": 9.93890930763882,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.891228767259363e-06,
+      "loss": 0.4903,
+      "step": 200110
+    },
+    {
+      "epoch": 9.93940597993444,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 4.851494983609815e-06,
+      "loss": 0.5034,
+      "step": 200120
+    },
+    {
+      "epoch": 9.939902652230058,
+      "grad_norm": 0.236328125,
+      "learning_rate": 4.811761199960267e-06,
+      "loss": 0.4839,
+      "step": 200130
+    },
+    {
+      "epoch": 9.940399324525679,
+      "grad_norm": 0.201171875,
+      "learning_rate": 4.7720274163107185e-06,
+      "loss": 0.4631,
+      "step": 200140
+    },
+    {
+      "epoch": 9.940895996821297,
+      "grad_norm": 0.205078125,
+      "learning_rate": 4.73229363266117e-06,
+      "loss": 0.4924,
+      "step": 200150
+    },
+    {
+      "epoch": 9.941392669116917,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 4.692559849011622e-06,
+      "loss": 0.4681,
+      "step": 200160
+    },
+    {
+      "epoch": 9.941889341412535,
+      "grad_norm": 0.201171875,
+      "learning_rate": 4.652826065362074e-06,
+      "loss": 0.4945,
+      "step": 200170
+    },
+    {
+      "epoch": 9.942386013708155,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 4.613092281712526e-06,
+      "loss": 0.4998,
+      "step": 200180
+    },
+    {
+      "epoch": 9.942882686003774,
+      "grad_norm": 0.236328125,
+      "learning_rate": 4.573358498062978e-06,
+      "loss": 0.5073,
+      "step": 200190
+    },
+    {
+      "epoch": 9.943379358299394,
+      "grad_norm": 0.20703125,
+      "learning_rate": 4.53362471441343e-06,
+      "loss": 0.5035,
+      "step": 200200
+    },
+    {
+      "epoch": 9.943876030595014,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 4.4938909307638825e-06,
+      "loss": 0.5029,
+      "step": 200210
+    },
+    {
+      "epoch": 9.944372702890632,
+      "grad_norm": 0.21484375,
+      "learning_rate": 4.454157147114334e-06,
+      "loss": 0.511,
+      "step": 200220
+    },
+    {
+      "epoch": 9.944869375186252,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.414423363464786e-06,
+      "loss": 0.5102,
+      "step": 200230
+    },
+    {
+      "epoch": 9.94536604748187,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 4.374689579815238e-06,
+      "loss": 0.5198,
+      "step": 200240
+    },
+    {
+      "epoch": 9.945862719777491,
+      "grad_norm": 0.21484375,
+      "learning_rate": 4.33495579616569e-06,
+      "loss": 0.482,
+      "step": 200250
+    },
+    {
+      "epoch": 9.94635939207311,
+      "grad_norm": 0.2109375,
+      "learning_rate": 4.295222012516142e-06,
+      "loss": 0.5265,
+      "step": 200260
+    },
+    {
+      "epoch": 9.94685606436873,
+      "grad_norm": 0.21484375,
+      "learning_rate": 4.2554882288665946e-06,
+      "loss": 0.5336,
+      "step": 200270
+    },
+    {
+      "epoch": 9.94735273666435,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.2157544452170464e-06,
+      "loss": 0.5028,
+      "step": 200280
+    },
+    {
+      "epoch": 9.947849408959968,
+      "grad_norm": 0.228515625,
+      "learning_rate": 4.176020661567498e-06,
+      "loss": 0.4794,
+      "step": 200290
+    },
+    {
+      "epoch": 9.948346081255588,
+      "grad_norm": 0.234375,
+      "learning_rate": 4.13628687791795e-06,
+      "loss": 0.5099,
+      "step": 200300
+    },
+    {
+      "epoch": 9.948842753551206,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 4.096553094268401e-06,
+      "loss": 0.4611,
+      "step": 200310
+    },
+    {
+      "epoch": 9.949339425846826,
+      "grad_norm": 0.197265625,
+      "learning_rate": 4.056819310618854e-06,
+      "loss": 0.466,
+      "step": 200320
+    },
+    {
+      "epoch": 9.949836098142445,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 4.017085526969306e-06,
+      "loss": 0.482,
+      "step": 200330
+    },
+    {
+      "epoch": 9.950332770438065,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.977351743319758e-06,
+      "loss": 0.4685,
+      "step": 200340
+    },
+    {
+      "epoch": 9.950829442733685,
+      "grad_norm": 0.2109375,
+      "learning_rate": 3.9376179596702096e-06,
+      "loss": 0.4698,
+      "step": 200350
+    },
+    {
+      "epoch": 9.951326115029303,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.8978841760206614e-06,
+      "loss": 0.4589,
+      "step": 200360
+    },
+    {
+      "epoch": 9.951822787324923,
+      "grad_norm": 0.205078125,
+      "learning_rate": 3.858150392371114e-06,
+      "loss": 0.4901,
+      "step": 200370
+    },
+    {
+      "epoch": 9.952319459620542,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.818416608721566e-06,
+      "loss": 0.4826,
+      "step": 200380
+    },
+    {
+      "epoch": 9.952816131916162,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 3.778682825072018e-06,
+      "loss": 0.4923,
+      "step": 200390
+    },
+    {
+      "epoch": 9.95331280421178,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.7389490414224694e-06,
+      "loss": 0.4776,
+      "step": 200400
+    },
+    {
+      "epoch": 9.9538094765074,
+      "grad_norm": 0.236328125,
+      "learning_rate": 3.6992152577729212e-06,
+      "loss": 0.4819,
+      "step": 200410
+    },
+    {
+      "epoch": 9.95430614880302,
+      "grad_norm": 0.201171875,
+      "learning_rate": 3.659481474123374e-06,
+      "loss": 0.4963,
+      "step": 200420
+    },
+    {
+      "epoch": 9.954802821098639,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.619747690473826e-06,
+      "loss": 0.4885,
+      "step": 200430
+    },
+    {
+      "epoch": 9.955299493394259,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 3.5800139068242777e-06,
+      "loss": 0.466,
+      "step": 200440
+    },
+    {
+      "epoch": 9.955796165689877,
+      "grad_norm": 0.201171875,
+      "learning_rate": 3.5402801231747296e-06,
+      "loss": 0.4784,
+      "step": 200450
+    },
+    {
+      "epoch": 9.956292837985497,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 3.5005463395251814e-06,
+      "loss": 0.5032,
+      "step": 200460
+    },
+    {
+      "epoch": 9.956789510281116,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.4608125558756333e-06,
+      "loss": 0.509,
+      "step": 200470
+    },
+    {
+      "epoch": 9.957286182576736,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.4210787722260856e-06,
+      "loss": 0.4765,
+      "step": 200480
+    },
+    {
+      "epoch": 9.957782854872356,
+      "grad_norm": 0.21484375,
+      "learning_rate": 3.3813449885765375e-06,
+      "loss": 0.492,
+      "step": 200490
+    },
+    {
+      "epoch": 9.958279527167974,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 3.3416112049269894e-06,
+      "loss": 0.5059,
+      "step": 200500
+    },
+    {
+      "epoch": 9.958776199463594,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.3018774212774412e-06,
+      "loss": 0.4879,
+      "step": 200510
+    },
+    {
+      "epoch": 9.959272871759213,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 3.262143637627893e-06,
+      "loss": 0.4838,
+      "step": 200520
+    },
+    {
+      "epoch": 9.959769544054833,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.2224098539783454e-06,
+      "loss": 0.4625,
+      "step": 200530
+    },
+    {
+      "epoch": 9.960266216350451,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 3.1826760703287973e-06,
+      "loss": 0.5267,
+      "step": 200540
+    },
+    {
+      "epoch": 9.960762888646071,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.142942286679249e-06,
+      "loss": 0.4634,
+      "step": 200550
+    },
+    {
+      "epoch": 9.961259560941691,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.103208503029701e-06,
+      "loss": 0.4889,
+      "step": 200560
+    },
+    {
+      "epoch": 9.96175623323731,
+      "grad_norm": 0.251953125,
+      "learning_rate": 3.0634747193801533e-06,
+      "loss": 0.4917,
+      "step": 200570
+    },
+    {
+      "epoch": 9.96225290553293,
+      "grad_norm": 0.205078125,
+      "learning_rate": 3.023740935730605e-06,
+      "loss": 0.496,
+      "step": 200580
+    },
+    {
+      "epoch": 9.962749577828548,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.984007152081057e-06,
+      "loss": 0.4951,
+      "step": 200590
+    },
+    {
+      "epoch": 9.963246250124168,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.944273368431509e-06,
+      "loss": 0.5108,
+      "step": 200600
+    },
+    {
+      "epoch": 9.963742922419787,
+      "grad_norm": 0.236328125,
+      "learning_rate": 2.904539584781961e-06,
+      "loss": 0.4966,
+      "step": 200610
+    },
+    {
+      "epoch": 9.964239594715407,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.864805801132413e-06,
+      "loss": 0.517,
+      "step": 200620
+    },
+    {
+      "epoch": 9.964736267011027,
+      "grad_norm": 0.2109375,
+      "learning_rate": 2.825072017482865e-06,
+      "loss": 0.4991,
+      "step": 200630
+    },
+    {
+      "epoch": 9.965232939306645,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.785338233833317e-06,
+      "loss": 0.4462,
+      "step": 200640
+    },
+    {
+      "epoch": 9.965729611602265,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.745604450183769e-06,
+      "loss": 0.5046,
+      "step": 200650
+    },
+    {
+      "epoch": 9.966226283897884,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.7058706665342206e-06,
+      "loss": 0.4744,
+      "step": 200660
+    },
+    {
+      "epoch": 9.966722956193504,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 2.666136882884673e-06,
+      "loss": 0.498,
+      "step": 200670
+    },
+    {
+      "epoch": 9.967219628489122,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.6264030992351248e-06,
+      "loss": 0.5009,
+      "step": 200680
+    },
+    {
+      "epoch": 9.967716300784742,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.5866693155855767e-06,
+      "loss": 0.4931,
+      "step": 200690
+    },
+    {
+      "epoch": 9.968212973080362,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.546935531936029e-06,
+      "loss": 0.4721,
+      "step": 200700
+    },
+    {
+      "epoch": 9.96870964537598,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.507201748286481e-06,
+      "loss": 0.4798,
+      "step": 200710
+    },
+    {
+      "epoch": 9.9692063176716,
+      "grad_norm": 0.2109375,
+      "learning_rate": 2.4674679646369327e-06,
+      "loss": 0.4699,
+      "step": 200720
+    },
+    {
+      "epoch": 9.96970298996722,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.4277341809873846e-06,
+      "loss": 0.5015,
+      "step": 200730
+    },
+    {
+      "epoch": 9.97019966226284,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.3880003973378365e-06,
+      "loss": 0.4943,
+      "step": 200740
+    },
+    {
+      "epoch": 9.970696334558458,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.3482666136882887e-06,
+      "loss": 0.4559,
+      "step": 200750
+    },
+    {
+      "epoch": 9.971193006854078,
+      "grad_norm": 0.234375,
+      "learning_rate": 2.3085328300387406e-06,
+      "loss": 0.503,
+      "step": 200760
+    },
+    {
+      "epoch": 9.971689679149698,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 2.2687990463891925e-06,
+      "loss": 0.5,
+      "step": 200770
+    },
+    {
+      "epoch": 9.972186351445316,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 2.229065262739645e-06,
+      "loss": 0.4934,
+      "step": 200780
+    },
+    {
+      "epoch": 9.972683023740936,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.1893314790900962e-06,
+      "loss": 0.5046,
+      "step": 200790
+    },
+    {
+      "epoch": 9.973179696036555,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.149597695440548e-06,
+      "loss": 0.5024,
+      "step": 200800
+    },
+    {
+      "epoch": 9.973676368332175,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.1098639117910004e-06,
+      "loss": 0.4862,
+      "step": 200810
+    },
+    {
+      "epoch": 9.974173040627793,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.0701301281414523e-06,
+      "loss": 0.4961,
+      "step": 200820
+    },
+    {
+      "epoch": 9.974669712923413,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.0303963444919046e-06,
+      "loss": 0.4749,
+      "step": 200830
+    },
+    {
+      "epoch": 9.975166385219033,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.9906625608423565e-06,
+      "loss": 0.4709,
+      "step": 200840
+    },
+    {
+      "epoch": 9.975663057514652,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.9509287771928083e-06,
+      "loss": 0.5292,
+      "step": 200850
+    },
+    {
+      "epoch": 9.976159729810272,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.91119499354326e-06,
+      "loss": 0.4629,
+      "step": 200860
+    },
+    {
+      "epoch": 9.97665640210589,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.871461209893712e-06,
+      "loss": 0.4717,
+      "step": 200870
+    },
+    {
+      "epoch": 9.97715307440151,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.831727426244164e-06,
+      "loss": 0.4822,
+      "step": 200880
+    },
+    {
+      "epoch": 9.977649746697129,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.7919936425946163e-06,
+      "loss": 0.5102,
+      "step": 200890
+    },
+    {
+      "epoch": 9.978146418992749,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 1.7522598589450681e-06,
+      "loss": 0.4861,
+      "step": 200900
+    },
+    {
+      "epoch": 9.978643091288369,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 1.7125260752955202e-06,
+      "loss": 0.4787,
+      "step": 200910
+    },
+    {
+      "epoch": 9.979139763583987,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.672792291645972e-06,
+      "loss": 0.4888,
+      "step": 200920
+    },
+    {
+      "epoch": 9.979636435879607,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.633058507996424e-06,
+      "loss": 0.4792,
+      "step": 200930
+    },
+    {
+      "epoch": 9.980133108175226,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.593324724346876e-06,
+      "loss": 0.5013,
+      "step": 200940
+    },
+    {
+      "epoch": 9.980629780470846,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.553590940697328e-06,
+      "loss": 0.4623,
+      "step": 200950
+    },
+    {
+      "epoch": 9.981126452766464,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.51385715704778e-06,
+      "loss": 0.4763,
+      "step": 200960
+    },
+    {
+      "epoch": 9.981623125062084,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.4741233733982319e-06,
+      "loss": 0.5116,
+      "step": 200970
+    },
+    {
+      "epoch": 9.982119797357704,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 1.434389589748684e-06,
+      "loss": 0.484,
+      "step": 200980
+    },
+    {
+      "epoch": 9.982616469653323,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.3946558060991358e-06,
+      "loss": 0.4718,
+      "step": 200990
+    },
+    {
+      "epoch": 9.983113141948943,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.3549220224495877e-06,
+      "loss": 0.4857,
+      "step": 201000
+    },
+    {
+      "epoch": 9.983609814244561,
+      "grad_norm": 0.19921875,
+      "learning_rate": 1.3151882388000398e-06,
+      "loss": 0.4885,
+      "step": 201010
+    },
+    {
+      "epoch": 9.984106486540181,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.2754544551504919e-06,
+      "loss": 0.4994,
+      "step": 201020
+    },
+    {
+      "epoch": 9.9846031588358,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.2357206715009438e-06,
+      "loss": 0.5013,
+      "step": 201030
+    },
+    {
+      "epoch": 9.98509983113142,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 1.1959868878513956e-06,
+      "loss": 0.4846,
+      "step": 201040
+    },
+    {
+      "epoch": 9.98559650342704,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.1562531042018477e-06,
+      "loss": 0.4439,
+      "step": 201050
+    },
+    {
+      "epoch": 9.986093175722658,
+      "grad_norm": 0.22265625,
+      "learning_rate": 1.1165193205522998e-06,
+      "loss": 0.4679,
+      "step": 201060
+    },
+    {
+      "epoch": 9.986589848018278,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 1.0767855369027515e-06,
+      "loss": 0.4808,
+      "step": 201070
+    },
+    {
+      "epoch": 9.987086520313897,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.0370517532532036e-06,
+      "loss": 0.5098,
+      "step": 201080
+    },
+    {
+      "epoch": 9.987583192609517,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.973179696036556e-07,
+      "loss": 0.4858,
+      "step": 201090
+    },
+    {
+      "epoch": 9.988079864905135,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.575841859541075e-07,
+      "loss": 0.4841,
+      "step": 201100
+    },
+    {
+      "epoch": 9.988576537200755,
+      "grad_norm": 0.220703125,
+      "learning_rate": 9.178504023045594e-07,
+      "loss": 0.4603,
+      "step": 201110
+    },
+    {
+      "epoch": 9.989073209496375,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 8.781166186550115e-07,
+      "loss": 0.4741,
+      "step": 201120
+    },
+    {
+      "epoch": 9.989569881791994,
+      "grad_norm": 0.1953125,
+      "learning_rate": 8.383828350054634e-07,
+      "loss": 0.4818,
+      "step": 201130
+    },
+    {
+      "epoch": 9.990066554087614,
+      "grad_norm": 0.216796875,
+      "learning_rate": 7.986490513559154e-07,
+      "loss": 0.4931,
+      "step": 201140
+    },
+    {
+      "epoch": 9.990563226383232,
+      "grad_norm": 0.212890625,
+      "learning_rate": 7.589152677063674e-07,
+      "loss": 0.51,
+      "step": 201150
+    },
+    {
+      "epoch": 9.991059898678852,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.191814840568193e-07,
+      "loss": 0.4486,
+      "step": 201160
+    },
+    {
+      "epoch": 9.99155657097447,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 6.794477004072713e-07,
+      "loss": 0.5048,
+      "step": 201170
+    },
+    {
+      "epoch": 9.99205324327009,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.397139167577233e-07,
+      "loss": 0.5031,
+      "step": 201180
+    },
+    {
+      "epoch": 9.992549915565709,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.999801331081752e-07,
+      "loss": 0.5323,
+      "step": 201190
+    },
+    {
+      "epoch": 9.993046587861329,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 5.602463494586273e-07,
+      "loss": 0.4506,
+      "step": 201200
+    },
+    {
+      "epoch": 9.99354326015695,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.205125658090792e-07,
+      "loss": 0.4974,
+      "step": 201210
+    },
+    {
+      "epoch": 9.994039932452567,
+      "grad_norm": 0.25,
+      "learning_rate": 4.807787821595312e-07,
+      "loss": 0.5025,
+      "step": 201220
+    },
+    {
+      "epoch": 9.994536604748188,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 4.4104499850998314e-07,
+      "loss": 0.5047,
+      "step": 201230
+    },
+    {
+      "epoch": 9.995033277043806,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 4.013112148604351e-07,
+      "loss": 0.4562,
+      "step": 201240
+    },
+    {
+      "epoch": 9.995529949339426,
+      "grad_norm": 0.2109375,
+      "learning_rate": 3.6157743121088705e-07,
+      "loss": 0.4832,
+      "step": 201250
+    },
+    {
+      "epoch": 9.996026621635044,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.2184364756133903e-07,
+      "loss": 0.5173,
+      "step": 201260
+    },
+    {
+      "epoch": 9.996523293930665,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.82109863911791e-07,
+      "loss": 0.4914,
+      "step": 201270
+    },
+    {
+      "epoch": 9.997019966226285,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.42376080262243e-07,
+      "loss": 0.5044,
+      "step": 201280
+    },
+    {
+      "epoch": 9.997516638521903,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.0264229661269497e-07,
+      "loss": 0.4748,
+      "step": 201290
+    },
+    {
+      "epoch": 9.998013310817523,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.6290851296314692e-07,
+      "loss": 0.4938,
+      "step": 201300
+    },
+    {
+      "epoch": 9.998509983113141,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.231747293135989e-07,
+      "loss": 0.5312,
+      "step": 201310
+    },
+    {
+      "epoch": 9.999006655408762,
+      "grad_norm": 0.1875,
+      "learning_rate": 8.344094566405085e-08,
+      "loss": 0.4645,
+      "step": 201320
+    },
+    {
+      "epoch": 9.99950332770438,
+      "grad_norm": 0.2421875,
+      "learning_rate": 4.370716201450283e-08,
+      "loss": 0.4832,
+      "step": 201330
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.203125,
+      "learning_rate": 3.9733783649548035e-09,
+      "loss": 0.4821,
+      "step": 201340
+    },
+    {
+      "epoch": 10.0,
+      "step": 201340,
+      "total_flos": 1.6346195493598003e+18,
+      "train_loss": 0.5436877673617467,
+      "train_runtime": 108051.857,
+      "train_samples_per_second": 7.453,
+      "train_steps_per_second": 1.863
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 201340,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.6346195493598003e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}