diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,37485 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999953267752413,
+  "eval_steps": 500,
+  "global_step": 26748,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0001869289903497909,
+      "grad_norm": 2.171574354171753,
+      "learning_rate": 1.9999998275643267e-05,
+      "loss": 0.8437,
+      "step": 5
+    },
+    {
+      "epoch": 0.0003738579806995818,
+      "grad_norm": 2.646630048751831,
+      "learning_rate": 1.999999310257367e-05,
+      "loss": 0.6249,
+      "step": 10
+    },
+    {
+      "epoch": 0.0005607869710493726,
+      "grad_norm": 1.4575234651565552,
+      "learning_rate": 1.9999984480792985e-05,
+      "loss": 0.4614,
+      "step": 15
+    },
+    {
+      "epoch": 0.0007477159613991635,
+      "grad_norm": 0.6195482015609741,
+      "learning_rate": 1.9999972410304184e-05,
+      "loss": 0.4029,
+      "step": 20
+    },
+    {
+      "epoch": 0.0009346449517489544,
+      "grad_norm": 1.0422546863555908,
+      "learning_rate": 1.999995689111144e-05,
+      "loss": 0.4514,
+      "step": 25
+    },
+    {
+      "epoch": 0.0011215739420987452,
+      "grad_norm": 1.1793930530548096,
+      "learning_rate": 1.9999937923220094e-05,
+      "loss": 0.3913,
+      "step": 30
+    },
+    {
+      "epoch": 0.0013085029324485362,
+      "grad_norm": 0.8712885975837708,
+      "learning_rate": 1.9999915506636697e-05,
+      "loss": 0.4137,
+      "step": 35
+    },
+    {
+      "epoch": 0.001495431922798327,
+      "grad_norm": 0.795931875705719,
+      "learning_rate": 1.9999889641368975e-05,
+      "loss": 0.4616,
+      "step": 40
+    },
+    {
+      "epoch": 0.0016823609131481178,
+      "grad_norm": 1.453735589981079,
+      "learning_rate": 1.9999860327425846e-05,
+      "loss": 0.4453,
+      "step": 45
+    },
+    {
+      "epoch": 0.0018692899034979088,
+      "grad_norm": 0.865328311920166,
+      "learning_rate": 1.9999827564817424e-05,
+      "loss": 0.358,
+      "step": 50
+    },
+    {
+      "epoch": 0.0020562188938476997,
+      "grad_norm": 0.9025278091430664,
+      "learning_rate": 1.9999791353555008e-05,
+      "loss": 0.4263,
+      "step": 55
+    },
+    {
+      "epoch": 0.0022431478841974904,
+      "grad_norm": 0.8391484618186951,
+      "learning_rate": 1.9999751693651083e-05,
+      "loss": 0.3838,
+      "step": 60
+    },
+    {
+      "epoch": 0.002430076874547281,
+      "grad_norm": 0.7536634206771851,
+      "learning_rate": 1.9999708585119328e-05,
+      "loss": 0.5194,
+      "step": 65
+    },
+    {
+      "epoch": 0.0026170058648970723,
+      "grad_norm": 1.1140562295913696,
+      "learning_rate": 1.999966202797461e-05,
+      "loss": 0.379,
+      "step": 70
+    },
+    {
+      "epoch": 0.002803934855246863,
+      "grad_norm": 0.9421278238296509,
+      "learning_rate": 1.9999612022232982e-05,
+      "loss": 0.3929,
+      "step": 75
+    },
+    {
+      "epoch": 0.002990863845596654,
+      "grad_norm": 0.5064008831977844,
+      "learning_rate": 1.9999558567911697e-05,
+      "loss": 0.4323,
+      "step": 80
+    },
+    {
+      "epoch": 0.003177792835946445,
+      "grad_norm": 0.7780022621154785,
+      "learning_rate": 1.9999501665029185e-05,
+      "loss": 0.3719,
+      "step": 85
+    },
+    {
+      "epoch": 0.0033647218262962356,
+      "grad_norm": 1.0911474227905273,
+      "learning_rate": 1.9999441313605068e-05,
+      "loss": 0.3523,
+      "step": 90
+    },
+    {
+      "epoch": 0.003551650816646027,
+      "grad_norm": 0.665628969669342,
+      "learning_rate": 1.9999377513660167e-05,
+      "loss": 0.3953,
+      "step": 95
+    },
+    {
+      "epoch": 0.0037385798069958175,
+      "grad_norm": 1.2322030067443848,
+      "learning_rate": 1.9999310265216472e-05,
+      "loss": 0.4117,
+      "step": 100
+    },
+    {
+      "epoch": 0.003925508797345608,
+      "grad_norm": 0.8758599758148193,
+      "learning_rate": 1.999923956829719e-05,
+      "loss": 0.3031,
+      "step": 105
+    },
+    {
+      "epoch": 0.004112437787695399,
+      "grad_norm": 0.8419223427772522,
+      "learning_rate": 1.9999165422926696e-05,
+      "loss": 0.3016,
+      "step": 110
+    },
+    {
+      "epoch": 0.00429936677804519,
+      "grad_norm": 1.1489617824554443,
+      "learning_rate": 1.9999087829130554e-05,
+      "loss": 0.4597,
+      "step": 115
+    },
+    {
+      "epoch": 0.004486295768394981,
+      "grad_norm": 0.46470773220062256,
+      "learning_rate": 1.9999006786935532e-05,
+      "loss": 0.337,
+      "step": 120
+    },
+    {
+      "epoch": 0.004673224758744772,
+      "grad_norm": 0.4822236895561218,
+      "learning_rate": 1.999892229636958e-05,
+      "loss": 0.3476,
+      "step": 125
+    },
+    {
+      "epoch": 0.004860153749094562,
+      "grad_norm": 0.691170334815979,
+      "learning_rate": 1.9998834357461834e-05,
+      "loss": 0.3542,
+      "step": 130
+    },
+    {
+      "epoch": 0.0050470827394443534,
+      "grad_norm": 0.6940286755561829,
+      "learning_rate": 1.9998742970242614e-05,
+      "loss": 0.3268,
+      "step": 135
+    },
+    {
+      "epoch": 0.005234011729794145,
+      "grad_norm": 0.943358302116394,
+      "learning_rate": 1.999864813474345e-05,
+      "loss": 0.3943,
+      "step": 140
+    },
+    {
+      "epoch": 0.005420940720143935,
+      "grad_norm": 1.0193995237350464,
+      "learning_rate": 1.9998549850997044e-05,
+      "loss": 0.3066,
+      "step": 145
+    },
+    {
+      "epoch": 0.005607869710493726,
+      "grad_norm": 0.7878953814506531,
+      "learning_rate": 1.999844811903728e-05,
+      "loss": 0.3635,
+      "step": 150
+    },
+    {
+      "epoch": 0.005794798700843517,
+      "grad_norm": 0.4438617527484894,
+      "learning_rate": 1.9998342938899257e-05,
+      "loss": 0.4297,
+      "step": 155
+    },
+    {
+      "epoch": 0.005981727691193308,
+      "grad_norm": 0.35109004378318787,
+      "learning_rate": 1.999823431061924e-05,
+      "loss": 0.4432,
+      "step": 160
+    },
+    {
+      "epoch": 0.006168656681543099,
+      "grad_norm": 0.7942967414855957,
+      "learning_rate": 1.99981222342347e-05,
+      "loss": 0.4196,
+      "step": 165
+    },
+    {
+      "epoch": 0.00635558567189289,
+      "grad_norm": 0.8898890018463135,
+      "learning_rate": 1.9998006709784283e-05,
+      "loss": 0.3688,
+      "step": 170
+    },
+    {
+      "epoch": 0.006542514662242681,
+      "grad_norm": 1.1046160459518433,
+      "learning_rate": 1.999788773730783e-05,
+      "loss": 0.3895,
+      "step": 175
+    },
+    {
+      "epoch": 0.006729443652592471,
+      "grad_norm": 0.5191996693611145,
+      "learning_rate": 1.999776531684637e-05,
+      "loss": 0.3963,
+      "step": 180
+    },
+    {
+      "epoch": 0.006916372642942262,
+      "grad_norm": 0.5099379420280457,
+      "learning_rate": 1.9997639448442125e-05,
+      "loss": 0.3328,
+      "step": 185
+    },
+    {
+      "epoch": 0.007103301633292054,
+      "grad_norm": 1.2763317823410034,
+      "learning_rate": 1.9997510132138505e-05,
+      "loss": 0.2926,
+      "step": 190
+    },
+    {
+      "epoch": 0.007290230623641844,
+      "grad_norm": 1.150545358657837,
+      "learning_rate": 1.9997377367980104e-05,
+      "loss": 0.4188,
+      "step": 195
+    },
+    {
+      "epoch": 0.007477159613991635,
+      "grad_norm": 0.8739756941795349,
+      "learning_rate": 1.999724115601271e-05,
+      "loss": 0.4185,
+      "step": 200
+    },
+    {
+      "epoch": 0.007664088604341426,
+      "grad_norm": 1.0264869928359985,
+      "learning_rate": 1.99971014962833e-05,
+      "loss": 0.3185,
+      "step": 205
+    },
+    {
+      "epoch": 0.007851017594691216,
+      "grad_norm": 0.5631968975067139,
+      "learning_rate": 1.9996958388840036e-05,
+      "loss": 0.3156,
+      "step": 210
+    },
+    {
+      "epoch": 0.008037946585041008,
+      "grad_norm": 0.6525606513023376,
+      "learning_rate": 1.999681183373227e-05,
+      "loss": 0.3579,
+      "step": 215
+    },
+    {
+      "epoch": 0.008224875575390799,
+      "grad_norm": 0.894953727722168,
+      "learning_rate": 1.999666183101055e-05,
+      "loss": 0.3544,
+      "step": 220
+    },
+    {
+      "epoch": 0.00841180456574059,
+      "grad_norm": 0.5953158140182495,
+      "learning_rate": 1.9996508380726608e-05,
+      "loss": 0.3366,
+      "step": 225
+    },
+    {
+      "epoch": 0.00859873355609038,
+      "grad_norm": 0.47904253005981445,
+      "learning_rate": 1.9996351482933355e-05,
+      "loss": 0.448,
+      "step": 230
+    },
+    {
+      "epoch": 0.00878566254644017,
+      "grad_norm": 0.9660627841949463,
+      "learning_rate": 1.9996191137684913e-05,
+      "loss": 0.4367,
+      "step": 235
+    },
+    {
+      "epoch": 0.008972591536789962,
+      "grad_norm": 1.3929287195205688,
+      "learning_rate": 1.9996027345036574e-05,
+      "loss": 0.4314,
+      "step": 240
+    },
+    {
+      "epoch": 0.009159520527139753,
+      "grad_norm": 0.5760064721107483,
+      "learning_rate": 1.999586010504482e-05,
+      "loss": 0.389,
+      "step": 245
+    },
+    {
+      "epoch": 0.009346449517489544,
+      "grad_norm": 1.9712883234024048,
+      "learning_rate": 1.999568941776734e-05,
+      "loss": 0.3424,
+      "step": 250
+    },
+    {
+      "epoch": 0.009533378507839335,
+      "grad_norm": 0.6370198726654053,
+      "learning_rate": 1.9995515283262993e-05,
+      "loss": 0.4801,
+      "step": 255
+    },
+    {
+      "epoch": 0.009720307498189125,
+      "grad_norm": 1.003278136253357,
+      "learning_rate": 1.999533770159183e-05,
+      "loss": 0.3515,
+      "step": 260
+    },
+    {
+      "epoch": 0.009907236488538916,
+      "grad_norm": 0.4942162036895752,
+      "learning_rate": 1.9995156672815096e-05,
+      "loss": 0.3689,
+      "step": 265
+    },
+    {
+      "epoch": 0.010094165478888707,
+      "grad_norm": 0.5202749371528625,
+      "learning_rate": 1.9994972196995223e-05,
+      "loss": 0.3163,
+      "step": 270
+    },
+    {
+      "epoch": 0.010281094469238498,
+      "grad_norm": 0.6780266165733337,
+      "learning_rate": 1.9994784274195834e-05,
+      "loss": 0.3699,
+      "step": 275
+    },
+    {
+      "epoch": 0.01046802345958829,
+      "grad_norm": 0.40867510437965393,
+      "learning_rate": 1.9994592904481732e-05,
+      "loss": 0.3337,
+      "step": 280
+    },
+    {
+      "epoch": 0.01065495244993808,
+      "grad_norm": 0.36129143834114075,
+      "learning_rate": 1.999439808791892e-05,
+      "loss": 0.3753,
+      "step": 285
+    },
+    {
+      "epoch": 0.01084188144028787,
+      "grad_norm": 0.4958774447441101,
+      "learning_rate": 1.9994199824574583e-05,
+      "loss": 0.3724,
+      "step": 290
+    },
+    {
+      "epoch": 0.011028810430637661,
+      "grad_norm": 0.9592061042785645,
+      "learning_rate": 1.9993998114517096e-05,
+      "loss": 0.4238,
+      "step": 295
+    },
+    {
+      "epoch": 0.011215739420987452,
+      "grad_norm": 0.6660071015357971,
+      "learning_rate": 1.9993792957816027e-05,
+      "loss": 0.3774,
+      "step": 300
+    },
+    {
+      "epoch": 0.011402668411337243,
+      "grad_norm": 1.37131929397583,
+      "learning_rate": 1.999358435454212e-05,
+      "loss": 0.4088,
+      "step": 305
+    },
+    {
+      "epoch": 0.011589597401687034,
+      "grad_norm": 0.8485438227653503,
+      "learning_rate": 1.9993372304767327e-05,
+      "loss": 0.3734,
+      "step": 310
+    },
+    {
+      "epoch": 0.011776526392036826,
+      "grad_norm": 0.41537079215049744,
+      "learning_rate": 1.999315680856477e-05,
+      "loss": 0.2756,
+      "step": 315
+    },
+    {
+      "epoch": 0.011963455382386617,
+      "grad_norm": 0.36821046471595764,
+      "learning_rate": 1.999293786600877e-05,
+      "loss": 0.3323,
+      "step": 320
+    },
+    {
+      "epoch": 0.012150384372736406,
+      "grad_norm": 0.6895926594734192,
+      "learning_rate": 1.9992715477174832e-05,
+      "loss": 0.2989,
+      "step": 325
+    },
+    {
+      "epoch": 0.012337313363086197,
+      "grad_norm": 0.7400907278060913,
+      "learning_rate": 1.9992489642139654e-05,
+      "loss": 0.4937,
+      "step": 330
+    },
+    {
+      "epoch": 0.012524242353435988,
+      "grad_norm": 0.6539915800094604,
+      "learning_rate": 1.999226036098112e-05,
+      "loss": 0.2889,
+      "step": 335
+    },
+    {
+      "epoch": 0.01271117134378578,
+      "grad_norm": 0.7676389217376709,
+      "learning_rate": 1.9992027633778303e-05,
+      "loss": 0.3172,
+      "step": 340
+    },
+    {
+      "epoch": 0.01289810033413557,
+      "grad_norm": 0.6981975436210632,
+      "learning_rate": 1.9991791460611464e-05,
+      "loss": 0.3242,
+      "step": 345
+    },
+    {
+      "epoch": 0.013085029324485362,
+      "grad_norm": 0.6231608986854553,
+      "learning_rate": 1.999155184156205e-05,
+      "loss": 0.4696,
+      "step": 350
+    },
+    {
+      "epoch": 0.013271958314835151,
+      "grad_norm": 0.5723335146903992,
+      "learning_rate": 1.9991308776712697e-05,
+      "loss": 0.3813,
+      "step": 355
+    },
+    {
+      "epoch": 0.013458887305184943,
+      "grad_norm": 0.5270363092422485,
+      "learning_rate": 1.9991062266147237e-05,
+      "loss": 0.3927,
+      "step": 360
+    },
+    {
+      "epoch": 0.013645816295534734,
+      "grad_norm": 1.0495415925979614,
+      "learning_rate": 1.999081230995068e-05,
+      "loss": 0.3547,
+      "step": 365
+    },
+    {
+      "epoch": 0.013832745285884525,
+      "grad_norm": 0.9163959622383118,
+      "learning_rate": 1.9990558908209234e-05,
+      "loss": 0.3432,
+      "step": 370
+    },
+    {
+      "epoch": 0.014019674276234316,
+      "grad_norm": 1.1537407636642456,
+      "learning_rate": 1.9990302061010282e-05,
+      "loss": 0.3514,
+      "step": 375
+    },
+    {
+      "epoch": 0.014206603266584107,
+      "grad_norm": 0.7320118546485901,
+      "learning_rate": 1.999004176844241e-05,
+      "loss": 0.4305,
+      "step": 380
+    },
+    {
+      "epoch": 0.014393532256933897,
+      "grad_norm": 0.47620531916618347,
+      "learning_rate": 1.9989778030595385e-05,
+      "loss": 0.3117,
+      "step": 385
+    },
+    {
+      "epoch": 0.014580461247283688,
+      "grad_norm": 0.5840729475021362,
+      "learning_rate": 1.9989510847560157e-05,
+      "loss": 0.3125,
+      "step": 390
+    },
+    {
+      "epoch": 0.014767390237633479,
+      "grad_norm": 0.5005705952644348,
+      "learning_rate": 1.9989240219428873e-05,
+      "loss": 0.2977,
+      "step": 395
+    },
+    {
+      "epoch": 0.01495431922798327,
+      "grad_norm": 0.8225285410881042,
+      "learning_rate": 1.9988966146294867e-05,
+      "loss": 0.3511,
+      "step": 400
+    },
+    {
+      "epoch": 0.015141248218333061,
+      "grad_norm": 0.5237758755683899,
+      "learning_rate": 1.9988688628252656e-05,
+      "loss": 0.3672,
+      "step": 405
+    },
+    {
+      "epoch": 0.015328177208682852,
+      "grad_norm": 0.6276610493659973,
+      "learning_rate": 1.9988407665397952e-05,
+      "loss": 0.3592,
+      "step": 410
+    },
+    {
+      "epoch": 0.015515106199032642,
+      "grad_norm": 0.6155140995979309,
+      "learning_rate": 1.9988123257827646e-05,
+      "loss": 0.3913,
+      "step": 415
+    },
+    {
+      "epoch": 0.015702035189382433,
+      "grad_norm": 1.1031259298324585,
+      "learning_rate": 1.9987835405639827e-05,
+      "loss": 0.3597,
+      "step": 420
+    },
+    {
+      "epoch": 0.015888964179732224,
+      "grad_norm": 0.9590380787849426,
+      "learning_rate": 1.9987544108933758e-05,
+      "loss": 0.3436,
+      "step": 425
+    },
+    {
+      "epoch": 0.016075893170082015,
+      "grad_norm": 0.6944395899772644,
+      "learning_rate": 1.998724936780991e-05,
+      "loss": 0.4223,
+      "step": 430
+    },
+    {
+      "epoch": 0.016262822160431806,
+      "grad_norm": 0.8819741606712341,
+      "learning_rate": 1.9986951182369923e-05,
+      "loss": 0.3349,
+      "step": 435
+    },
+    {
+      "epoch": 0.016449751150781598,
+      "grad_norm": 0.8658530712127686,
+      "learning_rate": 1.998664955271664e-05,
+      "loss": 0.3447,
+      "step": 440
+    },
+    {
+      "epoch": 0.01663668014113139,
+      "grad_norm": 0.7974675297737122,
+      "learning_rate": 1.9986344478954078e-05,
+      "loss": 0.4101,
+      "step": 445
+    },
+    {
+      "epoch": 0.01682360913148118,
+      "grad_norm": 0.46253588795661926,
+      "learning_rate": 1.998603596118745e-05,
+      "loss": 0.3763,
+      "step": 450
+    },
+    {
+      "epoch": 0.01701053812183097,
+      "grad_norm": 0.67314213514328,
+      "learning_rate": 1.9985723999523154e-05,
+      "loss": 0.3419,
+      "step": 455
+    },
+    {
+      "epoch": 0.01719746711218076,
+      "grad_norm": 0.5673795938491821,
+      "learning_rate": 1.998540859406878e-05,
+      "loss": 0.3533,
+      "step": 460
+    },
+    {
+      "epoch": 0.01738439610253055,
+      "grad_norm": 0.34820371866226196,
+      "learning_rate": 1.99850897449331e-05,
+      "loss": 0.3095,
+      "step": 465
+    },
+    {
+      "epoch": 0.01757132509288034,
+      "grad_norm": 1.0526245832443237,
+      "learning_rate": 1.998476745222607e-05,
+      "loss": 0.3775,
+      "step": 470
+    },
+    {
+      "epoch": 0.017758254083230132,
+      "grad_norm": 0.2607547342777252,
+      "learning_rate": 1.9984441716058855e-05,
+      "loss": 0.319,
+      "step": 475
+    },
+    {
+      "epoch": 0.017945183073579923,
+      "grad_norm": 0.5493923425674438,
+      "learning_rate": 1.9984112536543774e-05,
+      "loss": 0.2914,
+      "step": 480
+    },
+    {
+      "epoch": 0.018132112063929715,
+      "grad_norm": 0.6481868028640747,
+      "learning_rate": 1.9983779913794366e-05,
+      "loss": 0.39,
+      "step": 485
+    },
+    {
+      "epoch": 0.018319041054279506,
+      "grad_norm": 0.4423280954360962,
+      "learning_rate": 1.9983443847925334e-05,
+      "loss": 0.383,
+      "step": 490
+    },
+    {
+      "epoch": 0.018505970044629297,
+      "grad_norm": 0.24521741271018982,
+      "learning_rate": 1.998310433905258e-05,
+      "loss": 0.3797,
+      "step": 495
+    },
+    {
+      "epoch": 0.018692899034979088,
+      "grad_norm": 0.4319721460342407,
+      "learning_rate": 1.9982761387293196e-05,
+      "loss": 0.3635,
+      "step": 500
+    },
+    {
+      "epoch": 0.01887982802532888,
+      "grad_norm": 0.6990941166877747,
+      "learning_rate": 1.998241499276545e-05,
+      "loss": 0.2869,
+      "step": 505
+    },
+    {
+      "epoch": 0.01906675701567867,
+      "grad_norm": 0.6501472592353821,
+      "learning_rate": 1.9982065155588803e-05,
+      "loss": 0.4004,
+      "step": 510
+    },
+    {
+      "epoch": 0.01925368600602846,
+      "grad_norm": 0.5194607973098755,
+      "learning_rate": 1.9981711875883908e-05,
+      "loss": 0.3062,
+      "step": 515
+    },
+    {
+      "epoch": 0.01944061499637825,
+      "grad_norm": 0.5444199442863464,
+      "learning_rate": 1.9981355153772603e-05,
+      "loss": 0.3218,
+      "step": 520
+    },
+    {
+      "epoch": 0.01962754398672804,
+      "grad_norm": 0.8375357985496521,
+      "learning_rate": 1.9980994989377902e-05,
+      "loss": 0.2581,
+      "step": 525
+    },
+    {
+      "epoch": 0.01981447297707783,
+      "grad_norm": 0.47779718041419983,
+      "learning_rate": 1.998063138282402e-05,
+      "loss": 0.2845,
+      "step": 530
+    },
+    {
+      "epoch": 0.020001401967427623,
+      "grad_norm": 0.5198971629142761,
+      "learning_rate": 1.998026433423636e-05,
+      "loss": 0.3539,
+      "step": 535
+    },
+    {
+      "epoch": 0.020188330957777414,
+      "grad_norm": 0.6025378108024597,
+      "learning_rate": 1.9979893843741498e-05,
+      "loss": 0.3003,
+      "step": 540
+    },
+    {
+      "epoch": 0.020375259948127205,
+      "grad_norm": 0.454045832157135,
+      "learning_rate": 1.997951991146721e-05,
+      "loss": 0.3132,
+      "step": 545
+    },
+    {
+      "epoch": 0.020562188938476996,
+      "grad_norm": 0.9200149178504944,
+      "learning_rate": 1.9979142537542455e-05,
+      "loss": 0.3469,
+      "step": 550
+    },
+    {
+      "epoch": 0.020749117928826787,
+      "grad_norm": 0.5452691912651062,
+      "learning_rate": 1.997876172209738e-05,
+      "loss": 0.3445,
+      "step": 555
+    },
+    {
+      "epoch": 0.02093604691917658,
+      "grad_norm": 0.46097660064697266,
+      "learning_rate": 1.997837746526331e-05,
+      "loss": 0.3302,
+      "step": 560
+    },
+    {
+      "epoch": 0.02112297590952637,
+      "grad_norm": 0.7622095346450806,
+      "learning_rate": 1.997798976717277e-05,
+      "loss": 0.3539,
+      "step": 565
+    },
+    {
+      "epoch": 0.02130990489987616,
+      "grad_norm": 0.44673827290534973,
+      "learning_rate": 1.9977598627959467e-05,
+      "loss": 0.3418,
+      "step": 570
+    },
+    {
+      "epoch": 0.021496833890225952,
+      "grad_norm": 0.38425078988075256,
+      "learning_rate": 1.9977204047758293e-05,
+      "loss": 0.342,
+      "step": 575
+    },
+    {
+      "epoch": 0.02168376288057574,
+      "grad_norm": 1.0525974035263062,
+      "learning_rate": 1.997680602670532e-05,
+      "loss": 0.3249,
+      "step": 580
+    },
+    {
+      "epoch": 0.02187069187092553,
+      "grad_norm": 0.5911397933959961,
+      "learning_rate": 1.9976404564937825e-05,
+      "loss": 0.2806,
+      "step": 585
+    },
+    {
+      "epoch": 0.022057620861275322,
+      "grad_norm": 0.7274040579795837,
+      "learning_rate": 1.9975999662594254e-05,
+      "loss": 0.3742,
+      "step": 590
+    },
+    {
+      "epoch": 0.022244549851625113,
+      "grad_norm": 0.8570606708526611,
+      "learning_rate": 1.9975591319814248e-05,
+      "loss": 0.4207,
+      "step": 595
+    },
+    {
+      "epoch": 0.022431478841974904,
+      "grad_norm": 0.6169318556785583,
+      "learning_rate": 1.9975179536738633e-05,
+      "loss": 0.3025,
+      "step": 600
+    },
+    {
+      "epoch": 0.022618407832324695,
+      "grad_norm": 0.7750716805458069,
+      "learning_rate": 1.997476431350942e-05,
+      "loss": 0.2713,
+      "step": 605
+    },
+    {
+      "epoch": 0.022805336822674487,
+      "grad_norm": 0.6509002447128296,
+      "learning_rate": 1.9974345650269812e-05,
+      "loss": 0.2948,
+      "step": 610
+    },
+    {
+      "epoch": 0.022992265813024278,
+      "grad_norm": 0.875956654548645,
+      "learning_rate": 1.9973923547164183e-05,
+      "loss": 0.4136,
+      "step": 615
+    },
+    {
+      "epoch": 0.02317919480337407,
+      "grad_norm": 0.5139532089233398,
+      "learning_rate": 1.9973498004338115e-05,
+      "loss": 0.3333,
+      "step": 620
+    },
+    {
+      "epoch": 0.02336612379372386,
+      "grad_norm": 0.7778888940811157,
+      "learning_rate": 1.9973069021938366e-05,
+      "loss": 0.3016,
+      "step": 625
+    },
+    {
+      "epoch": 0.02355305278407365,
+      "grad_norm": 0.4906633794307709,
+      "learning_rate": 1.9972636600112873e-05,
+      "loss": 0.3478,
+      "step": 630
+    },
+    {
+      "epoch": 0.023739981774423442,
+      "grad_norm": 0.43176034092903137,
+      "learning_rate": 1.997220073901077e-05,
+      "loss": 0.2925,
+      "step": 635
+    },
+    {
+      "epoch": 0.023926910764773233,
+      "grad_norm": 0.6139638423919678,
+      "learning_rate": 1.997176143878237e-05,
+      "loss": 0.336,
+      "step": 640
+    },
+    {
+      "epoch": 0.02411383975512302,
+      "grad_norm": 0.5035068988800049,
+      "learning_rate": 1.9971318699579177e-05,
+      "loss": 0.322,
+      "step": 645
+    },
+    {
+      "epoch": 0.024300768745472812,
+      "grad_norm": 0.8413116931915283,
+      "learning_rate": 1.997087252155388e-05,
+      "loss": 0.437,
+      "step": 650
+    },
+    {
+      "epoch": 0.024487697735822603,
+      "grad_norm": 0.33662039041519165,
+      "learning_rate": 1.9970422904860352e-05,
+      "loss": 0.2931,
+      "step": 655
+    },
+    {
+      "epoch": 0.024674626726172395,
+      "grad_norm": 0.6480875611305237,
+      "learning_rate": 1.9969969849653652e-05,
+      "loss": 0.2838,
+      "step": 660
+    },
+    {
+      "epoch": 0.024861555716522186,
+      "grad_norm": 1.1229877471923828,
+      "learning_rate": 1.9969513356090027e-05,
+      "loss": 0.3157,
+      "step": 665
+    },
+    {
+      "epoch": 0.025048484706871977,
+      "grad_norm": 0.5545793771743774,
+      "learning_rate": 1.9969053424326908e-05,
+      "loss": 0.4423,
+      "step": 670
+    },
+    {
+      "epoch": 0.025235413697221768,
+      "grad_norm": 0.5844056010246277,
+      "learning_rate": 1.9968590054522914e-05,
+      "loss": 0.4168,
+      "step": 675
+    },
+    {
+      "epoch": 0.02542234268757156,
+      "grad_norm": 0.5306451916694641,
+      "learning_rate": 1.996812324683784e-05,
+      "loss": 0.3978,
+      "step": 680
+    },
+    {
+      "epoch": 0.02560927167792135,
+      "grad_norm": 0.4964573085308075,
+      "learning_rate": 1.996765300143269e-05,
+      "loss": 0.351,
+      "step": 685
+    },
+    {
+      "epoch": 0.02579620066827114,
+      "grad_norm": 0.4789879024028778,
+      "learning_rate": 1.9967179318469626e-05,
+      "loss": 0.3372,
+      "step": 690
+    },
+    {
+      "epoch": 0.025983129658620933,
+      "grad_norm": 0.48183321952819824,
+      "learning_rate": 1.996670219811201e-05,
+      "loss": 0.2788,
+      "step": 695
+    },
+    {
+      "epoch": 0.026170058648970724,
+      "grad_norm": 0.2901974320411682,
+      "learning_rate": 1.996622164052439e-05,
+      "loss": 0.3617,
+      "step": 700
+    },
+    {
+      "epoch": 0.02635698763932051,
+      "grad_norm": 0.8643401861190796,
+      "learning_rate": 1.996573764587249e-05,
+      "loss": 0.3683,
+      "step": 705
+    },
+    {
+      "epoch": 0.026543916629670303,
+      "grad_norm": 0.6790117621421814,
+      "learning_rate": 1.9965250214323232e-05,
+      "loss": 0.4397,
+      "step": 710
+    },
+    {
+      "epoch": 0.026730845620020094,
+      "grad_norm": 0.7502917051315308,
+      "learning_rate": 1.996475934604472e-05,
+      "loss": 0.3743,
+      "step": 715
+    },
+    {
+      "epoch": 0.026917774610369885,
+      "grad_norm": 0.6247079372406006,
+      "learning_rate": 1.996426504120623e-05,
+      "loss": 0.3414,
+      "step": 720
+    },
+    {
+      "epoch": 0.027104703600719676,
+      "grad_norm": 1.0865947008132935,
+      "learning_rate": 1.9963767299978243e-05,
+      "loss": 0.332,
+      "step": 725
+    },
+    {
+      "epoch": 0.027291632591069467,
+      "grad_norm": 0.8259230256080627,
+      "learning_rate": 1.9963266122532416e-05,
+      "loss": 0.3709,
+      "step": 730
+    },
+    {
+      "epoch": 0.02747856158141926,
+      "grad_norm": 0.7781583070755005,
+      "learning_rate": 1.9962761509041578e-05,
+      "loss": 0.357,
+      "step": 735
+    },
+    {
+      "epoch": 0.02766549057176905,
+      "grad_norm": 0.4106627404689789,
+      "learning_rate": 1.996225345967977e-05,
+      "loss": 0.4027,
+      "step": 740
+    },
+    {
+      "epoch": 0.02785241956211884,
+      "grad_norm": 0.538112223148346,
+      "learning_rate": 1.99617419746222e-05,
+      "loss": 0.3817,
+      "step": 745
+    },
+    {
+      "epoch": 0.028039348552468632,
+      "grad_norm": 0.4181678593158722,
+      "learning_rate": 1.996122705404526e-05,
+      "loss": 0.3379,
+      "step": 750
+    },
+    {
+      "epoch": 0.028226277542818423,
+      "grad_norm": 0.7423063516616821,
+      "learning_rate": 1.9960708698126536e-05,
+      "loss": 0.3501,
+      "step": 755
+    },
+    {
+      "epoch": 0.028413206533168214,
+      "grad_norm": 0.9700421690940857,
+      "learning_rate": 1.996018690704479e-05,
+      "loss": 0.3911,
+      "step": 760
+    },
+    {
+      "epoch": 0.028600135523518002,
+      "grad_norm": 0.6143549084663391,
+      "learning_rate": 1.995966168097998e-05,
+      "loss": 0.3379,
+      "step": 765
+    },
+    {
+      "epoch": 0.028787064513867793,
+      "grad_norm": 0.5459659695625305,
+      "learning_rate": 1.995913302011323e-05,
+      "loss": 0.3502,
+      "step": 770
+    },
+    {
+      "epoch": 0.028973993504217584,
+      "grad_norm": 0.8577987551689148,
+      "learning_rate": 1.9958600924626873e-05,
+      "loss": 0.3926,
+      "step": 775
+    },
+    {
+      "epoch": 0.029160922494567375,
+      "grad_norm": 0.8069790601730347,
+      "learning_rate": 1.9958065394704406e-05,
+      "loss": 0.3648,
+      "step": 780
+    },
+    {
+      "epoch": 0.029347851484917167,
+      "grad_norm": 1.1032791137695312,
+      "learning_rate": 1.9957526430530514e-05,
+      "loss": 0.3606,
+      "step": 785
+    },
+    {
+      "epoch": 0.029534780475266958,
+      "grad_norm": 0.5585815906524658,
+      "learning_rate": 1.995698403229108e-05,
+      "loss": 0.3777,
+      "step": 790
+    },
+    {
+      "epoch": 0.02972170946561675,
+      "grad_norm": 0.524401843547821,
+      "learning_rate": 1.9956438200173155e-05,
+      "loss": 0.2764,
+      "step": 795
+    },
+    {
+      "epoch": 0.02990863845596654,
+      "grad_norm": 0.4294147789478302,
+      "learning_rate": 1.9955888934364985e-05,
+      "loss": 0.3334,
+      "step": 800
+    },
+    {
+      "epoch": 0.03009556744631633,
+      "grad_norm": 0.6010425686836243,
+      "learning_rate": 1.995533623505599e-05,
+      "loss": 0.3192,
+      "step": 805
+    },
+    {
+      "epoch": 0.030282496436666122,
+      "grad_norm": 0.8194842338562012,
+      "learning_rate": 1.9954780102436786e-05,
+      "loss": 0.3757,
+      "step": 810
+    },
+    {
+      "epoch": 0.030469425427015914,
+      "grad_norm": 0.7027654647827148,
+      "learning_rate": 1.995422053669917e-05,
+      "loss": 0.3285,
+      "step": 815
+    },
+    {
+      "epoch": 0.030656354417365705,
+      "grad_norm": 0.3135683238506317,
+      "learning_rate": 1.9953657538036105e-05,
+      "loss": 0.3776,
+      "step": 820
+    },
+    {
+      "epoch": 0.030843283407715492,
+      "grad_norm": 0.45088908076286316,
+      "learning_rate": 1.9953091106641772e-05,
+      "loss": 0.3567,
+      "step": 825
+    },
+    {
+      "epoch": 0.031030212398065284,
+      "grad_norm": 0.6652802228927612,
+      "learning_rate": 1.9952521242711504e-05,
+      "loss": 0.3103,
+      "step": 830
+    },
+    {
+      "epoch": 0.031217141388415075,
+      "grad_norm": 0.396379292011261,
+      "learning_rate": 1.9951947946441835e-05,
+      "loss": 0.3106,
+      "step": 835
+    },
+    {
+      "epoch": 0.031404070378764866,
+      "grad_norm": 0.5185825228691101,
+      "learning_rate": 1.9951371218030483e-05,
+      "loss": 0.3243,
+      "step": 840
+    },
+    {
+      "epoch": 0.03159099936911466,
+      "grad_norm": 0.48012250661849976,
+      "learning_rate": 1.995079105767634e-05,
+      "loss": 0.352,
+      "step": 845
+    },
+    {
+      "epoch": 0.03177792835946445,
+      "grad_norm": 0.576318621635437,
+      "learning_rate": 1.9950207465579483e-05,
+      "loss": 0.4734,
+      "step": 850
+    },
+    {
+      "epoch": 0.03196485734981424,
+      "grad_norm": 0.41435566544532776,
+      "learning_rate": 1.9949620441941183e-05,
+      "loss": 0.3589,
+      "step": 855
+    },
+    {
+      "epoch": 0.03215178634016403,
+      "grad_norm": 0.5953770875930786,
+      "learning_rate": 1.994902998696389e-05,
+      "loss": 0.3151,
+      "step": 860
+    },
+    {
+      "epoch": 0.03233871533051382,
+      "grad_norm": 0.43497735261917114,
+      "learning_rate": 1.9948436100851224e-05,
+      "loss": 0.3793,
+      "step": 865
+    },
+    {
+      "epoch": 0.03252564432086361,
+      "grad_norm": 0.8335877656936646,
+      "learning_rate": 1.9947838783808005e-05,
+      "loss": 0.362,
+      "step": 870
+    },
+    {
+      "epoch": 0.032712573311213404,
+      "grad_norm": 0.9302036762237549,
+      "learning_rate": 1.9947238036040234e-05,
+      "loss": 0.2667,
+      "step": 875
+    },
+    {
+      "epoch": 0.032899502301563195,
+      "grad_norm": 0.733805239200592,
+      "learning_rate": 1.994663385775509e-05,
+      "loss": 0.3825,
+      "step": 880
+    },
+    {
+      "epoch": 0.033086431291912986,
+      "grad_norm": 0.46970003843307495,
+      "learning_rate": 1.9946026249160935e-05,
+      "loss": 0.2545,
+      "step": 885
+    },
+    {
+      "epoch": 0.03327336028226278,
+      "grad_norm": 0.9111189246177673,
+      "learning_rate": 1.9945415210467316e-05,
+      "loss": 0.3019,
+      "step": 890
+    },
+    {
+      "epoch": 0.03346028927261257,
+      "grad_norm": 0.35354894399642944,
+      "learning_rate": 1.9944800741884963e-05,
+      "loss": 0.296,
+      "step": 895
+    },
+    {
+      "epoch": 0.03364721826296236,
+      "grad_norm": 0.5118780732154846,
+      "learning_rate": 1.9944182843625786e-05,
+      "loss": 0.3321,
+      "step": 900
+    },
+    {
+      "epoch": 0.03383414725331215,
+      "grad_norm": 0.6458944082260132,
+      "learning_rate": 1.9943561515902886e-05,
+      "loss": 0.2565,
+      "step": 905
+    },
+    {
+      "epoch": 0.03402107624366194,
+      "grad_norm": 0.34854066371917725,
+      "learning_rate": 1.9942936758930537e-05,
+      "loss": 0.3219,
+      "step": 910
+    },
+    {
+      "epoch": 0.034208005234011726,
+      "grad_norm": 0.4601725935935974,
+      "learning_rate": 1.9942308572924204e-05,
+      "loss": 0.274,
+      "step": 915
+    },
+    {
+      "epoch": 0.03439493422436152,
+      "grad_norm": 0.998417854309082,
+      "learning_rate": 1.9941676958100526e-05,
+      "loss": 0.3507,
+      "step": 920
+    },
+    {
+      "epoch": 0.03458186321471131,
+      "grad_norm": 0.7837041616439819,
+      "learning_rate": 1.9941041914677327e-05,
+      "loss": 0.3192,
+      "step": 925
+    },
+    {
+      "epoch": 0.0347687922050611,
+      "grad_norm": 0.541322648525238,
+      "learning_rate": 1.994040344287362e-05,
+      "loss": 0.3525,
+      "step": 930
+    },
+    {
+      "epoch": 0.03495572119541089,
+      "grad_norm": 0.7367837429046631,
+      "learning_rate": 1.9939761542909594e-05,
+      "loss": 0.3349,
+      "step": 935
+    },
+    {
+      "epoch": 0.03514265018576068,
+      "grad_norm": 0.6418364644050598,
+      "learning_rate": 1.9939116215006626e-05,
+      "loss": 0.3164,
+      "step": 940
+    },
+    {
+      "epoch": 0.03532957917611047,
+      "grad_norm": 0.6306796669960022,
+      "learning_rate": 1.993846745938726e-05,
+      "loss": 0.3192,
+      "step": 945
+    },
+    {
+      "epoch": 0.035516508166460264,
+      "grad_norm": 0.2718103229999542,
+      "learning_rate": 1.9937815276275247e-05,
+      "loss": 0.3019,
+      "step": 950
+    },
+    {
+      "epoch": 0.035703437156810056,
+      "grad_norm": 0.8403748869895935,
+      "learning_rate": 1.9937159665895494e-05,
+      "loss": 0.2813,
+      "step": 955
+    },
+    {
+      "epoch": 0.03589036614715985,
+      "grad_norm": 0.4756128191947937,
+      "learning_rate": 1.9936500628474115e-05,
+      "loss": 0.4187,
+      "step": 960
+    },
+    {
+      "epoch": 0.03607729513750964,
+      "grad_norm": 0.8436702489852905,
+      "learning_rate": 1.993583816423838e-05,
+      "loss": 0.3813,
+      "step": 965
+    },
+    {
+      "epoch": 0.03626422412785943,
+      "grad_norm": 0.3379393517971039,
+      "learning_rate": 1.9935172273416762e-05,
+      "loss": 0.3279,
+      "step": 970
+    },
+    {
+      "epoch": 0.03645115311820922,
+      "grad_norm": 0.5459058880805969,
+      "learning_rate": 1.9934502956238905e-05,
+      "loss": 0.3037,
+      "step": 975
+    },
+    {
+      "epoch": 0.03663808210855901,
+      "grad_norm": 0.6367349028587341,
+      "learning_rate": 1.9933830212935637e-05,
+      "loss": 0.4326,
+      "step": 980
+    },
+    {
+      "epoch": 0.0368250110989088,
+      "grad_norm": 0.35184261202812195,
+      "learning_rate": 1.993315404373897e-05,
+      "loss": 0.309,
+      "step": 985
+    },
+    {
+      "epoch": 0.037011940089258594,
+      "grad_norm": 0.49862006306648254,
+      "learning_rate": 1.9932474448882097e-05,
+      "loss": 0.3977,
+      "step": 990
+    },
+    {
+      "epoch": 0.037198869079608385,
+      "grad_norm": 0.4049328863620758,
+      "learning_rate": 1.9931791428599386e-05,
+      "loss": 0.3397,
+      "step": 995
+    },
+    {
+      "epoch": 0.037385798069958176,
+      "grad_norm": 0.4314810633659363,
+      "learning_rate": 1.993110498312639e-05,
+      "loss": 0.3494,
+      "step": 1000
+    },
+    {
+      "epoch": 0.03757272706030797,
+      "grad_norm": 0.5296122431755066,
+      "learning_rate": 1.993041511269985e-05,
+      "loss": 0.2799,
+      "step": 1005
+    },
+    {
+      "epoch": 0.03775965605065776,
+      "grad_norm": 0.506058394908905,
+      "learning_rate": 1.9929721817557682e-05,
+      "loss": 0.2591,
+      "step": 1010
+    },
+    {
+      "epoch": 0.03794658504100755,
+      "grad_norm": 0.49406224489212036,
+      "learning_rate": 1.9929025097938978e-05,
+      "loss": 0.4447,
+      "step": 1015
+    },
+    {
+      "epoch": 0.03813351403135734,
+      "grad_norm": 0.4546676576137543,
+      "learning_rate": 1.9928324954084023e-05,
+      "loss": 0.445,
+      "step": 1020
+    },
+    {
+      "epoch": 0.03832044302170713,
+      "grad_norm": 0.3749759793281555,
+      "learning_rate": 1.9927621386234274e-05,
+      "loss": 0.2486,
+      "step": 1025
+    },
+    {
+      "epoch": 0.03850737201205692,
+      "grad_norm": 1.062105655670166,
+      "learning_rate": 1.9926914394632368e-05,
+      "loss": 0.4197,
+      "step": 1030
+    },
+    {
+      "epoch": 0.038694301002406714,
+      "grad_norm": 0.4789983034133911,
+      "learning_rate": 1.992620397952213e-05,
+      "loss": 0.3374,
+      "step": 1035
+    },
+    {
+      "epoch": 0.0388812299927565,
+      "grad_norm": 0.4073324501514435,
+      "learning_rate": 1.9925490141148564e-05,
+      "loss": 0.3918,
+      "step": 1040
+    },
+    {
+      "epoch": 0.03906815898310629,
+      "grad_norm": 0.6208945512771606,
+      "learning_rate": 1.9924772879757848e-05,
+      "loss": 0.3542,
+      "step": 1045
+    },
+    {
+      "epoch": 0.03925508797345608,
+      "grad_norm": 0.4267159104347229,
+      "learning_rate": 1.9924052195597346e-05,
+      "loss": 0.3543,
+      "step": 1050
+    },
+    {
+      "epoch": 0.03944201696380587,
+      "grad_norm": 0.7129971385002136,
+      "learning_rate": 1.9923328088915603e-05,
+      "loss": 0.3433,
+      "step": 1055
+    },
+    {
+      "epoch": 0.03962894595415566,
+      "grad_norm": 0.7188255786895752,
+      "learning_rate": 1.992260055996234e-05,
+      "loss": 0.311,
+      "step": 1060
+    },
+    {
+      "epoch": 0.039815874944505454,
+      "grad_norm": 0.760067343711853,
+      "learning_rate": 1.9921869608988464e-05,
+      "loss": 0.3193,
+      "step": 1065
+    },
+    {
+      "epoch": 0.040002803934855245,
+      "grad_norm": 0.5148400664329529,
+      "learning_rate": 1.9921135236246058e-05,
+      "loss": 0.316,
+      "step": 1070
+    },
+    {
+      "epoch": 0.040189732925205036,
+      "grad_norm": 0.6598585844039917,
+      "learning_rate": 1.9920397441988384e-05,
+      "loss": 0.3337,
+      "step": 1075
+    },
+    {
+      "epoch": 0.04037666191555483,
+      "grad_norm": 0.40405112504959106,
+      "learning_rate": 1.991965622646989e-05,
+      "loss": 0.3585,
+      "step": 1080
+    },
+    {
+      "epoch": 0.04056359090590462,
+      "grad_norm": 0.6789656281471252,
+      "learning_rate": 1.9918911589946193e-05,
+      "loss": 0.3226,
+      "step": 1085
+    },
+    {
+      "epoch": 0.04075051989625441,
+      "grad_norm": 0.6832015514373779,
+      "learning_rate": 1.99181635326741e-05,
+      "loss": 0.3428,
+      "step": 1090
+    },
+    {
+      "epoch": 0.0409374488866042,
+      "grad_norm": 0.4636186361312866,
+      "learning_rate": 1.99174120549116e-05,
+      "loss": 0.4286,
+      "step": 1095
+    },
+    {
+      "epoch": 0.04112437787695399,
+      "grad_norm": 0.673987865447998,
+      "learning_rate": 1.9916657156917852e-05,
+      "loss": 0.3035,
+      "step": 1100
+    },
+    {
+      "epoch": 0.04131130686730378,
+      "grad_norm": 0.7312000393867493,
+      "learning_rate": 1.99158988389532e-05,
+      "loss": 0.3005,
+      "step": 1105
+    },
+    {
+      "epoch": 0.041498235857653575,
+      "grad_norm": 0.45022085309028625,
+      "learning_rate": 1.9915137101279163e-05,
+      "loss": 0.2549,
+      "step": 1110
+    },
+    {
+      "epoch": 0.041685164848003366,
+      "grad_norm": 0.695288360118866,
+      "learning_rate": 1.9914371944158445e-05,
+      "loss": 0.367,
+      "step": 1115
+    },
+    {
+      "epoch": 0.04187209383835316,
+      "grad_norm": 0.4332289695739746,
+      "learning_rate": 1.9913603367854927e-05,
+      "loss": 0.2729,
+      "step": 1120
+    },
+    {
+      "epoch": 0.04205902282870295,
+      "grad_norm": 0.36939793825149536,
+      "learning_rate": 1.9912831372633665e-05,
+      "loss": 0.3085,
+      "step": 1125
+    },
+    {
+      "epoch": 0.04224595181905274,
+      "grad_norm": 0.5997341871261597,
+      "learning_rate": 1.99120559587609e-05,
+      "loss": 0.3238,
+      "step": 1130
+    },
+    {
+      "epoch": 0.04243288080940253,
+      "grad_norm": 0.5867207646369934,
+      "learning_rate": 1.9911277126504056e-05,
+      "loss": 0.3768,
+      "step": 1135
+    },
+    {
+      "epoch": 0.04261980979975232,
+      "grad_norm": 0.45231306552886963,
+      "learning_rate": 1.9910494876131726e-05,
+      "loss": 0.3349,
+      "step": 1140
+    },
+    {
+      "epoch": 0.04280673879010211,
+      "grad_norm": 0.9641021490097046,
+      "learning_rate": 1.990970920791368e-05,
+      "loss": 0.3352,
+      "step": 1145
+    },
+    {
+      "epoch": 0.042993667780451904,
+      "grad_norm": 0.24853017926216125,
+      "learning_rate": 1.990892012212088e-05,
+      "loss": 0.3269,
+      "step": 1150
+    },
+    {
+      "epoch": 0.043180596770801695,
+      "grad_norm": 0.357064813375473,
+      "learning_rate": 1.9908127619025458e-05,
+      "loss": 0.2854,
+      "step": 1155
+    },
+    {
+      "epoch": 0.04336752576115148,
+      "grad_norm": 0.6567901968955994,
+      "learning_rate": 1.990733169890072e-05,
+      "loss": 0.4157,
+      "step": 1160
+    },
+    {
+      "epoch": 0.04355445475150127,
+      "grad_norm": 0.7041186690330505,
+      "learning_rate": 1.9906532362021164e-05,
+      "loss": 0.3406,
+      "step": 1165
+    },
+    {
+      "epoch": 0.04374138374185106,
+      "grad_norm": 0.7448705434799194,
+      "learning_rate": 1.990572960866245e-05,
+      "loss": 0.3272,
+      "step": 1170
+    },
+    {
+      "epoch": 0.04392831273220085,
+      "grad_norm": 0.35884958505630493,
+      "learning_rate": 1.9904923439101432e-05,
+      "loss": 0.3539,
+      "step": 1175
+    },
+    {
+      "epoch": 0.044115241722550644,
+      "grad_norm": 0.8289523124694824,
+      "learning_rate": 1.9904113853616128e-05,
+      "loss": 0.296,
+      "step": 1180
+    },
+    {
+      "epoch": 0.044302170712900435,
+      "grad_norm": 0.5754719972610474,
+      "learning_rate": 1.990330085248575e-05,
+      "loss": 0.3425,
+      "step": 1185
+    },
+    {
+      "epoch": 0.044489099703250226,
+      "grad_norm": 0.6339451670646667,
+      "learning_rate": 1.990248443599067e-05,
+      "loss": 0.4721,
+      "step": 1190
+    },
+    {
+      "epoch": 0.04467602869360002,
+      "grad_norm": 0.35451385378837585,
+      "learning_rate": 1.9901664604412453e-05,
+      "loss": 0.3476,
+      "step": 1195
+    },
+    {
+      "epoch": 0.04486295768394981,
+      "grad_norm": 0.466686487197876,
+      "learning_rate": 1.990084135803383e-05,
+      "loss": 0.3168,
+      "step": 1200
+    },
+    {
+      "epoch": 0.0450498866742996,
+      "grad_norm": 0.3746946454048157,
+      "learning_rate": 1.9900014697138718e-05,
+      "loss": 0.2576,
+      "step": 1205
+    },
+    {
+      "epoch": 0.04523681566464939,
+      "grad_norm": 0.4259583353996277,
+      "learning_rate": 1.9899184622012208e-05,
+      "loss": 0.3146,
+      "step": 1210
+    },
+    {
+      "epoch": 0.04542374465499918,
+      "grad_norm": 0.6210399270057678,
+      "learning_rate": 1.989835113294057e-05,
+      "loss": 0.3105,
+      "step": 1215
+    },
+    {
+      "epoch": 0.04561067364534897,
+      "grad_norm": 0.45747750997543335,
+      "learning_rate": 1.989751423021125e-05,
+      "loss": 0.2911,
+      "step": 1220
+    },
+    {
+      "epoch": 0.045797602635698764,
+      "grad_norm": 0.4624209702014923,
+      "learning_rate": 1.989667391411287e-05,
+      "loss": 0.3643,
+      "step": 1225
+    },
+    {
+      "epoch": 0.045984531626048555,
+      "grad_norm": 0.7682859301567078,
+      "learning_rate": 1.9895830184935233e-05,
+      "loss": 0.3374,
+      "step": 1230
+    },
+    {
+      "epoch": 0.04617146061639835,
+      "grad_norm": 0.6099391579627991,
+      "learning_rate": 1.989498304296932e-05,
+      "loss": 0.299,
+      "step": 1235
+    },
+    {
+      "epoch": 0.04635838960674814,
+      "grad_norm": 0.5577886700630188,
+      "learning_rate": 1.989413248850728e-05,
+      "loss": 0.3821,
+      "step": 1240
+    },
+    {
+      "epoch": 0.04654531859709793,
+      "grad_norm": 0.4436083436012268,
+      "learning_rate": 1.9893278521842448e-05,
+      "loss": 0.358,
+      "step": 1245
+    },
+    {
+      "epoch": 0.04673224758744772,
+      "grad_norm": 0.6322022676467896,
+      "learning_rate": 1.989242114326933e-05,
+      "loss": 0.2856,
+      "step": 1250
+    },
+    {
+      "epoch": 0.04691917657779751,
+      "grad_norm": 0.4859846234321594,
+      "learning_rate": 1.9891560353083616e-05,
+      "loss": 0.2951,
+      "step": 1255
+    },
+    {
+      "epoch": 0.0471061055681473,
+      "grad_norm": 0.4296378791332245,
+      "learning_rate": 1.989069615158217e-05,
+      "loss": 0.3627,
+      "step": 1260
+    },
+    {
+      "epoch": 0.047293034558497093,
+      "grad_norm": 0.4696434438228607,
+      "learning_rate": 1.9889828539063017e-05,
+      "loss": 0.3404,
+      "step": 1265
+    },
+    {
+      "epoch": 0.047479963548846885,
+      "grad_norm": 0.5244821310043335,
+      "learning_rate": 1.9888957515825383e-05,
+      "loss": 0.3347,
+      "step": 1270
+    },
+    {
+      "epoch": 0.047666892539196676,
+      "grad_norm": 0.5117040872573853,
+      "learning_rate": 1.9888083082169657e-05,
+      "loss": 0.3017,
+      "step": 1275
+    },
+    {
+      "epoch": 0.04785382152954647,
+      "grad_norm": 0.6786810159683228,
+      "learning_rate": 1.9887205238397405e-05,
+      "loss": 0.3495,
+      "step": 1280
+    },
+    {
+      "epoch": 0.04804075051989625,
+      "grad_norm": 0.41801556944847107,
+      "learning_rate": 1.988632398481137e-05,
+      "loss": 0.2713,
+      "step": 1285
+    },
+    {
+      "epoch": 0.04822767951024604,
+      "grad_norm": 0.3679557740688324,
+      "learning_rate": 1.988543932171547e-05,
+      "loss": 0.2617,
+      "step": 1290
+    },
+    {
+      "epoch": 0.048414608500595833,
+      "grad_norm": 0.3665943145751953,
+      "learning_rate": 1.9884551249414806e-05,
+      "loss": 0.3392,
+      "step": 1295
+    },
+    {
+      "epoch": 0.048601537490945625,
+      "grad_norm": 0.6276513934135437,
+      "learning_rate": 1.9883659768215642e-05,
+      "loss": 0.3797,
+      "step": 1300
+    },
+    {
+      "epoch": 0.048788466481295416,
+      "grad_norm": 0.3099709451198578,
+      "learning_rate": 1.988276487842543e-05,
+      "loss": 0.384,
+      "step": 1305
+    },
+    {
+      "epoch": 0.04897539547164521,
+      "grad_norm": 0.3573191463947296,
+      "learning_rate": 1.9881866580352783e-05,
+      "loss": 0.3194,
+      "step": 1310
+    },
+    {
+      "epoch": 0.049162324461995,
+      "grad_norm": 0.4190731644630432,
+      "learning_rate": 1.9880964874307506e-05,
+      "loss": 0.284,
+      "step": 1315
+    },
+    {
+      "epoch": 0.04934925345234479,
+      "grad_norm": 0.4039645791053772,
+      "learning_rate": 1.988005976060057e-05,
+      "loss": 0.2548,
+      "step": 1320
+    },
+    {
+      "epoch": 0.04953618244269458,
+      "grad_norm": 0.4496673345565796,
+      "learning_rate": 1.987915123954412e-05,
+      "loss": 0.3529,
+      "step": 1325
+    },
+    {
+      "epoch": 0.04972311143304437,
+      "grad_norm": 0.8639070987701416,
+      "learning_rate": 1.9878239311451483e-05,
+      "loss": 0.3073,
+      "step": 1330
+    },
+    {
+      "epoch": 0.04991004042339416,
+      "grad_norm": 0.5019544959068298,
+      "learning_rate": 1.9877323976637153e-05,
+      "loss": 0.3829,
+      "step": 1335
+    },
+    {
+      "epoch": 0.050096969413743954,
+      "grad_norm": 0.5682312846183777,
+      "learning_rate": 1.9876405235416808e-05,
+      "loss": 0.4139,
+      "step": 1340
+    },
+    {
+      "epoch": 0.050283898404093745,
+      "grad_norm": 0.671947717666626,
+      "learning_rate": 1.987548308810729e-05,
+      "loss": 0.2905,
+      "step": 1345
+    },
+    {
+      "epoch": 0.050470827394443536,
+      "grad_norm": 0.2840399444103241,
+      "learning_rate": 1.9874557535026623e-05,
+      "loss": 0.3269,
+      "step": 1350
+    },
+    {
+      "epoch": 0.05065775638479333,
+      "grad_norm": 0.5511133074760437,
+      "learning_rate": 1.9873628576494004e-05,
+      "loss": 0.3221,
+      "step": 1355
+    },
+    {
+      "epoch": 0.05084468537514312,
+      "grad_norm": 0.4360320270061493,
+      "learning_rate": 1.9872696212829804e-05,
+      "loss": 0.3094,
+      "step": 1360
+    },
+    {
+      "epoch": 0.05103161436549291,
+      "grad_norm": 0.342729777097702,
+      "learning_rate": 1.987176044435557e-05,
+      "loss": 0.3401,
+      "step": 1365
+    },
+    {
+      "epoch": 0.0512185433558427,
+      "grad_norm": 0.43248656392097473,
+      "learning_rate": 1.987082127139402e-05,
+      "loss": 0.2907,
+      "step": 1370
+    },
+    {
+      "epoch": 0.05140547234619249,
+      "grad_norm": 0.6353283524513245,
+      "learning_rate": 1.9869878694269048e-05,
+      "loss": 0.3599,
+      "step": 1375
+    },
+    {
+      "epoch": 0.05159240133654228,
+      "grad_norm": 0.5560038685798645,
+      "learning_rate": 1.9868932713305723e-05,
+      "loss": 0.3177,
+      "step": 1380
+    },
+    {
+      "epoch": 0.051779330326892074,
+      "grad_norm": 0.49626341462135315,
+      "learning_rate": 1.9867983328830283e-05,
+      "loss": 0.3419,
+      "step": 1385
+    },
+    {
+      "epoch": 0.051966259317241865,
+      "grad_norm": 0.5227237939834595,
+      "learning_rate": 1.986703054117015e-05,
+      "loss": 0.2357,
+      "step": 1390
+    },
+    {
+      "epoch": 0.05215318830759166,
+      "grad_norm": 0.45654603838920593,
+      "learning_rate": 1.986607435065391e-05,
+      "loss": 0.2313,
+      "step": 1395
+    },
+    {
+      "epoch": 0.05234011729794145,
+      "grad_norm": 0.4107493460178375,
+      "learning_rate": 1.9865114757611322e-05,
+      "loss": 0.3056,
+      "step": 1400
+    },
+    {
+      "epoch": 0.05252704628829123,
+      "grad_norm": 0.9286690354347229,
+      "learning_rate": 1.9864151762373323e-05,
+      "loss": 0.3116,
+      "step": 1405
+    },
+    {
+      "epoch": 0.05271397527864102,
+      "grad_norm": 0.5546656847000122,
+      "learning_rate": 1.986318536527203e-05,
+      "loss": 0.3392,
+      "step": 1410
+    },
+    {
+      "epoch": 0.052900904268990814,
+      "grad_norm": 0.45263174176216125,
+      "learning_rate": 1.986221556664072e-05,
+      "loss": 0.3175,
+      "step": 1415
+    },
+    {
+      "epoch": 0.053087833259340605,
+      "grad_norm": 0.2992209792137146,
+      "learning_rate": 1.9861242366813846e-05,
+      "loss": 0.2616,
+      "step": 1420
+    },
+    {
+      "epoch": 0.0532747622496904,
+      "grad_norm": 0.39666202664375305,
+      "learning_rate": 1.9860265766127045e-05,
+      "loss": 0.3728,
+      "step": 1425
+    },
+    {
+      "epoch": 0.05346169124004019,
+      "grad_norm": 0.616473376750946,
+      "learning_rate": 1.9859285764917108e-05,
+      "loss": 0.3823,
+      "step": 1430
+    },
+    {
+      "epoch": 0.05364862023038998,
+      "grad_norm": 0.3824799954891205,
+      "learning_rate": 1.985830236352202e-05,
+      "loss": 0.2945,
+      "step": 1435
+    },
+    {
+      "epoch": 0.05383554922073977,
+      "grad_norm": 0.45680978894233704,
+      "learning_rate": 1.9857315562280923e-05,
+      "loss": 0.3025,
+      "step": 1440
+    },
+    {
+      "epoch": 0.05402247821108956,
+      "grad_norm": 0.6660309433937073,
+      "learning_rate": 1.9856325361534133e-05,
+      "loss": 0.2585,
+      "step": 1445
+    },
+    {
+      "epoch": 0.05420940720143935,
+      "grad_norm": 0.6838333010673523,
+      "learning_rate": 1.9855331761623143e-05,
+      "loss": 0.2684,
+      "step": 1450
+    },
+    {
+      "epoch": 0.054396336191789144,
+      "grad_norm": 0.36576396226882935,
+      "learning_rate": 1.9854334762890626e-05,
+      "loss": 0.3123,
+      "step": 1455
+    },
+    {
+      "epoch": 0.054583265182138935,
+      "grad_norm": 0.4945361614227295,
+      "learning_rate": 1.9853334365680408e-05,
+      "loss": 0.3113,
+      "step": 1460
+    },
+    {
+      "epoch": 0.054770194172488726,
+      "grad_norm": 0.4163508117198944,
+      "learning_rate": 1.98523305703375e-05,
+      "loss": 0.3394,
+      "step": 1465
+    },
+    {
+      "epoch": 0.05495712316283852,
+      "grad_norm": 0.42047226428985596,
+      "learning_rate": 1.985132337720809e-05,
+      "loss": 0.2486,
+      "step": 1470
+    },
+    {
+      "epoch": 0.05514405215318831,
+      "grad_norm": 0.4733279347419739,
+      "learning_rate": 1.9850312786639513e-05,
+      "loss": 0.3164,
+      "step": 1475
+    },
+    {
+      "epoch": 0.0553309811435381,
+      "grad_norm": 0.36004531383514404,
+      "learning_rate": 1.984929879898031e-05,
+      "loss": 0.2725,
+      "step": 1480
+    },
+    {
+      "epoch": 0.05551791013388789,
+      "grad_norm": 0.32165494561195374,
+      "learning_rate": 1.9848281414580167e-05,
+      "loss": 0.3624,
+      "step": 1485
+    },
+    {
+      "epoch": 0.05570483912423768,
+      "grad_norm": 0.7197104096412659,
+      "learning_rate": 1.9847260633789953e-05,
+      "loss": 0.3337,
+      "step": 1490
+    },
+    {
+      "epoch": 0.05589176811458747,
+      "grad_norm": 0.4705203175544739,
+      "learning_rate": 1.984623645696171e-05,
+      "loss": 0.3503,
+      "step": 1495
+    },
+    {
+      "epoch": 0.056078697104937264,
+      "grad_norm": 0.6553797125816345,
+      "learning_rate": 1.984520888444864e-05,
+      "loss": 0.4203,
+      "step": 1500
+    },
+    {
+      "epoch": 0.056265626095287055,
+      "grad_norm": 0.46082812547683716,
+      "learning_rate": 1.9844177916605126e-05,
+      "loss": 0.299,
+      "step": 1505
+    },
+    {
+      "epoch": 0.056452555085636846,
+      "grad_norm": 0.6781812906265259,
+      "learning_rate": 1.984314355378672e-05,
+      "loss": 0.3063,
+      "step": 1510
+    },
+    {
+      "epoch": 0.05663948407598664,
+      "grad_norm": 0.6019364595413208,
+      "learning_rate": 1.9842105796350143e-05,
+      "loss": 0.389,
+      "step": 1515
+    },
+    {
+      "epoch": 0.05682641306633643,
+      "grad_norm": 0.4727480113506317,
+      "learning_rate": 1.9841064644653293e-05,
+      "loss": 0.367,
+      "step": 1520
+    },
+    {
+      "epoch": 0.05701334205668621,
+      "grad_norm": 0.5794846415519714,
+      "learning_rate": 1.9840020099055226e-05,
+      "loss": 0.3148,
+      "step": 1525
+    },
+    {
+      "epoch": 0.057200271047036004,
+      "grad_norm": 0.48816463351249695,
+      "learning_rate": 1.983897215991618e-05,
+      "loss": 0.3408,
+      "step": 1530
+    },
+    {
+      "epoch": 0.057387200037385795,
+      "grad_norm": 0.2970341145992279,
+      "learning_rate": 1.983792082759756e-05,
+      "loss": 0.3218,
+      "step": 1535
+    },
+    {
+      "epoch": 0.057574129027735586,
+      "grad_norm": 0.2880699336528778,
+      "learning_rate": 1.9836866102461933e-05,
+      "loss": 0.3219,
+      "step": 1540
+    },
+    {
+      "epoch": 0.05776105801808538,
+      "grad_norm": 0.368488609790802,
+      "learning_rate": 1.9835807984873055e-05,
+      "loss": 0.3782,
+      "step": 1545
+    },
+    {
+      "epoch": 0.05794798700843517,
+      "grad_norm": 0.5185800194740295,
+      "learning_rate": 1.983474647519583e-05,
+      "loss": 0.2973,
+      "step": 1550
+    },
+    {
+      "epoch": 0.05813491599878496,
+      "grad_norm": 0.5686115026473999,
+      "learning_rate": 1.9833681573796352e-05,
+      "loss": 0.317,
+      "step": 1555
+    },
+    {
+      "epoch": 0.05832184498913475,
+      "grad_norm": 0.6750605702400208,
+      "learning_rate": 1.983261328104187e-05,
+      "loss": 0.3243,
+      "step": 1560
+    },
+    {
+      "epoch": 0.05850877397948454,
+      "grad_norm": 0.36497625708580017,
+      "learning_rate": 1.9831541597300804e-05,
+      "loss": 0.3339,
+      "step": 1565
+    },
+    {
+      "epoch": 0.05869570296983433,
+      "grad_norm": 0.5015937089920044,
+      "learning_rate": 1.983046652294275e-05,
+      "loss": 0.3349,
+      "step": 1570
+    },
+    {
+      "epoch": 0.058882631960184124,
+      "grad_norm": 0.36055922508239746,
+      "learning_rate": 1.982938805833847e-05,
+      "loss": 0.2826,
+      "step": 1575
+    },
+    {
+      "epoch": 0.059069560950533916,
+      "grad_norm": 0.36943939328193665,
+      "learning_rate": 1.9828306203859896e-05,
+      "loss": 0.3517,
+      "step": 1580
+    },
+    {
+      "epoch": 0.05925648994088371,
+      "grad_norm": 0.47426244616508484,
+      "learning_rate": 1.982722095988013e-05,
+      "loss": 0.3198,
+      "step": 1585
+    },
+    {
+      "epoch": 0.0594434189312335,
+      "grad_norm": 0.5041938424110413,
+      "learning_rate": 1.9826132326773443e-05,
+      "loss": 0.3636,
+      "step": 1590
+    },
+    {
+      "epoch": 0.05963034792158329,
+      "grad_norm": 0.5388785600662231,
+      "learning_rate": 1.982504030491527e-05,
+      "loss": 0.3527,
+      "step": 1595
+    },
+    {
+      "epoch": 0.05981727691193308,
+      "grad_norm": 0.399564653635025,
+      "learning_rate": 1.982394489468222e-05,
+      "loss": 0.3333,
+      "step": 1600
+    },
+    {
+      "epoch": 0.06000420590228287,
+      "grad_norm": 1.5852733850479126,
+      "learning_rate": 1.9822846096452064e-05,
+      "loss": 0.3851,
+      "step": 1605
+    },
+    {
+      "epoch": 0.06019113489263266,
+      "grad_norm": 0.3177071809768677,
+      "learning_rate": 1.982174391060375e-05,
+      "loss": 0.3981,
+      "step": 1610
+    },
+    {
+      "epoch": 0.060378063882982454,
+      "grad_norm": 0.9603758454322815,
+      "learning_rate": 1.982063833751739e-05,
+      "loss": 0.3691,
+      "step": 1615
+    },
+    {
+      "epoch": 0.060564992873332245,
+      "grad_norm": 0.3188650608062744,
+      "learning_rate": 1.9819529377574265e-05,
+      "loss": 0.3143,
+      "step": 1620
+    },
+    {
+      "epoch": 0.060751921863682036,
+      "grad_norm": 0.48925361037254333,
+      "learning_rate": 1.9818417031156826e-05,
+      "loss": 0.3063,
+      "step": 1625
+    },
+    {
+      "epoch": 0.06093885085403183,
+      "grad_norm": 0.4794778823852539,
+      "learning_rate": 1.9817301298648683e-05,
+      "loss": 0.3278,
+      "step": 1630
+    },
+    {
+      "epoch": 0.06112577984438162,
+      "grad_norm": 0.4850408434867859,
+      "learning_rate": 1.9816182180434622e-05,
+      "loss": 0.3245,
+      "step": 1635
+    },
+    {
+      "epoch": 0.06131270883473141,
+      "grad_norm": 0.5055322647094727,
+      "learning_rate": 1.9815059676900597e-05,
+      "loss": 0.2969,
+      "step": 1640
+    },
+    {
+      "epoch": 0.0614996378250812,
+      "grad_norm": 0.8014519214630127,
+      "learning_rate": 1.9813933788433724e-05,
+      "loss": 0.2473,
+      "step": 1645
+    },
+    {
+      "epoch": 0.061686566815430985,
+      "grad_norm": 0.3348430395126343,
+      "learning_rate": 1.9812804515422298e-05,
+      "loss": 0.3928,
+      "step": 1650
+    },
+    {
+      "epoch": 0.061873495805780776,
+      "grad_norm": 0.6837722659111023,
+      "learning_rate": 1.9811671858255764e-05,
+      "loss": 0.2898,
+      "step": 1655
+    },
+    {
+      "epoch": 0.06206042479613057,
+      "grad_norm": 0.422375351190567,
+      "learning_rate": 1.9810535817324746e-05,
+      "loss": 0.3522,
+      "step": 1660
+    },
+    {
+      "epoch": 0.06224735378648036,
+      "grad_norm": 0.8504202365875244,
+      "learning_rate": 1.980939639302103e-05,
+      "loss": 0.4159,
+      "step": 1665
+    },
+    {
+      "epoch": 0.06243428277683015,
+      "grad_norm": 0.31550297141075134,
+      "learning_rate": 1.9808253585737577e-05,
+      "loss": 0.3423,
+      "step": 1670
+    },
+    {
+      "epoch": 0.06262121176717994,
+      "grad_norm": 0.52349454164505,
+      "learning_rate": 1.9807107395868503e-05,
+      "loss": 0.337,
+      "step": 1675
+    },
+    {
+      "epoch": 0.06280814075752973,
+      "grad_norm": 0.9710644483566284,
+      "learning_rate": 1.9805957823809095e-05,
+      "loss": 0.3331,
+      "step": 1680
+    },
+    {
+      "epoch": 0.06299506974787952,
+      "grad_norm": 0.5250943899154663,
+      "learning_rate": 1.9804804869955815e-05,
+      "loss": 0.3059,
+      "step": 1685
+    },
+    {
+      "epoch": 0.06318199873822931,
+      "grad_norm": 0.6633884310722351,
+      "learning_rate": 1.980364853470627e-05,
+      "loss": 0.3314,
+      "step": 1690
+    },
+    {
+      "epoch": 0.0633689277285791,
+      "grad_norm": 0.35779082775115967,
+      "learning_rate": 1.9802488818459263e-05,
+      "loss": 0.2959,
+      "step": 1695
+    },
+    {
+      "epoch": 0.0635558567189289,
+      "grad_norm": 0.3499528169631958,
+      "learning_rate": 1.980132572161474e-05,
+      "loss": 0.3142,
+      "step": 1700
+    },
+    {
+      "epoch": 0.06374278570927869,
+      "grad_norm": 0.45091912150382996,
+      "learning_rate": 1.9800159244573813e-05,
+      "loss": 0.342,
+      "step": 1705
+    },
+    {
+      "epoch": 0.06392971469962848,
+      "grad_norm": 0.46202903985977173,
+      "learning_rate": 1.9798989387738776e-05,
+      "loss": 0.3052,
+      "step": 1710
+    },
+    {
+      "epoch": 0.06411664368997827,
+      "grad_norm": 0.4897025227546692,
+      "learning_rate": 1.9797816151513075e-05,
+      "loss": 0.3864,
+      "step": 1715
+    },
+    {
+      "epoch": 0.06430357268032806,
+      "grad_norm": 0.44095808267593384,
+      "learning_rate": 1.979663953630133e-05,
+      "loss": 0.3165,
+      "step": 1720
+    },
+    {
+      "epoch": 0.06449050167067785,
+      "grad_norm": 0.4061107337474823,
+      "learning_rate": 1.9795459542509314e-05,
+      "loss": 0.3123,
+      "step": 1725
+    },
+    {
+      "epoch": 0.06467743066102764,
+      "grad_norm": 0.37612634897232056,
+      "learning_rate": 1.9794276170543974e-05,
+      "loss": 0.3726,
+      "step": 1730
+    },
+    {
+      "epoch": 0.06486435965137743,
+      "grad_norm": 0.8302399516105652,
+      "learning_rate": 1.979308942081343e-05,
+      "loss": 0.3406,
+      "step": 1735
+    },
+    {
+      "epoch": 0.06505128864172723,
+      "grad_norm": 0.6120762228965759,
+      "learning_rate": 1.9791899293726947e-05,
+      "loss": 0.3493,
+      "step": 1740
+    },
+    {
+      "epoch": 0.06523821763207702,
+      "grad_norm": 1.227718710899353,
+      "learning_rate": 1.9790705789694977e-05,
+      "loss": 0.3243,
+      "step": 1745
+    },
+    {
+      "epoch": 0.06542514662242681,
+      "grad_norm": 0.2927764356136322,
+      "learning_rate": 1.978950890912912e-05,
+      "loss": 0.2685,
+      "step": 1750
+    },
+    {
+      "epoch": 0.0656120756127766,
+      "grad_norm": 0.18635421991348267,
+      "learning_rate": 1.9788308652442137e-05,
+      "loss": 0.326,
+      "step": 1755
+    },
+    {
+      "epoch": 0.06579900460312639,
+      "grad_norm": 0.5031612515449524,
+      "learning_rate": 1.978710502004798e-05,
+      "loss": 0.2777,
+      "step": 1760
+    },
+    {
+      "epoch": 0.06598593359347618,
+      "grad_norm": 0.5212199091911316,
+      "learning_rate": 1.9785898012361732e-05,
+      "loss": 0.3813,
+      "step": 1765
+    },
+    {
+      "epoch": 0.06617286258382597,
+      "grad_norm": 0.40999093651771545,
+      "learning_rate": 1.978468762979966e-05,
+      "loss": 0.3218,
+      "step": 1770
+    },
+    {
+      "epoch": 0.06635979157417576,
+      "grad_norm": 0.29708200693130493,
+      "learning_rate": 1.9783473872779192e-05,
+      "loss": 0.2983,
+      "step": 1775
+    },
+    {
+      "epoch": 0.06654672056452555,
+      "grad_norm": 0.48487138748168945,
+      "learning_rate": 1.978225674171892e-05,
+      "loss": 0.2956,
+      "step": 1780
+    },
+    {
+      "epoch": 0.06673364955487535,
+      "grad_norm": 0.7466865181922913,
+      "learning_rate": 1.9781036237038593e-05,
+      "loss": 0.4038,
+      "step": 1785
+    },
+    {
+      "epoch": 0.06692057854522514,
+      "grad_norm": 0.5608679056167603,
+      "learning_rate": 1.977981235915913e-05,
+      "loss": 0.31,
+      "step": 1790
+    },
+    {
+      "epoch": 0.06710750753557493,
+      "grad_norm": 0.3457610011100769,
+      "learning_rate": 1.9778585108502613e-05,
+      "loss": 0.3171,
+      "step": 1795
+    },
+    {
+      "epoch": 0.06729443652592472,
+      "grad_norm": 0.3912654221057892,
+      "learning_rate": 1.977735448549228e-05,
+      "loss": 0.2944,
+      "step": 1800
+    },
+    {
+      "epoch": 0.06748136551627451,
+      "grad_norm": 0.660007119178772,
+      "learning_rate": 1.9776120490552545e-05,
+      "loss": 0.3354,
+      "step": 1805
+    },
+    {
+      "epoch": 0.0676682945066243,
+      "grad_norm": 0.5889789462089539,
+      "learning_rate": 1.9774883124108975e-05,
+      "loss": 0.2601,
+      "step": 1810
+    },
+    {
+      "epoch": 0.06785522349697409,
+      "grad_norm": 0.43226462602615356,
+      "learning_rate": 1.97736423865883e-05,
+      "loss": 0.26,
+      "step": 1815
+    },
+    {
+      "epoch": 0.06804215248732388,
+      "grad_norm": 0.38729366660118103,
+      "learning_rate": 1.9772398278418414e-05,
+      "loss": 0.2675,
+      "step": 1820
+    },
+    {
+      "epoch": 0.06822908147767368,
+      "grad_norm": 0.472239226102829,
+      "learning_rate": 1.9771150800028376e-05,
+      "loss": 0.3722,
+      "step": 1825
+    },
+    {
+      "epoch": 0.06841601046802345,
+      "grad_norm": 0.5041793584823608,
+      "learning_rate": 1.976989995184841e-05,
+      "loss": 0.2839,
+      "step": 1830
+    },
+    {
+      "epoch": 0.06860293945837324,
+      "grad_norm": 0.5347047448158264,
+      "learning_rate": 1.9768645734309896e-05,
+      "loss": 0.365,
+      "step": 1835
+    },
+    {
+      "epoch": 0.06878986844872303,
+      "grad_norm": 0.3948882818222046,
+      "learning_rate": 1.976738814784537e-05,
+      "loss": 0.3337,
+      "step": 1840
+    },
+    {
+      "epoch": 0.06897679743907283,
+      "grad_norm": 0.8375716209411621,
+      "learning_rate": 1.9766127192888543e-05,
+      "loss": 0.2908,
+      "step": 1845
+    },
+    {
+      "epoch": 0.06916372642942262,
+      "grad_norm": 0.6502942442893982,
+      "learning_rate": 1.9764862869874282e-05,
+      "loss": 0.2642,
+      "step": 1850
+    },
+    {
+      "epoch": 0.06935065541977241,
+      "grad_norm": 0.36091551184654236,
+      "learning_rate": 1.9763595179238617e-05,
+      "loss": 0.3685,
+      "step": 1855
+    },
+    {
+      "epoch": 0.0695375844101222,
+      "grad_norm": 0.6670817732810974,
+      "learning_rate": 1.9762324121418735e-05,
+      "loss": 0.2686,
+      "step": 1860
+    },
+    {
+      "epoch": 0.06972451340047199,
+      "grad_norm": 0.50603187084198,
+      "learning_rate": 1.9761049696852996e-05,
+      "loss": 0.4413,
+      "step": 1865
+    },
+    {
+      "epoch": 0.06991144239082178,
+      "grad_norm": 0.4863389730453491,
+      "learning_rate": 1.97597719059809e-05,
+      "loss": 0.216,
+      "step": 1870
+    },
+    {
+      "epoch": 0.07009837138117157,
+      "grad_norm": 0.3261346220970154,
+      "learning_rate": 1.975849074924313e-05,
+      "loss": 0.3634,
+      "step": 1875
+    },
+    {
+      "epoch": 0.07028530037152136,
+      "grad_norm": 0.44674980640411377,
+      "learning_rate": 1.9757206227081514e-05,
+      "loss": 0.268,
+      "step": 1880
+    },
+    {
+      "epoch": 0.07047222936187116,
+      "grad_norm": 0.4099423587322235,
+      "learning_rate": 1.975591833993905e-05,
+      "loss": 0.3505,
+      "step": 1885
+    },
+    {
+      "epoch": 0.07065915835222095,
+      "grad_norm": 1.4500172138214111,
+      "learning_rate": 1.9754627088259894e-05,
+      "loss": 0.3023,
+      "step": 1890
+    },
+    {
+      "epoch": 0.07084608734257074,
+      "grad_norm": 0.49058449268341064,
+      "learning_rate": 1.975333247248936e-05,
+      "loss": 0.2925,
+      "step": 1895
+    },
+    {
+      "epoch": 0.07103301633292053,
+      "grad_norm": 0.4941057860851288,
+      "learning_rate": 1.9752034493073924e-05,
+      "loss": 0.2957,
+      "step": 1900
+    },
+    {
+      "epoch": 0.07121994532327032,
+      "grad_norm": 0.8845500349998474,
+      "learning_rate": 1.9750733150461225e-05,
+      "loss": 0.3244,
+      "step": 1905
+    },
+    {
+      "epoch": 0.07140687431362011,
+      "grad_norm": 1.2214986085891724,
+      "learning_rate": 1.9749428445100053e-05,
+      "loss": 0.3882,
+      "step": 1910
+    },
+    {
+      "epoch": 0.0715938033039699,
+      "grad_norm": 0.6646692752838135,
+      "learning_rate": 1.974812037744037e-05,
+      "loss": 0.4436,
+      "step": 1915
+    },
+    {
+      "epoch": 0.0717807322943197,
+      "grad_norm": 0.5141096115112305,
+      "learning_rate": 1.9746808947933285e-05,
+      "loss": 0.2812,
+      "step": 1920
+    },
+    {
+      "epoch": 0.07196766128466948,
+      "grad_norm": 0.39117810130119324,
+      "learning_rate": 1.9745494157031075e-05,
+      "loss": 0.2966,
+      "step": 1925
+    },
+    {
+      "epoch": 0.07215459027501928,
+      "grad_norm": 0.5144658088684082,
+      "learning_rate": 1.9744176005187173e-05,
+      "loss": 0.3422,
+      "step": 1930
+    },
+    {
+      "epoch": 0.07234151926536907,
+      "grad_norm": 0.6618564128875732,
+      "learning_rate": 1.9742854492856178e-05,
+      "loss": 0.3295,
+      "step": 1935
+    },
+    {
+      "epoch": 0.07252844825571886,
+      "grad_norm": 0.7013601660728455,
+      "learning_rate": 1.9741529620493833e-05,
+      "loss": 0.3461,
+      "step": 1940
+    },
+    {
+      "epoch": 0.07271537724606865,
+      "grad_norm": 0.5347566604614258,
+      "learning_rate": 1.9740201388557053e-05,
+      "loss": 0.278,
+      "step": 1945
+    },
+    {
+      "epoch": 0.07290230623641844,
+      "grad_norm": 0.4391633868217468,
+      "learning_rate": 1.9738869797503905e-05,
+      "loss": 0.2967,
+      "step": 1950
+    },
+    {
+      "epoch": 0.07308923522676823,
+      "grad_norm": 0.4526468813419342,
+      "learning_rate": 1.9737534847793622e-05,
+      "loss": 0.2962,
+      "step": 1955
+    },
+    {
+      "epoch": 0.07327616421711802,
+      "grad_norm": 0.652113676071167,
+      "learning_rate": 1.9736196539886578e-05,
+      "loss": 0.2854,
+      "step": 1960
+    },
+    {
+      "epoch": 0.07346309320746781,
+      "grad_norm": 0.39846867322921753,
+      "learning_rate": 1.9734854874244332e-05,
+      "loss": 0.2783,
+      "step": 1965
+    },
+    {
+      "epoch": 0.0736500221978176,
+      "grad_norm": 0.3684355318546295,
+      "learning_rate": 1.9733509851329574e-05,
+      "loss": 0.2901,
+      "step": 1970
+    },
+    {
+      "epoch": 0.0738369511881674,
+      "grad_norm": 0.6221095323562622,
+      "learning_rate": 1.973216147160617e-05,
+      "loss": 0.3012,
+      "step": 1975
+    },
+    {
+      "epoch": 0.07402388017851719,
+      "grad_norm": 0.41872674226760864,
+      "learning_rate": 1.9730809735539134e-05,
+      "loss": 0.3818,
+      "step": 1980
+    },
+    {
+      "epoch": 0.07421080916886698,
+      "grad_norm": 0.35373455286026,
+      "learning_rate": 1.9729454643594646e-05,
+      "loss": 0.354,
+      "step": 1985
+    },
+    {
+      "epoch": 0.07439773815921677,
+      "grad_norm": 0.795148491859436,
+      "learning_rate": 1.9728096196240035e-05,
+      "loss": 0.3374,
+      "step": 1990
+    },
+    {
+      "epoch": 0.07458466714956656,
+      "grad_norm": 0.32079294323921204,
+      "learning_rate": 1.972673439394379e-05,
+      "loss": 0.2758,
+      "step": 1995
+    },
+    {
+      "epoch": 0.07477159613991635,
+      "grad_norm": 0.3288242518901825,
+      "learning_rate": 1.9725369237175562e-05,
+      "loss": 0.2796,
+      "step": 2000
+    },
+    {
+      "epoch": 0.07495852513026614,
+      "grad_norm": 0.428270548582077,
+      "learning_rate": 1.9724000726406148e-05,
+      "loss": 0.2982,
+      "step": 2005
+    },
+    {
+      "epoch": 0.07514545412061593,
+      "grad_norm": 0.38480132818222046,
+      "learning_rate": 1.9722628862107514e-05,
+      "loss": 0.2905,
+      "step": 2010
+    },
+    {
+      "epoch": 0.07533238311096573,
+      "grad_norm": 0.5246176719665527,
+      "learning_rate": 1.9721253644752774e-05,
+      "loss": 0.3551,
+      "step": 2015
+    },
+    {
+      "epoch": 0.07551931210131552,
+      "grad_norm": 0.5422767400741577,
+      "learning_rate": 1.97198750748162e-05,
+      "loss": 0.3718,
+      "step": 2020
+    },
+    {
+      "epoch": 0.07570624109166531,
+      "grad_norm": 0.5160179138183594,
+      "learning_rate": 1.971849315277322e-05,
+      "loss": 0.3526,
+      "step": 2025
+    },
+    {
+      "epoch": 0.0758931700820151,
+      "grad_norm": 0.5089318752288818,
+      "learning_rate": 1.9717107879100426e-05,
+      "loss": 0.344,
+      "step": 2030
+    },
+    {
+      "epoch": 0.07608009907236489,
+      "grad_norm": 0.6274924874305725,
+      "learning_rate": 1.971571925427555e-05,
+      "loss": 0.2923,
+      "step": 2035
+    },
+    {
+      "epoch": 0.07626702806271468,
+      "grad_norm": 0.5537222027778625,
+      "learning_rate": 1.9714327278777495e-05,
+      "loss": 0.3163,
+      "step": 2040
+    },
+    {
+      "epoch": 0.07645395705306447,
+      "grad_norm": 0.7514776587486267,
+      "learning_rate": 1.9712931953086314e-05,
+      "loss": 0.4283,
+      "step": 2045
+    },
+    {
+      "epoch": 0.07664088604341426,
+      "grad_norm": 0.33425864577293396,
+      "learning_rate": 1.9711533277683214e-05,
+      "loss": 0.2521,
+      "step": 2050
+    },
+    {
+      "epoch": 0.07682781503376405,
+      "grad_norm": 0.4531971514225006,
+      "learning_rate": 1.9710131253050555e-05,
+      "loss": 0.3491,
+      "step": 2055
+    },
+    {
+      "epoch": 0.07701474402411385,
+      "grad_norm": 0.2874973714351654,
+      "learning_rate": 1.9708725879671856e-05,
+      "loss": 0.334,
+      "step": 2060
+    },
+    {
+      "epoch": 0.07720167301446364,
+      "grad_norm": 0.4508047103881836,
+      "learning_rate": 1.9707317158031794e-05,
+      "loss": 0.2811,
+      "step": 2065
+    },
+    {
+      "epoch": 0.07738860200481343,
+      "grad_norm": 0.233233243227005,
+      "learning_rate": 1.9705905088616195e-05,
+      "loss": 0.2521,
+      "step": 2070
+    },
+    {
+      "epoch": 0.0775755309951632,
+      "grad_norm": 0.31307104229927063,
+      "learning_rate": 1.970448967191204e-05,
+      "loss": 0.322,
+      "step": 2075
+    },
+    {
+      "epoch": 0.077762459985513,
+      "grad_norm": 0.6601381897926331,
+      "learning_rate": 1.9703070908407463e-05,
+      "loss": 0.3456,
+      "step": 2080
+    },
+    {
+      "epoch": 0.07794938897586279,
+      "grad_norm": 0.37102803587913513,
+      "learning_rate": 1.970164879859176e-05,
+      "loss": 0.374,
+      "step": 2085
+    },
+    {
+      "epoch": 0.07813631796621258,
+      "grad_norm": 0.37135210633277893,
+      "learning_rate": 1.9700223342955374e-05,
+      "loss": 0.3463,
+      "step": 2090
+    },
+    {
+      "epoch": 0.07832324695656237,
+      "grad_norm": 0.5516648292541504,
+      "learning_rate": 1.9698794541989903e-05,
+      "loss": 0.4849,
+      "step": 2095
+    },
+    {
+      "epoch": 0.07851017594691216,
+      "grad_norm": 0.5477395057678223,
+      "learning_rate": 1.96973623961881e-05,
+      "loss": 0.2745,
+      "step": 2100
+    },
+    {
+      "epoch": 0.07869710493726195,
+      "grad_norm": 0.5191754102706909,
+      "learning_rate": 1.969592690604387e-05,
+      "loss": 0.3044,
+      "step": 2105
+    },
+    {
+      "epoch": 0.07888403392761174,
+      "grad_norm": 1.8438035249710083,
+      "learning_rate": 1.9694488072052275e-05,
+      "loss": 0.3786,
+      "step": 2110
+    },
+    {
+      "epoch": 0.07907096291796153,
+      "grad_norm": 0.3772248327732086,
+      "learning_rate": 1.9693045894709524e-05,
+      "loss": 0.2567,
+      "step": 2115
+    },
+    {
+      "epoch": 0.07925789190831133,
+      "grad_norm": 0.6036641597747803,
+      "learning_rate": 1.9691600374512988e-05,
+      "loss": 0.3108,
+      "step": 2120
+    },
+    {
+      "epoch": 0.07944482089866112,
+      "grad_norm": 0.4737173020839691,
+      "learning_rate": 1.969015151196118e-05,
+      "loss": 0.2473,
+      "step": 2125
+    },
+    {
+      "epoch": 0.07963174988901091,
+      "grad_norm": 0.741744339466095,
+      "learning_rate": 1.9688699307553774e-05,
+      "loss": 0.3534,
+      "step": 2130
+    },
+    {
+      "epoch": 0.0798186788793607,
+      "grad_norm": 0.4742072522640228,
+      "learning_rate": 1.9687243761791595e-05,
+      "loss": 0.3127,
+      "step": 2135
+    },
+    {
+      "epoch": 0.08000560786971049,
+      "grad_norm": 0.4502277374267578,
+      "learning_rate": 1.9685784875176613e-05,
+      "loss": 0.3633,
+      "step": 2140
+    },
+    {
+      "epoch": 0.08019253686006028,
+      "grad_norm": 0.9466716647148132,
+      "learning_rate": 1.9684322648211964e-05,
+      "loss": 0.3007,
+      "step": 2145
+    },
+    {
+      "epoch": 0.08037946585041007,
+      "grad_norm": 0.4492679834365845,
+      "learning_rate": 1.9682857081401923e-05,
+      "loss": 0.2861,
+      "step": 2150
+    },
+    {
+      "epoch": 0.08056639484075986,
+      "grad_norm": 0.8264315724372864,
+      "learning_rate": 1.9681388175251925e-05,
+      "loss": 0.3445,
+      "step": 2155
+    },
+    {
+      "epoch": 0.08075332383110966,
+      "grad_norm": 0.7091488242149353,
+      "learning_rate": 1.9679915930268553e-05,
+      "loss": 0.3153,
+      "step": 2160
+    },
+    {
+      "epoch": 0.08094025282145945,
+      "grad_norm": 0.5118813514709473,
+      "learning_rate": 1.967844034695954e-05,
+      "loss": 0.3658,
+      "step": 2165
+    },
+    {
+      "epoch": 0.08112718181180924,
+      "grad_norm": 0.6051716208457947,
+      "learning_rate": 1.967696142583377e-05,
+      "loss": 0.3552,
+      "step": 2170
+    },
+    {
+      "epoch": 0.08131411080215903,
+      "grad_norm": 0.5814588665962219,
+      "learning_rate": 1.967547916740129e-05,
+      "loss": 0.3133,
+      "step": 2175
+    },
+    {
+      "epoch": 0.08150103979250882,
+      "grad_norm": 0.6085551977157593,
+      "learning_rate": 1.967399357217328e-05,
+      "loss": 0.3033,
+      "step": 2180
+    },
+    {
+      "epoch": 0.08168796878285861,
+      "grad_norm": 0.5651482939720154,
+      "learning_rate": 1.9672504640662083e-05,
+      "loss": 0.3819,
+      "step": 2185
+    },
+    {
+      "epoch": 0.0818748977732084,
+      "grad_norm": 0.3710572123527527,
+      "learning_rate": 1.9671012373381188e-05,
+      "loss": 0.3037,
+      "step": 2190
+    },
+    {
+      "epoch": 0.0820618267635582,
+      "grad_norm": 0.43990039825439453,
+      "learning_rate": 1.9669516770845233e-05,
+      "loss": 0.4436,
+      "step": 2195
+    },
+    {
+      "epoch": 0.08224875575390798,
+      "grad_norm": 0.6509809494018555,
+      "learning_rate": 1.966801783357001e-05,
+      "loss": 0.2811,
+      "step": 2200
+    },
+    {
+      "epoch": 0.08243568474425778,
+      "grad_norm": 0.5239195823669434,
+      "learning_rate": 1.9666515562072463e-05,
+      "loss": 0.3285,
+      "step": 2205
+    },
+    {
+      "epoch": 0.08262261373460757,
+      "grad_norm": 0.6424049139022827,
+      "learning_rate": 1.9665009956870678e-05,
+      "loss": 0.2595,
+      "step": 2210
+    },
+    {
+      "epoch": 0.08280954272495736,
+      "grad_norm": 0.3653183579444885,
+      "learning_rate": 1.9663501018483897e-05,
+      "loss": 0.3201,
+      "step": 2215
+    },
+    {
+      "epoch": 0.08299647171530715,
+      "grad_norm": 0.6920386552810669,
+      "learning_rate": 1.9661988747432508e-05,
+      "loss": 0.2718,
+      "step": 2220
+    },
+    {
+      "epoch": 0.08318340070565694,
+      "grad_norm": 0.4126833379268646,
+      "learning_rate": 1.966047314423805e-05,
+      "loss": 0.336,
+      "step": 2225
+    },
+    {
+      "epoch": 0.08337032969600673,
+      "grad_norm": 0.11903409659862518,
+      "learning_rate": 1.9658954209423214e-05,
+      "loss": 0.3856,
+      "step": 2230
+    },
+    {
+      "epoch": 0.08355725868635652,
+      "grad_norm": 0.44001504778862,
+      "learning_rate": 1.9657431943511837e-05,
+      "loss": 0.3288,
+      "step": 2235
+    },
+    {
+      "epoch": 0.08374418767670631,
+      "grad_norm": 0.2602495849132538,
+      "learning_rate": 1.96559063470289e-05,
+      "loss": 0.2417,
+      "step": 2240
+    },
+    {
+      "epoch": 0.0839311166670561,
+      "grad_norm": 0.48337632417678833,
+      "learning_rate": 1.965437742050054e-05,
+      "loss": 0.4299,
+      "step": 2245
+    },
+    {
+      "epoch": 0.0841180456574059,
+      "grad_norm": 0.40766072273254395,
+      "learning_rate": 1.9652845164454044e-05,
+      "loss": 0.3488,
+      "step": 2250
+    },
+    {
+      "epoch": 0.08430497464775569,
+      "grad_norm": 0.4196990132331848,
+      "learning_rate": 1.9651309579417835e-05,
+      "loss": 0.3123,
+      "step": 2255
+    },
+    {
+      "epoch": 0.08449190363810548,
+      "grad_norm": 0.4198533296585083,
+      "learning_rate": 1.96497706659215e-05,
+      "loss": 0.2318,
+      "step": 2260
+    },
+    {
+      "epoch": 0.08467883262845527,
+      "grad_norm": 0.48852840065956116,
+      "learning_rate": 1.9648228424495765e-05,
+      "loss": 0.2639,
+      "step": 2265
+    },
+    {
+      "epoch": 0.08486576161880506,
+      "grad_norm": 0.42771658301353455,
+      "learning_rate": 1.96466828556725e-05,
+      "loss": 0.2426,
+      "step": 2270
+    },
+    {
+      "epoch": 0.08505269060915485,
+      "grad_norm": 0.46135228872299194,
+      "learning_rate": 1.9645133959984733e-05,
+      "loss": 0.3736,
+      "step": 2275
+    },
+    {
+      "epoch": 0.08523961959950464,
+      "grad_norm": 0.673338770866394,
+      "learning_rate": 1.9643581737966628e-05,
+      "loss": 0.2744,
+      "step": 2280
+    },
+    {
+      "epoch": 0.08542654858985443,
+      "grad_norm": 0.5732733011245728,
+      "learning_rate": 1.964202619015351e-05,
+      "loss": 0.3472,
+      "step": 2285
+    },
+    {
+      "epoch": 0.08561347758020423,
+      "grad_norm": 0.6712455749511719,
+      "learning_rate": 1.9640467317081833e-05,
+      "loss": 0.2671,
+      "step": 2290
+    },
+    {
+      "epoch": 0.08580040657055402,
+      "grad_norm": 0.6656973958015442,
+      "learning_rate": 1.9638905119289215e-05,
+      "loss": 0.3475,
+      "step": 2295
+    },
+    {
+      "epoch": 0.08598733556090381,
+      "grad_norm": 0.3552950620651245,
+      "learning_rate": 1.963733959731441e-05,
+      "loss": 0.2821,
+      "step": 2300
+    },
+    {
+      "epoch": 0.0861742645512536,
+      "grad_norm": 0.7310354709625244,
+      "learning_rate": 1.9635770751697326e-05,
+      "loss": 0.3669,
+      "step": 2305
+    },
+    {
+      "epoch": 0.08636119354160339,
+      "grad_norm": 0.3826546370983124,
+      "learning_rate": 1.9634198582979005e-05,
+      "loss": 0.3274,
+      "step": 2310
+    },
+    {
+      "epoch": 0.08654812253195318,
+      "grad_norm": 0.8085774183273315,
+      "learning_rate": 1.963262309170165e-05,
+      "loss": 0.303,
+      "step": 2315
+    },
+    {
+      "epoch": 0.08673505152230296,
+      "grad_norm": 0.2076191008090973,
+      "learning_rate": 1.96310442784086e-05,
+      "loss": 0.2976,
+      "step": 2320
+    },
+    {
+      "epoch": 0.08692198051265275,
+      "grad_norm": 0.8541048169136047,
+      "learning_rate": 1.962946214364434e-05,
+      "loss": 0.3923,
+      "step": 2325
+    },
+    {
+      "epoch": 0.08710890950300254,
+      "grad_norm": 0.4931800663471222,
+      "learning_rate": 1.9627876687954508e-05,
+      "loss": 0.3641,
+      "step": 2330
+    },
+    {
+      "epoch": 0.08729583849335233,
+      "grad_norm": 0.33070865273475647,
+      "learning_rate": 1.9626287911885882e-05,
+      "loss": 0.3196,
+      "step": 2335
+    },
+    {
+      "epoch": 0.08748276748370212,
+      "grad_norm": 0.2637079060077667,
+      "learning_rate": 1.9624695815986383e-05,
+      "loss": 0.3229,
+      "step": 2340
+    },
+    {
+      "epoch": 0.08766969647405191,
+      "grad_norm": 0.34368112683296204,
+      "learning_rate": 1.9623100400805076e-05,
+      "loss": 0.2956,
+      "step": 2345
+    },
+    {
+      "epoch": 0.0878566254644017,
+      "grad_norm": 0.5603893995285034,
+      "learning_rate": 1.9621501666892178e-05,
+      "loss": 0.2941,
+      "step": 2350
+    },
+    {
+      "epoch": 0.0880435544547515,
+      "grad_norm": 0.5940320491790771,
+      "learning_rate": 1.9619899614799046e-05,
+      "loss": 0.315,
+      "step": 2355
+    },
+    {
+      "epoch": 0.08823048344510129,
+      "grad_norm": 0.36051857471466064,
+      "learning_rate": 1.9618294245078184e-05,
+      "loss": 0.391,
+      "step": 2360
+    },
+    {
+      "epoch": 0.08841741243545108,
+      "grad_norm": 0.5375169515609741,
+      "learning_rate": 1.9616685558283234e-05,
+      "loss": 0.3213,
+      "step": 2365
+    },
+    {
+      "epoch": 0.08860434142580087,
+      "grad_norm": 0.6388794779777527,
+      "learning_rate": 1.9615073554968988e-05,
+      "loss": 0.347,
+      "step": 2370
+    },
+    {
+      "epoch": 0.08879127041615066,
+      "grad_norm": 1.1879587173461914,
+      "learning_rate": 1.961345823569138e-05,
+      "loss": 0.2717,
+      "step": 2375
+    },
+    {
+      "epoch": 0.08897819940650045,
+      "grad_norm": 0.21704648435115814,
+      "learning_rate": 1.961183960100749e-05,
+      "loss": 0.3559,
+      "step": 2380
+    },
+    {
+      "epoch": 0.08916512839685024,
+      "grad_norm": 1.803924560546875,
+      "learning_rate": 1.961021765147553e-05,
+      "loss": 0.3177,
+      "step": 2385
+    },
+    {
+      "epoch": 0.08935205738720003,
+      "grad_norm": 0.6759556531906128,
+      "learning_rate": 1.9608592387654873e-05,
+      "loss": 0.2972,
+      "step": 2390
+    },
+    {
+      "epoch": 0.08953898637754983,
+      "grad_norm": 0.2949463725090027,
+      "learning_rate": 1.9606963810106023e-05,
+      "loss": 0.3087,
+      "step": 2395
+    },
+    {
+      "epoch": 0.08972591536789962,
+      "grad_norm": 0.34233054518699646,
+      "learning_rate": 1.9605331919390627e-05,
+      "loss": 0.4371,
+      "step": 2400
+    },
+    {
+      "epoch": 0.08991284435824941,
+      "grad_norm": 0.3700423836708069,
+      "learning_rate": 1.960369671607148e-05,
+      "loss": 0.2686,
+      "step": 2405
+    },
+    {
+      "epoch": 0.0900997733485992,
+      "grad_norm": 0.528294026851654,
+      "learning_rate": 1.9602058200712516e-05,
+      "loss": 0.3552,
+      "step": 2410
+    },
+    {
+      "epoch": 0.09028670233894899,
+      "grad_norm": 2.334455966949463,
+      "learning_rate": 1.9600416373878815e-05,
+      "loss": 0.3232,
+      "step": 2415
+    },
+    {
+      "epoch": 0.09047363132929878,
+      "grad_norm": 0.4088938236236572,
+      "learning_rate": 1.959877123613659e-05,
+      "loss": 0.2649,
+      "step": 2420
+    },
+    {
+      "epoch": 0.09066056031964857,
+      "grad_norm": 0.5448519587516785,
+      "learning_rate": 1.959712278805321e-05,
+      "loss": 0.3223,
+      "step": 2425
+    },
+    {
+      "epoch": 0.09084748930999836,
+      "grad_norm": 0.5493069291114807,
+      "learning_rate": 1.9595471030197165e-05,
+      "loss": 0.345,
+      "step": 2430
+    },
+    {
+      "epoch": 0.09103441830034815,
+      "grad_norm": 0.39371412992477417,
+      "learning_rate": 1.9593815963138115e-05,
+      "loss": 0.3379,
+      "step": 2435
+    },
+    {
+      "epoch": 0.09122134729069795,
+      "grad_norm": 0.5203854441642761,
+      "learning_rate": 1.9592157587446833e-05,
+      "loss": 0.2653,
+      "step": 2440
+    },
+    {
+      "epoch": 0.09140827628104774,
+      "grad_norm": 0.4320240616798401,
+      "learning_rate": 1.9590495903695248e-05,
+      "loss": 0.3827,
+      "step": 2445
+    },
+    {
+      "epoch": 0.09159520527139753,
+      "grad_norm": 0.35967013239860535,
+      "learning_rate": 1.958883091245643e-05,
+      "loss": 0.4025,
+      "step": 2450
+    },
+    {
+      "epoch": 0.09178213426174732,
+      "grad_norm": 0.47668302059173584,
+      "learning_rate": 1.9587162614304588e-05,
+      "loss": 0.2894,
+      "step": 2455
+    },
+    {
+      "epoch": 0.09196906325209711,
+      "grad_norm": 0.463037371635437,
+      "learning_rate": 1.958549100981506e-05,
+      "loss": 0.2681,
+      "step": 2460
+    },
+    {
+      "epoch": 0.0921559922424469,
+      "grad_norm": 0.5469281077384949,
+      "learning_rate": 1.9583816099564346e-05,
+      "loss": 0.346,
+      "step": 2465
+    },
+    {
+      "epoch": 0.0923429212327967,
+      "grad_norm": 0.40648287534713745,
+      "learning_rate": 1.958213788413007e-05,
+      "loss": 0.3613,
+      "step": 2470
+    },
+    {
+      "epoch": 0.09252985022314648,
+      "grad_norm": 0.35621514916419983,
+      "learning_rate": 1.9580456364091003e-05,
+      "loss": 0.3235,
+      "step": 2475
+    },
+    {
+      "epoch": 0.09271677921349628,
+      "grad_norm": 0.6592122316360474,
+      "learning_rate": 1.957877154002705e-05,
+      "loss": 0.3155,
+      "step": 2480
+    },
+    {
+      "epoch": 0.09290370820384607,
+      "grad_norm": 0.515332818031311,
+      "learning_rate": 1.9577083412519258e-05,
+      "loss": 0.2627,
+      "step": 2485
+    },
+    {
+      "epoch": 0.09309063719419586,
+      "grad_norm": 0.3281778395175934,
+      "learning_rate": 1.9575391982149814e-05,
+      "loss": 0.3066,
+      "step": 2490
+    },
+    {
+      "epoch": 0.09327756618454565,
+      "grad_norm": 0.4038369953632355,
+      "learning_rate": 1.9573697249502046e-05,
+      "loss": 0.2842,
+      "step": 2495
+    },
+    {
+      "epoch": 0.09346449517489544,
+      "grad_norm": 0.5899633169174194,
+      "learning_rate": 1.957199921516042e-05,
+      "loss": 0.2495,
+      "step": 2500
+    },
+    {
+      "epoch": 0.09365142416524523,
+      "grad_norm": 0.2653183043003082,
+      "learning_rate": 1.9570297879710533e-05,
+      "loss": 0.3687,
+      "step": 2505
+    },
+    {
+      "epoch": 0.09383835315559502,
+      "grad_norm": 0.5264580249786377,
+      "learning_rate": 1.9568593243739133e-05,
+      "loss": 0.3317,
+      "step": 2510
+    },
+    {
+      "epoch": 0.09402528214594481,
+      "grad_norm": 0.8166529536247253,
+      "learning_rate": 1.95668853078341e-05,
+      "loss": 0.3921,
+      "step": 2515
+    },
+    {
+      "epoch": 0.0942122111362946,
+      "grad_norm": 3.684605360031128,
+      "learning_rate": 1.9565174072584448e-05,
+      "loss": 0.383,
+      "step": 2520
+    },
+    {
+      "epoch": 0.0943991401266444,
+      "grad_norm": 0.3563925325870514,
+      "learning_rate": 1.9563459538580337e-05,
+      "loss": 0.2752,
+      "step": 2525
+    },
+    {
+      "epoch": 0.09458606911699419,
+      "grad_norm": 0.696751594543457,
+      "learning_rate": 1.9561741706413055e-05,
+      "loss": 0.2909,
+      "step": 2530
+    },
+    {
+      "epoch": 0.09477299810734398,
+      "grad_norm": 0.47342103719711304,
+      "learning_rate": 1.956002057667504e-05,
+      "loss": 0.3115,
+      "step": 2535
+    },
+    {
+      "epoch": 0.09495992709769377,
+      "grad_norm": 0.3338661193847656,
+      "learning_rate": 1.955829614995986e-05,
+      "loss": 0.3501,
+      "step": 2540
+    },
+    {
+      "epoch": 0.09514685608804356,
+      "grad_norm": 0.48812028765678406,
+      "learning_rate": 1.9556568426862214e-05,
+      "loss": 0.3285,
+      "step": 2545
+    },
+    {
+      "epoch": 0.09533378507839335,
+      "grad_norm": 0.5304372310638428,
+      "learning_rate": 1.9554837407977948e-05,
+      "loss": 0.3472,
+      "step": 2550
+    },
+    {
+      "epoch": 0.09552071406874314,
+      "grad_norm": 0.5442197322845459,
+      "learning_rate": 1.9553103093904043e-05,
+      "loss": 0.3138,
+      "step": 2555
+    },
+    {
+      "epoch": 0.09570764305909293,
+      "grad_norm": 0.34639909863471985,
+      "learning_rate": 1.955136548523861e-05,
+      "loss": 0.2917,
+      "step": 2560
+    },
+    {
+      "epoch": 0.09589457204944271,
+      "grad_norm": 0.43235430121421814,
+      "learning_rate": 1.9549624582580905e-05,
+      "loss": 0.382,
+      "step": 2565
+    },
+    {
+      "epoch": 0.0960815010397925,
+      "grad_norm": 0.37063688039779663,
+      "learning_rate": 1.9547880386531307e-05,
+      "loss": 0.3394,
+      "step": 2570
+    },
+    {
+      "epoch": 0.0962684300301423,
+      "grad_norm": 1.0126583576202393,
+      "learning_rate": 1.954613289769135e-05,
+      "loss": 0.3272,
+      "step": 2575
+    },
+    {
+      "epoch": 0.09645535902049208,
+      "grad_norm": 0.6808075308799744,
+      "learning_rate": 1.9544382116663687e-05,
+      "loss": 0.3496,
+      "step": 2580
+    },
+    {
+      "epoch": 0.09664228801084188,
+      "grad_norm": 0.4482937753200531,
+      "learning_rate": 1.9542628044052115e-05,
+      "loss": 0.3691,
+      "step": 2585
+    },
+    {
+      "epoch": 0.09682921700119167,
+      "grad_norm": 0.5843492150306702,
+      "learning_rate": 1.954087068046156e-05,
+      "loss": 0.3311,
+      "step": 2590
+    },
+    {
+      "epoch": 0.09701614599154146,
+      "grad_norm": 0.45115911960601807,
+      "learning_rate": 1.9539110026498085e-05,
+      "loss": 0.2578,
+      "step": 2595
+    },
+    {
+      "epoch": 0.09720307498189125,
+      "grad_norm": 0.4495108425617218,
+      "learning_rate": 1.9537346082768894e-05,
+      "loss": 0.3572,
+      "step": 2600
+    },
+    {
+      "epoch": 0.09739000397224104,
+      "grad_norm": 0.5485822558403015,
+      "learning_rate": 1.9535578849882318e-05,
+      "loss": 0.3271,
+      "step": 2605
+    },
+    {
+      "epoch": 0.09757693296259083,
+      "grad_norm": 0.3917326331138611,
+      "learning_rate": 1.9533808328447828e-05,
+      "loss": 0.2825,
+      "step": 2610
+    },
+    {
+      "epoch": 0.09776386195294062,
+      "grad_norm": 0.7909905314445496,
+      "learning_rate": 1.953203451907602e-05,
+      "loss": 0.299,
+      "step": 2615
+    },
+    {
+      "epoch": 0.09795079094329041,
+      "grad_norm": 0.6403763294219971,
+      "learning_rate": 1.9530257422378635e-05,
+      "loss": 0.4017,
+      "step": 2620
+    },
+    {
+      "epoch": 0.0981377199336402,
+      "grad_norm": 0.5497360229492188,
+      "learning_rate": 1.9528477038968542e-05,
+      "loss": 0.4055,
+      "step": 2625
+    },
+    {
+      "epoch": 0.09832464892399,
+      "grad_norm": 0.35415706038475037,
+      "learning_rate": 1.9526693369459747e-05,
+      "loss": 0.2735,
+      "step": 2630
+    },
+    {
+      "epoch": 0.09851157791433979,
+      "grad_norm": 0.6400545239448547,
+      "learning_rate": 1.9524906414467376e-05,
+      "loss": 0.3236,
+      "step": 2635
+    },
+    {
+      "epoch": 0.09869850690468958,
+      "grad_norm": 0.6510913968086243,
+      "learning_rate": 1.952311617460771e-05,
+      "loss": 0.2668,
+      "step": 2640
+    },
+    {
+      "epoch": 0.09888543589503937,
+      "grad_norm": 0.36324068903923035,
+      "learning_rate": 1.9521322650498148e-05,
+      "loss": 0.3129,
+      "step": 2645
+    },
+    {
+      "epoch": 0.09907236488538916,
+      "grad_norm": 0.4775453209877014,
+      "learning_rate": 1.9519525842757223e-05,
+      "loss": 0.265,
+      "step": 2650
+    },
+    {
+      "epoch": 0.09925929387573895,
+      "grad_norm": 0.6153637170791626,
+      "learning_rate": 1.9517725752004605e-05,
+      "loss": 0.2791,
+      "step": 2655
+    },
+    {
+      "epoch": 0.09944622286608874,
+      "grad_norm": 0.704619288444519,
+      "learning_rate": 1.951592237886109e-05,
+      "loss": 0.2513,
+      "step": 2660
+    },
+    {
+      "epoch": 0.09963315185643853,
+      "grad_norm": 0.49600765109062195,
+      "learning_rate": 1.9514115723948612e-05,
+      "loss": 0.416,
+      "step": 2665
+    },
+    {
+      "epoch": 0.09982008084678833,
+      "grad_norm": 0.5411927103996277,
+      "learning_rate": 1.9512305787890237e-05,
+      "loss": 0.3892,
+      "step": 2670
+    },
+    {
+      "epoch": 0.10000700983713812,
+      "grad_norm": 0.4148860275745392,
+      "learning_rate": 1.9510492571310157e-05,
+      "loss": 0.316,
+      "step": 2675
+    },
+    {
+      "epoch": 0.10019393882748791,
+      "grad_norm": 0.7685550451278687,
+      "learning_rate": 1.95086760748337e-05,
+      "loss": 0.3357,
+      "step": 2680
+    },
+    {
+      "epoch": 0.1003808678178377,
+      "grad_norm": 0.7916879057884216,
+      "learning_rate": 1.950685629908732e-05,
+      "loss": 0.4379,
+      "step": 2685
+    },
+    {
+      "epoch": 0.10056779680818749,
+      "grad_norm": 0.4039478003978729,
+      "learning_rate": 1.9505033244698614e-05,
+      "loss": 0.2756,
+      "step": 2690
+    },
+    {
+      "epoch": 0.10075472579853728,
+      "grad_norm": 0.4937886595726013,
+      "learning_rate": 1.950320691229629e-05,
+      "loss": 0.2877,
+      "step": 2695
+    },
+    {
+      "epoch": 0.10094165478888707,
+      "grad_norm": 0.7463287711143494,
+      "learning_rate": 1.9501377302510204e-05,
+      "loss": 0.277,
+      "step": 2700
+    },
+    {
+      "epoch": 0.10112858377923686,
+      "grad_norm": 0.591262936592102,
+      "learning_rate": 1.9499544415971337e-05,
+      "loss": 0.316,
+      "step": 2705
+    },
+    {
+      "epoch": 0.10131551276958665,
+      "grad_norm": 0.3209913372993469,
+      "learning_rate": 1.94977082533118e-05,
+      "loss": 0.2486,
+      "step": 2710
+    },
+    {
+      "epoch": 0.10150244175993645,
+      "grad_norm": 0.5596148371696472,
+      "learning_rate": 1.9495868815164827e-05,
+      "loss": 0.3891,
+      "step": 2715
+    },
+    {
+      "epoch": 0.10168937075028624,
+      "grad_norm": 0.3874911367893219,
+      "learning_rate": 1.949402610216479e-05,
+      "loss": 0.3855,
+      "step": 2720
+    },
+    {
+      "epoch": 0.10187629974063603,
+      "grad_norm": 0.5168285369873047,
+      "learning_rate": 1.9492180114947187e-05,
+      "loss": 0.2817,
+      "step": 2725
+    },
+    {
+      "epoch": 0.10206322873098582,
+      "grad_norm": 0.5476559996604919,
+      "learning_rate": 1.949033085414865e-05,
+      "loss": 0.3411,
+      "step": 2730
+    },
+    {
+      "epoch": 0.10225015772133561,
+      "grad_norm": 0.8609138131141663,
+      "learning_rate": 1.9488478320406937e-05,
+      "loss": 0.3765,
+      "step": 2735
+    },
+    {
+      "epoch": 0.1024370867116854,
+      "grad_norm": 0.5176308751106262,
+      "learning_rate": 1.9486622514360928e-05,
+      "loss": 0.2958,
+      "step": 2740
+    },
+    {
+      "epoch": 0.10262401570203519,
+      "grad_norm": 0.49085232615470886,
+      "learning_rate": 1.9484763436650637e-05,
+      "loss": 0.3155,
+      "step": 2745
+    },
+    {
+      "epoch": 0.10281094469238498,
+      "grad_norm": 0.5792192816734314,
+      "learning_rate": 1.948290108791721e-05,
+      "loss": 0.3046,
+      "step": 2750
+    },
+    {
+      "epoch": 0.10299787368273478,
+      "grad_norm": 0.5799537301063538,
+      "learning_rate": 1.9481035468802922e-05,
+      "loss": 0.3144,
+      "step": 2755
+    },
+    {
+      "epoch": 0.10318480267308457,
+      "grad_norm": 2.7372591495513916,
+      "learning_rate": 1.9479166579951162e-05,
+      "loss": 0.5132,
+      "step": 2760
+    },
+    {
+      "epoch": 0.10337173166343436,
+      "grad_norm": 0.6942354440689087,
+      "learning_rate": 1.9477294422006462e-05,
+      "loss": 0.312,
+      "step": 2765
+    },
+    {
+      "epoch": 0.10355866065378415,
+      "grad_norm": 0.42621228098869324,
+      "learning_rate": 1.947541899561448e-05,
+      "loss": 0.3342,
+      "step": 2770
+    },
+    {
+      "epoch": 0.10374558964413394,
+      "grad_norm": 0.2672087252140045,
+      "learning_rate": 1.9473540301421985e-05,
+      "loss": 0.3697,
+      "step": 2775
+    },
+    {
+      "epoch": 0.10393251863448373,
+      "grad_norm": 0.3490724563598633,
+      "learning_rate": 1.9471658340076895e-05,
+      "loss": 0.397,
+      "step": 2780
+    },
+    {
+      "epoch": 0.10411944762483352,
+      "grad_norm": 0.4320790767669678,
+      "learning_rate": 1.9469773112228237e-05,
+      "loss": 0.354,
+      "step": 2785
+    },
+    {
+      "epoch": 0.10430637661518331,
+      "grad_norm": 0.19510655105113983,
+      "learning_rate": 1.946788461852618e-05,
+      "loss": 0.3459,
+      "step": 2790
+    },
+    {
+      "epoch": 0.1044933056055331,
+      "grad_norm": 0.4755474627017975,
+      "learning_rate": 1.9465992859622006e-05,
+      "loss": 0.3761,
+      "step": 2795
+    },
+    {
+      "epoch": 0.1046802345958829,
+      "grad_norm": 0.3486561179161072,
+      "learning_rate": 1.946409783616813e-05,
+      "loss": 0.2303,
+      "step": 2800
+    },
+    {
+      "epoch": 0.10486716358623269,
+      "grad_norm": 0.7743209004402161,
+      "learning_rate": 1.946219954881809e-05,
+      "loss": 0.3435,
+      "step": 2805
+    },
+    {
+      "epoch": 0.10505409257658246,
+      "grad_norm": 1.9734418392181396,
+      "learning_rate": 1.9460297998226552e-05,
+      "loss": 0.3507,
+      "step": 2810
+    },
+    {
+      "epoch": 0.10524102156693226,
+      "grad_norm": 0.5270185470581055,
+      "learning_rate": 1.9458393185049303e-05,
+      "loss": 0.4341,
+      "step": 2815
+    },
+    {
+      "epoch": 0.10542795055728205,
+      "grad_norm": 0.41794025897979736,
+      "learning_rate": 1.945648510994327e-05,
+      "loss": 0.3172,
+      "step": 2820
+    },
+    {
+      "epoch": 0.10561487954763184,
+      "grad_norm": 0.22198650240898132,
+      "learning_rate": 1.9454573773566478e-05,
+      "loss": 0.2623,
+      "step": 2825
+    },
+    {
+      "epoch": 0.10580180853798163,
+      "grad_norm": 0.36959734559059143,
+      "learning_rate": 1.94526591765781e-05,
+      "loss": 0.2379,
+      "step": 2830
+    },
+    {
+      "epoch": 0.10598873752833142,
+      "grad_norm": 0.21317122876644135,
+      "learning_rate": 1.945074131963843e-05,
+      "loss": 0.3395,
+      "step": 2835
+    },
+    {
+      "epoch": 0.10617566651868121,
+      "grad_norm": 0.5071830153465271,
+      "learning_rate": 1.944882020340887e-05,
+      "loss": 0.3268,
+      "step": 2840
+    },
+    {
+      "epoch": 0.106362595509031,
+      "grad_norm": 0.4251273274421692,
+      "learning_rate": 1.944689582855197e-05,
+      "loss": 0.3505,
+      "step": 2845
+    },
+    {
+      "epoch": 0.1065495244993808,
+      "grad_norm": 0.6162049174308777,
+      "learning_rate": 1.9444968195731384e-05,
+      "loss": 0.308,
+      "step": 2850
+    },
+    {
+      "epoch": 0.10673645348973058,
+      "grad_norm": 0.522985577583313,
+      "learning_rate": 1.94430373056119e-05,
+      "loss": 0.2516,
+      "step": 2855
+    },
+    {
+      "epoch": 0.10692338248008038,
+      "grad_norm": 0.7239044308662415,
+      "learning_rate": 1.9441103158859427e-05,
+      "loss": 0.3029,
+      "step": 2860
+    },
+    {
+      "epoch": 0.10711031147043017,
+      "grad_norm": 0.5046908855438232,
+      "learning_rate": 1.9439165756141e-05,
+      "loss": 0.3179,
+      "step": 2865
+    },
+    {
+      "epoch": 0.10729724046077996,
+      "grad_norm": 0.4853164553642273,
+      "learning_rate": 1.9437225098124765e-05,
+      "loss": 0.2714,
+      "step": 2870
+    },
+    {
+      "epoch": 0.10748416945112975,
+      "grad_norm": 0.472434401512146,
+      "learning_rate": 1.9435281185480007e-05,
+      "loss": 0.2768,
+      "step": 2875
+    },
+    {
+      "epoch": 0.10767109844147954,
+      "grad_norm": 0.549471914768219,
+      "learning_rate": 1.943333401887712e-05,
+      "loss": 0.3117,
+      "step": 2880
+    },
+    {
+      "epoch": 0.10785802743182933,
+      "grad_norm": 0.3362756669521332,
+      "learning_rate": 1.943138359898763e-05,
+      "loss": 0.2906,
+      "step": 2885
+    },
+    {
+      "epoch": 0.10804495642217912,
+      "grad_norm": 0.6930930614471436,
+      "learning_rate": 1.9429429926484184e-05,
+      "loss": 0.2813,
+      "step": 2890
+    },
+    {
+      "epoch": 0.10823188541252891,
+      "grad_norm": 1.1984028816223145,
+      "learning_rate": 1.942747300204054e-05,
+      "loss": 0.3159,
+      "step": 2895
+    },
+    {
+      "epoch": 0.1084188144028787,
+      "grad_norm": 0.6312728524208069,
+      "learning_rate": 1.9425512826331593e-05,
+      "loss": 0.3663,
+      "step": 2900
+    },
+    {
+      "epoch": 0.1086057433932285,
+      "grad_norm": 0.4989972710609436,
+      "learning_rate": 1.9423549400033344e-05,
+      "loss": 0.3155,
+      "step": 2905
+    },
+    {
+      "epoch": 0.10879267238357829,
+      "grad_norm": 0.22321349382400513,
+      "learning_rate": 1.9421582723822926e-05,
+      "loss": 0.3021,
+      "step": 2910
+    },
+    {
+      "epoch": 0.10897960137392808,
+      "grad_norm": 0.4356032609939575,
+      "learning_rate": 1.9419612798378588e-05,
+      "loss": 0.341,
+      "step": 2915
+    },
+    {
+      "epoch": 0.10916653036427787,
+      "grad_norm": 0.6972177028656006,
+      "learning_rate": 1.9417639624379704e-05,
+      "loss": 0.31,
+      "step": 2920
+    },
+    {
+      "epoch": 0.10935345935462766,
+      "grad_norm": 0.6108505725860596,
+      "learning_rate": 1.9415663202506757e-05,
+      "loss": 0.2865,
+      "step": 2925
+    },
+    {
+      "epoch": 0.10954038834497745,
+      "grad_norm": 0.29747897386550903,
+      "learning_rate": 1.941368353344137e-05,
+      "loss": 0.2782,
+      "step": 2930
+    },
+    {
+      "epoch": 0.10972731733532724,
+      "grad_norm": 0.46353158354759216,
+      "learning_rate": 1.9411700617866268e-05,
+      "loss": 0.3503,
+      "step": 2935
+    },
+    {
+      "epoch": 0.10991424632567703,
+      "grad_norm": 0.3761847913265228,
+      "learning_rate": 1.9409714456465303e-05,
+      "loss": 0.3263,
+      "step": 2940
+    },
+    {
+      "epoch": 0.11010117531602683,
+      "grad_norm": 0.3461897671222687,
+      "learning_rate": 1.9407725049923443e-05,
+      "loss": 0.2596,
+      "step": 2945
+    },
+    {
+      "epoch": 0.11028810430637662,
+      "grad_norm": 0.3752954602241516,
+      "learning_rate": 1.940573239892678e-05,
+      "loss": 0.3043,
+      "step": 2950
+    },
+    {
+      "epoch": 0.11047503329672641,
+      "grad_norm": 0.41294610500335693,
+      "learning_rate": 1.940373650416252e-05,
+      "loss": 0.327,
+      "step": 2955
+    },
+    {
+      "epoch": 0.1106619622870762,
+      "grad_norm": 0.4009648859500885,
+      "learning_rate": 1.940173736631899e-05,
+      "loss": 0.2781,
+      "step": 2960
+    },
+    {
+      "epoch": 0.11084889127742599,
+      "grad_norm": 0.20430001616477966,
+      "learning_rate": 1.9399734986085636e-05,
+      "loss": 0.3943,
+      "step": 2965
+    },
+    {
+      "epoch": 0.11103582026777578,
+      "grad_norm": 0.7819133400917053,
+      "learning_rate": 1.9397729364153025e-05,
+      "loss": 0.295,
+      "step": 2970
+    },
+    {
+      "epoch": 0.11122274925812557,
+      "grad_norm": 0.7665151953697205,
+      "learning_rate": 1.9395720501212833e-05,
+      "loss": 0.2599,
+      "step": 2975
+    },
+    {
+      "epoch": 0.11140967824847536,
+      "grad_norm": 0.38904261589050293,
+      "learning_rate": 1.9393708397957863e-05,
+      "loss": 0.3369,
+      "step": 2980
+    },
+    {
+      "epoch": 0.11159660723882515,
+      "grad_norm": 0.4791386127471924,
+      "learning_rate": 1.9391693055082028e-05,
+      "loss": 0.3276,
+      "step": 2985
+    },
+    {
+      "epoch": 0.11178353622917495,
+      "grad_norm": 0.5694244503974915,
+      "learning_rate": 1.9389674473280365e-05,
+      "loss": 0.3659,
+      "step": 2990
+    },
+    {
+      "epoch": 0.11197046521952474,
+      "grad_norm": 0.7917852401733398,
+      "learning_rate": 1.9387652653249023e-05,
+      "loss": 0.333,
+      "step": 2995
+    },
+    {
+      "epoch": 0.11215739420987453,
+      "grad_norm": 0.41513529419898987,
+      "learning_rate": 1.9385627595685275e-05,
+      "loss": 0.2983,
+      "step": 3000
+    },
+    {
+      "epoch": 0.11234432320022432,
+      "grad_norm": 0.39530879259109497,
+      "learning_rate": 1.9383599301287498e-05,
+      "loss": 0.2979,
+      "step": 3005
+    },
+    {
+      "epoch": 0.11253125219057411,
+      "grad_norm": 0.40975919365882874,
+      "learning_rate": 1.93815677707552e-05,
+      "loss": 0.3145,
+      "step": 3010
+    },
+    {
+      "epoch": 0.1127181811809239,
+      "grad_norm": 0.7045682072639465,
+      "learning_rate": 1.9379533004788992e-05,
+      "loss": 0.3777,
+      "step": 3015
+    },
+    {
+      "epoch": 0.11290511017127369,
+      "grad_norm": 0.5002708435058594,
+      "learning_rate": 1.9377495004090605e-05,
+      "loss": 0.3937,
+      "step": 3020
+    },
+    {
+      "epoch": 0.11309203916162348,
+      "grad_norm": 0.9641180634498596,
+      "learning_rate": 1.937545376936289e-05,
+      "loss": 0.2889,
+      "step": 3025
+    },
+    {
+      "epoch": 0.11327896815197327,
+      "grad_norm": 0.41359230875968933,
+      "learning_rate": 1.9373409301309817e-05,
+      "loss": 0.2924,
+      "step": 3030
+    },
+    {
+      "epoch": 0.11346589714232307,
+      "grad_norm": 0.5380722284317017,
+      "learning_rate": 1.9371361600636452e-05,
+      "loss": 0.3907,
+      "step": 3035
+    },
+    {
+      "epoch": 0.11365282613267286,
+      "grad_norm": 0.8563070297241211,
+      "learning_rate": 1.9369310668049e-05,
+      "loss": 0.341,
+      "step": 3040
+    },
+    {
+      "epoch": 0.11383975512302265,
+      "grad_norm": 1.5900834798812866,
+      "learning_rate": 1.936725650425476e-05,
+      "loss": 0.3633,
+      "step": 3045
+    },
+    {
+      "epoch": 0.11402668411337243,
+      "grad_norm": 0.4389062821865082,
+      "learning_rate": 1.936519910996216e-05,
+      "loss": 0.341,
+      "step": 3050
+    },
+    {
+      "epoch": 0.11421361310372222,
+      "grad_norm": 0.4638637602329254,
+      "learning_rate": 1.936313848588073e-05,
+      "loss": 0.3595,
+      "step": 3055
+    },
+    {
+      "epoch": 0.11440054209407201,
+      "grad_norm": 0.541275680065155,
+      "learning_rate": 1.9361074632721125e-05,
+      "loss": 0.3566,
+      "step": 3060
+    },
+    {
+      "epoch": 0.1145874710844218,
+      "grad_norm": 0.38814595341682434,
+      "learning_rate": 1.935900755119511e-05,
+      "loss": 0.28,
+      "step": 3065
+    },
+    {
+      "epoch": 0.11477440007477159,
+      "grad_norm": 0.3759201169013977,
+      "learning_rate": 1.935693724201556e-05,
+      "loss": 0.3574,
+      "step": 3070
+    },
+    {
+      "epoch": 0.11496132906512138,
+      "grad_norm": 0.4212549030780792,
+      "learning_rate": 1.9354863705896464e-05,
+      "loss": 0.2823,
+      "step": 3075
+    },
+    {
+      "epoch": 0.11514825805547117,
+      "grad_norm": 0.24448025226593018,
+      "learning_rate": 1.9352786943552925e-05,
+      "loss": 0.3305,
+      "step": 3080
+    },
+    {
+      "epoch": 0.11533518704582096,
+      "grad_norm": 1.0636529922485352,
+      "learning_rate": 1.9350706955701163e-05,
+      "loss": 0.4083,
+      "step": 3085
+    },
+    {
+      "epoch": 0.11552211603617075,
+      "grad_norm": 0.554562509059906,
+      "learning_rate": 1.9348623743058504e-05,
+      "loss": 0.3184,
+      "step": 3090
+    },
+    {
+      "epoch": 0.11570904502652055,
+      "grad_norm": 0.5117464661598206,
+      "learning_rate": 1.9346537306343384e-05,
+      "loss": 0.2759,
+      "step": 3095
+    },
+    {
+      "epoch": 0.11589597401687034,
+      "grad_norm": 0.2832476794719696,
+      "learning_rate": 1.9344447646275367e-05,
+      "loss": 0.2997,
+      "step": 3100
+    },
+    {
+      "epoch": 0.11608290300722013,
+      "grad_norm": 0.38118836283683777,
+      "learning_rate": 1.9342354763575103e-05,
+      "loss": 0.3417,
+      "step": 3105
+    },
+    {
+      "epoch": 0.11626983199756992,
+      "grad_norm": 0.42836499214172363,
+      "learning_rate": 1.9340258658964376e-05,
+      "loss": 0.2619,
+      "step": 3110
+    },
+    {
+      "epoch": 0.11645676098791971,
+      "grad_norm": 0.6346667408943176,
+      "learning_rate": 1.9338159333166063e-05,
+      "loss": 0.2843,
+      "step": 3115
+    },
+    {
+      "epoch": 0.1166436899782695,
+      "grad_norm": 0.4850265681743622,
+      "learning_rate": 1.9336056786904175e-05,
+      "loss": 0.3905,
+      "step": 3120
+    },
+    {
+      "epoch": 0.1168306189686193,
+      "grad_norm": 0.6557062864303589,
+      "learning_rate": 1.9333951020903812e-05,
+      "loss": 0.3158,
+      "step": 3125
+    },
+    {
+      "epoch": 0.11701754795896908,
+      "grad_norm": 0.30885252356529236,
+      "learning_rate": 1.9331842035891193e-05,
+      "loss": 0.2957,
+      "step": 3130
+    },
+    {
+      "epoch": 0.11720447694931888,
+      "grad_norm": 0.3093513250350952,
+      "learning_rate": 1.9329729832593646e-05,
+      "loss": 0.3044,
+      "step": 3135
+    },
+    {
+      "epoch": 0.11739140593966867,
+      "grad_norm": 0.4149554967880249,
+      "learning_rate": 1.932761441173961e-05,
+      "loss": 0.3545,
+      "step": 3140
+    },
+    {
+      "epoch": 0.11757833493001846,
+      "grad_norm": 0.5750139951705933,
+      "learning_rate": 1.932549577405864e-05,
+      "loss": 0.2389,
+      "step": 3145
+    },
+    {
+      "epoch": 0.11776526392036825,
+      "grad_norm": 0.5002467632293701,
+      "learning_rate": 1.932337392028138e-05,
+      "loss": 0.3064,
+      "step": 3150
+    },
+    {
+      "epoch": 0.11795219291071804,
+      "grad_norm": 0.7145326733589172,
+      "learning_rate": 1.9321248851139605e-05,
+      "loss": 0.3383,
+      "step": 3155
+    },
+    {
+      "epoch": 0.11813912190106783,
+      "grad_norm": 0.3954809010028839,
+      "learning_rate": 1.9319120567366186e-05,
+      "loss": 0.3623,
+      "step": 3160
+    },
+    {
+      "epoch": 0.11832605089141762,
+      "grad_norm": 0.637115478515625,
+      "learning_rate": 1.9316989069695112e-05,
+      "loss": 0.2777,
+      "step": 3165
+    },
+    {
+      "epoch": 0.11851297988176741,
+      "grad_norm": 0.45427948236465454,
+      "learning_rate": 1.9314854358861478e-05,
+      "loss": 0.5832,
+      "step": 3170
+    },
+    {
+      "epoch": 0.1186999088721172,
+      "grad_norm": 0.41445136070251465,
+      "learning_rate": 1.931271643560147e-05,
+      "loss": 0.3425,
+      "step": 3175
+    },
+    {
+      "epoch": 0.118886837862467,
+      "grad_norm": 0.6264867186546326,
+      "learning_rate": 1.9310575300652416e-05,
+      "loss": 0.2942,
+      "step": 3180
+    },
+    {
+      "epoch": 0.11907376685281679,
+      "grad_norm": 0.2800760269165039,
+      "learning_rate": 1.9308430954752717e-05,
+      "loss": 0.2636,
+      "step": 3185
+    },
+    {
+      "epoch": 0.11926069584316658,
+      "grad_norm": 0.49179184436798096,
+      "learning_rate": 1.9306283398641906e-05,
+      "loss": 0.2834,
+      "step": 3190
+    },
+    {
+      "epoch": 0.11944762483351637,
+      "grad_norm": 0.17769566178321838,
+      "learning_rate": 1.9304132633060605e-05,
+      "loss": 0.3386,
+      "step": 3195
+    },
+    {
+      "epoch": 0.11963455382386616,
+      "grad_norm": 0.35259896516799927,
+      "learning_rate": 1.930197865875056e-05,
+      "loss": 0.3855,
+      "step": 3200
+    },
+    {
+      "epoch": 0.11982148281421595,
+      "grad_norm": 0.44811856746673584,
+      "learning_rate": 1.929982147645461e-05,
+      "loss": 0.3237,
+      "step": 3205
+    },
+    {
+      "epoch": 0.12000841180456574,
+      "grad_norm": 0.6957955360412598,
+      "learning_rate": 1.9297661086916704e-05,
+      "loss": 0.3235,
+      "step": 3210
+    },
+    {
+      "epoch": 0.12019534079491553,
+      "grad_norm": 0.35209953784942627,
+      "learning_rate": 1.9295497490881902e-05,
+      "loss": 0.3576,
+      "step": 3215
+    },
+    {
+      "epoch": 0.12038226978526533,
+      "grad_norm": 0.30803924798965454,
+      "learning_rate": 1.9293330689096366e-05,
+      "loss": 0.313,
+      "step": 3220
+    },
+    {
+      "epoch": 0.12056919877561512,
+      "grad_norm": 0.39552509784698486,
+      "learning_rate": 1.9291160682307363e-05,
+      "loss": 0.2947,
+      "step": 3225
+    },
+    {
+      "epoch": 0.12075612776596491,
+      "grad_norm": 0.5420857667922974,
+      "learning_rate": 1.9288987471263266e-05,
+      "loss": 0.3885,
+      "step": 3230
+    },
+    {
+      "epoch": 0.1209430567563147,
+      "grad_norm": 0.36529114842414856,
+      "learning_rate": 1.928681105671355e-05,
+      "loss": 0.3692,
+      "step": 3235
+    },
+    {
+      "epoch": 0.12112998574666449,
+      "grad_norm": 0.27940261363983154,
+      "learning_rate": 1.9284631439408804e-05,
+      "loss": 0.2968,
+      "step": 3240
+    },
+    {
+      "epoch": 0.12131691473701428,
+      "grad_norm": 0.38085952401161194,
+      "learning_rate": 1.9282448620100716e-05,
+      "loss": 0.3231,
+      "step": 3245
+    },
+    {
+      "epoch": 0.12150384372736407,
+      "grad_norm": 0.4754383862018585,
+      "learning_rate": 1.928026259954207e-05,
+      "loss": 0.396,
+      "step": 3250
+    },
+    {
+      "epoch": 0.12169077271771386,
+      "grad_norm": 0.5790641903877258,
+      "learning_rate": 1.927807337848677e-05,
+      "loss": 0.2934,
+      "step": 3255
+    },
+    {
+      "epoch": 0.12187770170806365,
+      "grad_norm": 0.4564800560474396,
+      "learning_rate": 1.927588095768981e-05,
+      "loss": 0.2913,
+      "step": 3260
+    },
+    {
+      "epoch": 0.12206463069841345,
+      "grad_norm": 0.5212480425834656,
+      "learning_rate": 1.9273685337907295e-05,
+      "loss": 0.2929,
+      "step": 3265
+    },
+    {
+      "epoch": 0.12225155968876324,
+      "grad_norm": 0.43916815519332886,
+      "learning_rate": 1.9271486519896434e-05,
+      "loss": 0.345,
+      "step": 3270
+    },
+    {
+      "epoch": 0.12243848867911303,
+      "grad_norm": 0.47202667593955994,
+      "learning_rate": 1.926928450441553e-05,
+      "loss": 0.3599,
+      "step": 3275
+    },
+    {
+      "epoch": 0.12262541766946282,
+      "grad_norm": 0.3427533507347107,
+      "learning_rate": 1.9267079292224005e-05,
+      "loss": 0.2895,
+      "step": 3280
+    },
+    {
+      "epoch": 0.12281234665981261,
+      "grad_norm": 0.23820289969444275,
+      "learning_rate": 1.9264870884082362e-05,
+      "loss": 0.2214,
+      "step": 3285
+    },
+    {
+      "epoch": 0.1229992756501624,
+      "grad_norm": 0.7368319630622864,
+      "learning_rate": 1.9262659280752224e-05,
+      "loss": 0.2429,
+      "step": 3290
+    },
+    {
+      "epoch": 0.12318620464051218,
+      "grad_norm": 0.36256974935531616,
+      "learning_rate": 1.9260444482996313e-05,
+      "loss": 0.3072,
+      "step": 3295
+    },
+    {
+      "epoch": 0.12337313363086197,
+      "grad_norm": 0.44561657309532166,
+      "learning_rate": 1.9258226491578443e-05,
+      "loss": 0.2567,
+      "step": 3300
+    },
+    {
+      "epoch": 0.12356006262121176,
+      "grad_norm": 0.708128035068512,
+      "learning_rate": 1.9256005307263536e-05,
+      "loss": 0.3169,
+      "step": 3305
+    },
+    {
+      "epoch": 0.12374699161156155,
+      "grad_norm": 0.37106049060821533,
+      "learning_rate": 1.925378093081762e-05,
+      "loss": 0.2737,
+      "step": 3310
+    },
+    {
+      "epoch": 0.12393392060191134,
+      "grad_norm": 3.3657779693603516,
+      "learning_rate": 1.925155336300781e-05,
+      "loss": 0.3468,
+      "step": 3315
+    },
+    {
+      "epoch": 0.12412084959226113,
+      "grad_norm": 0.33611029386520386,
+      "learning_rate": 1.9249322604602342e-05,
+      "loss": 0.3304,
+      "step": 3320
+    },
+    {
+      "epoch": 0.12430777858261093,
+      "grad_norm": 0.4634288251399994,
+      "learning_rate": 1.9247088656370528e-05,
+      "loss": 0.2953,
+      "step": 3325
+    },
+    {
+      "epoch": 0.12449470757296072,
+      "grad_norm": 1.3842734098434448,
+      "learning_rate": 1.9244851519082802e-05,
+      "loss": 0.4189,
+      "step": 3330
+    },
+    {
+      "epoch": 0.12468163656331051,
+      "grad_norm": 0.5978188514709473,
+      "learning_rate": 1.924261119351069e-05,
+      "loss": 0.3122,
+      "step": 3335
+    },
+    {
+      "epoch": 0.1248685655536603,
+      "grad_norm": 0.30435624718666077,
+      "learning_rate": 1.9240367680426804e-05,
+      "loss": 0.3596,
+      "step": 3340
+    },
+    {
+      "epoch": 0.1250554945440101,
+      "grad_norm": 0.30297866463661194,
+      "learning_rate": 1.923812098060488e-05,
+      "loss": 0.3459,
+      "step": 3345
+    },
+    {
+      "epoch": 0.12524242353435988,
+      "grad_norm": 0.3620128929615021,
+      "learning_rate": 1.923587109481973e-05,
+      "loss": 0.342,
+      "step": 3350
+    },
+    {
+      "epoch": 0.1254293525247097,
+      "grad_norm": 0.3281639516353607,
+      "learning_rate": 1.9233618023847285e-05,
+      "loss": 0.2632,
+      "step": 3355
+    },
+    {
+      "epoch": 0.12561628151505946,
+      "grad_norm": 0.3437563478946686,
+      "learning_rate": 1.923136176846456e-05,
+      "loss": 0.3144,
+      "step": 3360
+    },
+    {
+      "epoch": 0.12580321050540927,
+      "grad_norm": 0.4065919816493988,
+      "learning_rate": 1.9229102329449674e-05,
+      "loss": 0.328,
+      "step": 3365
+    },
+    {
+      "epoch": 0.12599013949575905,
+      "grad_norm": 0.714131236076355,
+      "learning_rate": 1.9226839707581838e-05,
+      "loss": 0.3692,
+      "step": 3370
+    },
+    {
+      "epoch": 0.12617706848610885,
+      "grad_norm": 0.6316368579864502,
+      "learning_rate": 1.9224573903641374e-05,
+      "loss": 0.3259,
+      "step": 3375
+    },
+    {
+      "epoch": 0.12636399747645863,
+      "grad_norm": 0.23467862606048584,
+      "learning_rate": 1.9222304918409684e-05,
+      "loss": 0.2745,
+      "step": 3380
+    },
+    {
+      "epoch": 0.12655092646680843,
+      "grad_norm": 0.5963695645332336,
+      "learning_rate": 1.9220032752669282e-05,
+      "loss": 0.297,
+      "step": 3385
+    },
+    {
+      "epoch": 0.1267378554571582,
+      "grad_norm": 0.38383743166923523,
+      "learning_rate": 1.921775740720377e-05,
+      "loss": 0.3192,
+      "step": 3390
+    },
+    {
+      "epoch": 0.12692478444750802,
+      "grad_norm": 0.5108599662780762,
+      "learning_rate": 1.9215478882797852e-05,
+      "loss": 0.321,
+      "step": 3395
+    },
+    {
+      "epoch": 0.1271117134378578,
+      "grad_norm": 0.6184415221214294,
+      "learning_rate": 1.9213197180237325e-05,
+      "loss": 0.3903,
+      "step": 3400
+    },
+    {
+      "epoch": 0.1272986424282076,
+      "grad_norm": 0.4752747416496277,
+      "learning_rate": 1.921091230030908e-05,
+      "loss": 0.2943,
+      "step": 3405
+    },
+    {
+      "epoch": 0.12748557141855738,
+      "grad_norm": 0.4967573583126068,
+      "learning_rate": 1.9208624243801107e-05,
+      "loss": 0.3427,
+      "step": 3410
+    },
+    {
+      "epoch": 0.12767250040890718,
+      "grad_norm": 0.6884995102882385,
+      "learning_rate": 1.9206333011502497e-05,
+      "loss": 0.2472,
+      "step": 3415
+    },
+    {
+      "epoch": 0.12785942939925696,
+      "grad_norm": 0.49374425411224365,
+      "learning_rate": 1.9204038604203423e-05,
+      "loss": 0.3751,
+      "step": 3420
+    },
+    {
+      "epoch": 0.12804635838960673,
+      "grad_norm": 0.38376346230506897,
+      "learning_rate": 1.9201741022695165e-05,
+      "loss": 0.3392,
+      "step": 3425
+    },
+    {
+      "epoch": 0.12823328737995654,
+      "grad_norm": 0.4707247018814087,
+      "learning_rate": 1.919944026777009e-05,
+      "loss": 0.2892,
+      "step": 3430
+    },
+    {
+      "epoch": 0.12842021637030632,
+      "grad_norm": 0.5075487494468689,
+      "learning_rate": 1.9197136340221667e-05,
+      "loss": 0.2932,
+      "step": 3435
+    },
+    {
+      "epoch": 0.12860714536065612,
+      "grad_norm": 0.3079968988895416,
+      "learning_rate": 1.9194829240844448e-05,
+      "loss": 0.3229,
+      "step": 3440
+    },
+    {
+      "epoch": 0.1287940743510059,
+      "grad_norm": 0.41244247555732727,
+      "learning_rate": 1.919251897043409e-05,
+      "loss": 0.3227,
+      "step": 3445
+    },
+    {
+      "epoch": 0.1289810033413557,
+      "grad_norm": 0.6245453953742981,
+      "learning_rate": 1.9190205529787336e-05,
+      "loss": 0.3171,
+      "step": 3450
+    },
+    {
+      "epoch": 0.12916793233170548,
+      "grad_norm": 0.5110964775085449,
+      "learning_rate": 1.9187888919702035e-05,
+      "loss": 0.3235,
+      "step": 3455
+    },
+    {
+      "epoch": 0.1293548613220553,
+      "grad_norm": 0.5428165197372437,
+      "learning_rate": 1.9185569140977104e-05,
+      "loss": 0.2773,
+      "step": 3460
+    },
+    {
+      "epoch": 0.12954179031240506,
+      "grad_norm": 0.24074698984622955,
+      "learning_rate": 1.9183246194412583e-05,
+      "loss": 0.2858,
+      "step": 3465
+    },
+    {
+      "epoch": 0.12972871930275487,
+      "grad_norm": 0.5543522834777832,
+      "learning_rate": 1.9180920080809575e-05,
+      "loss": 0.2618,
+      "step": 3470
+    },
+    {
+      "epoch": 0.12991564829310465,
+      "grad_norm": 0.15020422637462616,
+      "learning_rate": 1.9178590800970302e-05,
+      "loss": 0.3799,
+      "step": 3475
+    },
+    {
+      "epoch": 0.13010257728345445,
+      "grad_norm": 0.5467866659164429,
+      "learning_rate": 1.9176258355698062e-05,
+      "loss": 0.335,
+      "step": 3480
+    },
+    {
+      "epoch": 0.13028950627380423,
+      "grad_norm": 0.3544618487358093,
+      "learning_rate": 1.917392274579725e-05,
+      "loss": 0.322,
+      "step": 3485
+    },
+    {
+      "epoch": 0.13047643526415403,
+      "grad_norm": 0.32355251908302307,
+      "learning_rate": 1.9171583972073345e-05,
+      "loss": 0.2544,
+      "step": 3490
+    },
+    {
+      "epoch": 0.1306633642545038,
+      "grad_norm": 0.3455124795436859,
+      "learning_rate": 1.916924203533293e-05,
+      "loss": 0.2684,
+      "step": 3495
+    },
+    {
+      "epoch": 0.13085029324485362,
+      "grad_norm": 0.5765972137451172,
+      "learning_rate": 1.9166896936383668e-05,
+      "loss": 0.3054,
+      "step": 3500
+    },
+    {
+      "epoch": 0.1310372222352034,
+      "grad_norm": 0.5195266008377075,
+      "learning_rate": 1.9164548676034312e-05,
+      "loss": 0.3135,
+      "step": 3505
+    },
+    {
+      "epoch": 0.1312241512255532,
+      "grad_norm": 0.35077986121177673,
+      "learning_rate": 1.9162197255094722e-05,
+      "loss": 0.3495,
+      "step": 3510
+    },
+    {
+      "epoch": 0.13141108021590298,
+      "grad_norm": 0.34569740295410156,
+      "learning_rate": 1.915984267437583e-05,
+      "loss": 0.2855,
+      "step": 3515
+    },
+    {
+      "epoch": 0.13159800920625278,
+      "grad_norm": 0.36829873919487,
+      "learning_rate": 1.915748493468966e-05,
+      "loss": 0.3321,
+      "step": 3520
+    },
+    {
+      "epoch": 0.13178493819660256,
+      "grad_norm": 0.5086042881011963,
+      "learning_rate": 1.915512403684933e-05,
+      "loss": 0.3072,
+      "step": 3525
+    },
+    {
+      "epoch": 0.13197186718695236,
+      "grad_norm": 0.3874140679836273,
+      "learning_rate": 1.9152759981669046e-05,
+      "loss": 0.2989,
+      "step": 3530
+    },
+    {
+      "epoch": 0.13215879617730214,
+      "grad_norm": 0.4253312945365906,
+      "learning_rate": 1.9150392769964106e-05,
+      "loss": 0.3197,
+      "step": 3535
+    },
+    {
+      "epoch": 0.13234572516765195,
+      "grad_norm": 0.309948205947876,
+      "learning_rate": 1.914802240255089e-05,
+      "loss": 0.3147,
+      "step": 3540
+    },
+    {
+      "epoch": 0.13253265415800172,
+      "grad_norm": 0.38400158286094666,
+      "learning_rate": 1.9145648880246877e-05,
+      "loss": 0.3448,
+      "step": 3545
+    },
+    {
+      "epoch": 0.13271958314835153,
+      "grad_norm": 0.2646450996398926,
+      "learning_rate": 1.914327220387062e-05,
+      "loss": 0.4444,
+      "step": 3550
+    },
+    {
+      "epoch": 0.1329065121387013,
+      "grad_norm": 0.4903238117694855,
+      "learning_rate": 1.914089237424176e-05,
+      "loss": 0.3533,
+      "step": 3555
+    },
+    {
+      "epoch": 0.1330934411290511,
+      "grad_norm": 0.48070570826530457,
+      "learning_rate": 1.9138509392181047e-05,
+      "loss": 0.2815,
+      "step": 3560
+    },
+    {
+      "epoch": 0.1332803701194009,
+      "grad_norm": 0.3975090980529785,
+      "learning_rate": 1.9136123258510292e-05,
+      "loss": 0.3287,
+      "step": 3565
+    },
+    {
+      "epoch": 0.1334672991097507,
+      "grad_norm": 0.3869096040725708,
+      "learning_rate": 1.9133733974052412e-05,
+      "loss": 0.3055,
+      "step": 3570
+    },
+    {
+      "epoch": 0.13365422810010047,
+      "grad_norm": 0.26971435546875,
+      "learning_rate": 1.9131341539631395e-05,
+      "loss": 0.3044,
+      "step": 3575
+    },
+    {
+      "epoch": 0.13384115709045027,
+      "grad_norm": 0.5639915466308594,
+      "learning_rate": 1.912894595607233e-05,
+      "loss": 0.2953,
+      "step": 3580
+    },
+    {
+      "epoch": 0.13402808608080005,
+      "grad_norm": 0.463853120803833,
+      "learning_rate": 1.912654722420138e-05,
+      "loss": 0.2509,
+      "step": 3585
+    },
+    {
+      "epoch": 0.13421501507114986,
+      "grad_norm": 0.38717523217201233,
+      "learning_rate": 1.9124145344845804e-05,
+      "loss": 0.3899,
+      "step": 3590
+    },
+    {
+      "epoch": 0.13440194406149963,
+      "grad_norm": 0.2911614179611206,
+      "learning_rate": 1.9121740318833938e-05,
+      "loss": 0.2567,
+      "step": 3595
+    },
+    {
+      "epoch": 0.13458887305184944,
+      "grad_norm": 0.5385882258415222,
+      "learning_rate": 1.9119332146995205e-05,
+      "loss": 0.265,
+      "step": 3600
+    },
+    {
+      "epoch": 0.13477580204219922,
+      "grad_norm": 0.3904414176940918,
+      "learning_rate": 1.9116920830160117e-05,
+      "loss": 0.251,
+      "step": 3605
+    },
+    {
+      "epoch": 0.13496273103254902,
+      "grad_norm": 0.41085904836654663,
+      "learning_rate": 1.9114506369160267e-05,
+      "loss": 0.3327,
+      "step": 3610
+    },
+    {
+      "epoch": 0.1351496600228988,
+      "grad_norm": 0.6957544088363647,
+      "learning_rate": 1.9112088764828335e-05,
+      "loss": 0.3716,
+      "step": 3615
+    },
+    {
+      "epoch": 0.1353365890132486,
+      "grad_norm": 0.4141400456428528,
+      "learning_rate": 1.910966801799808e-05,
+      "loss": 0.2635,
+      "step": 3620
+    },
+    {
+      "epoch": 0.13552351800359838,
+      "grad_norm": 0.773308277130127,
+      "learning_rate": 1.910724412950435e-05,
+      "loss": 0.2762,
+      "step": 3625
+    },
+    {
+      "epoch": 0.13571044699394819,
+      "grad_norm": 0.4335046708583832,
+      "learning_rate": 1.910481710018308e-05,
+      "loss": 0.3463,
+      "step": 3630
+    },
+    {
+      "epoch": 0.13589737598429796,
+      "grad_norm": 0.6109732985496521,
+      "learning_rate": 1.9102386930871276e-05,
+      "loss": 0.3244,
+      "step": 3635
+    },
+    {
+      "epoch": 0.13608430497464777,
+      "grad_norm": 0.3992200791835785,
+      "learning_rate": 1.9099953622407038e-05,
+      "loss": 0.4066,
+      "step": 3640
+    },
+    {
+      "epoch": 0.13627123396499755,
+      "grad_norm": 0.2817442715167999,
+      "learning_rate": 1.9097517175629535e-05,
+      "loss": 0.382,
+      "step": 3645
+    },
+    {
+      "epoch": 0.13645816295534735,
+      "grad_norm": 0.6066967248916626,
+      "learning_rate": 1.9095077591379044e-05,
+      "loss": 0.3066,
+      "step": 3650
+    },
+    {
+      "epoch": 0.13664509194569713,
+      "grad_norm": 0.638088047504425,
+      "learning_rate": 1.9092634870496892e-05,
+      "loss": 0.341,
+      "step": 3655
+    },
+    {
+      "epoch": 0.1368320209360469,
+      "grad_norm": 0.5067809224128723,
+      "learning_rate": 1.9090189013825515e-05,
+      "loss": 0.3039,
+      "step": 3660
+    },
+    {
+      "epoch": 0.1370189499263967,
+      "grad_norm": 0.3270426392555237,
+      "learning_rate": 1.908774002220841e-05,
+      "loss": 0.3165,
+      "step": 3665
+    },
+    {
+      "epoch": 0.1372058789167465,
+      "grad_norm": 0.46038034558296204,
+      "learning_rate": 1.908528789649017e-05,
+      "loss": 0.3287,
+      "step": 3670
+    },
+    {
+      "epoch": 0.1373928079070963,
+      "grad_norm": 0.35730430483818054,
+      "learning_rate": 1.9082832637516458e-05,
+      "loss": 0.2788,
+      "step": 3675
+    },
+    {
+      "epoch": 0.13757973689744607,
+      "grad_norm": 0.4533166289329529,
+      "learning_rate": 1.908037424613403e-05,
+      "loss": 0.3384,
+      "step": 3680
+    },
+    {
+      "epoch": 0.13776666588779587,
+      "grad_norm": 0.6102305054664612,
+      "learning_rate": 1.907791272319071e-05,
+      "loss": 0.3651,
+      "step": 3685
+    },
+    {
+      "epoch": 0.13795359487814565,
+      "grad_norm": 0.6289255023002625,
+      "learning_rate": 1.9075448069535406e-05,
+      "loss": 0.311,
+      "step": 3690
+    },
+    {
+      "epoch": 0.13814052386849546,
+      "grad_norm": 0.44820600748062134,
+      "learning_rate": 1.9072980286018104e-05,
+      "loss": 0.3194,
+      "step": 3695
+    },
+    {
+      "epoch": 0.13832745285884523,
+      "grad_norm": 0.36444956064224243,
+      "learning_rate": 1.907050937348988e-05,
+      "loss": 0.293,
+      "step": 3700
+    },
+    {
+      "epoch": 0.13851438184919504,
+      "grad_norm": 0.3149167001247406,
+      "learning_rate": 1.9068035332802874e-05,
+      "loss": 0.3288,
+      "step": 3705
+    },
+    {
+      "epoch": 0.13870131083954482,
+      "grad_norm": 0.4111466705799103,
+      "learning_rate": 1.9065558164810312e-05,
+      "loss": 0.2532,
+      "step": 3710
+    },
+    {
+      "epoch": 0.13888823982989462,
+      "grad_norm": 0.37402135133743286,
+      "learning_rate": 1.9063077870366504e-05,
+      "loss": 0.3662,
+      "step": 3715
+    },
+    {
+      "epoch": 0.1390751688202444,
+      "grad_norm": 0.706342875957489,
+      "learning_rate": 1.9060594450326824e-05,
+      "loss": 0.2887,
+      "step": 3720
+    },
+    {
+      "epoch": 0.1392620978105942,
+      "grad_norm": 0.5995466709136963,
+      "learning_rate": 1.9058107905547737e-05,
+      "loss": 0.3387,
+      "step": 3725
+    },
+    {
+      "epoch": 0.13944902680094398,
+      "grad_norm": 0.16998915374279022,
+      "learning_rate": 1.9055618236886784e-05,
+      "loss": 0.3133,
+      "step": 3730
+    },
+    {
+      "epoch": 0.1396359557912938,
+      "grad_norm": 0.3938615024089813,
+      "learning_rate": 1.9053125445202574e-05,
+      "loss": 0.3253,
+      "step": 3735
+    },
+    {
+      "epoch": 0.13982288478164356,
+      "grad_norm": 0.7660725712776184,
+      "learning_rate": 1.9050629531354806e-05,
+      "loss": 0.3604,
+      "step": 3740
+    },
+    {
+      "epoch": 0.14000981377199337,
+      "grad_norm": 0.4029218256473541,
+      "learning_rate": 1.9048130496204247e-05,
+      "loss": 0.3705,
+      "step": 3745
+    },
+    {
+      "epoch": 0.14019674276234315,
+      "grad_norm": 0.5488607883453369,
+      "learning_rate": 1.9045628340612737e-05,
+      "loss": 0.2867,
+      "step": 3750
+    },
+    {
+      "epoch": 0.14038367175269295,
+      "grad_norm": 0.38521480560302734,
+      "learning_rate": 1.904312306544321e-05,
+      "loss": 0.2713,
+      "step": 3755
+    },
+    {
+      "epoch": 0.14057060074304273,
+      "grad_norm": 0.23381686210632324,
+      "learning_rate": 1.9040614671559647e-05,
+      "loss": 0.3445,
+      "step": 3760
+    },
+    {
+      "epoch": 0.14075752973339253,
+      "grad_norm": 0.6382274031639099,
+      "learning_rate": 1.9038103159827136e-05,
+      "loss": 0.3306,
+      "step": 3765
+    },
+    {
+      "epoch": 0.1409444587237423,
+      "grad_norm": 0.17501187324523926,
+      "learning_rate": 1.9035588531111818e-05,
+      "loss": 0.2653,
+      "step": 3770
+    },
+    {
+      "epoch": 0.14113138771409212,
+      "grad_norm": 0.4778057038784027,
+      "learning_rate": 1.903307078628092e-05,
+      "loss": 0.3657,
+      "step": 3775
+    },
+    {
+      "epoch": 0.1413183167044419,
+      "grad_norm": 0.22998036444187164,
+      "learning_rate": 1.9030549926202732e-05,
+      "loss": 0.2792,
+      "step": 3780
+    },
+    {
+      "epoch": 0.1415052456947917,
+      "grad_norm": 0.35750317573547363,
+      "learning_rate": 1.902802595174664e-05,
+      "loss": 0.2231,
+      "step": 3785
+    },
+    {
+      "epoch": 0.14169217468514148,
+      "grad_norm": 0.5821340680122375,
+      "learning_rate": 1.902549886378308e-05,
+      "loss": 0.4172,
+      "step": 3790
+    },
+    {
+      "epoch": 0.14187910367549128,
+      "grad_norm": 0.6900777816772461,
+      "learning_rate": 1.902296866318357e-05,
+      "loss": 0.3648,
+      "step": 3795
+    },
+    {
+      "epoch": 0.14206603266584106,
+      "grad_norm": 0.4445832669734955,
+      "learning_rate": 1.9020435350820715e-05,
+      "loss": 0.2523,
+      "step": 3800
+    },
+    {
+      "epoch": 0.14225296165619086,
+      "grad_norm": 0.43811532855033875,
+      "learning_rate": 1.9017898927568173e-05,
+      "loss": 0.3028,
+      "step": 3805
+    },
+    {
+      "epoch": 0.14243989064654064,
+      "grad_norm": 0.4545620381832123,
+      "learning_rate": 1.9015359394300686e-05,
+      "loss": 0.3506,
+      "step": 3810
+    },
+    {
+      "epoch": 0.14262681963689045,
+      "grad_norm": 0.401100754737854,
+      "learning_rate": 1.901281675189407e-05,
+      "loss": 0.3335,
+      "step": 3815
+    },
+    {
+      "epoch": 0.14281374862724022,
+      "grad_norm": 0.9869058132171631,
+      "learning_rate": 1.9010271001225203e-05,
+      "loss": 0.3321,
+      "step": 3820
+    },
+    {
+      "epoch": 0.14300067761759003,
+      "grad_norm": 0.5079956650733948,
+      "learning_rate": 1.9007722143172046e-05,
+      "loss": 0.2671,
+      "step": 3825
+    },
+    {
+      "epoch": 0.1431876066079398,
+      "grad_norm": 0.5241971015930176,
+      "learning_rate": 1.9005170178613624e-05,
+      "loss": 0.2777,
+      "step": 3830
+    },
+    {
+      "epoch": 0.1433745355982896,
+      "grad_norm": 0.39962106943130493,
+      "learning_rate": 1.900261510843004e-05,
+      "loss": 0.3102,
+      "step": 3835
+    },
+    {
+      "epoch": 0.1435614645886394,
+      "grad_norm": 0.4119807183742523,
+      "learning_rate": 1.9000056933502466e-05,
+      "loss": 0.3591,
+      "step": 3840
+    },
+    {
+      "epoch": 0.1437483935789892,
+      "grad_norm": 0.4325822591781616,
+      "learning_rate": 1.8997495654713133e-05,
+      "loss": 0.3438,
+      "step": 3845
+    },
+    {
+      "epoch": 0.14393532256933897,
+      "grad_norm": 0.2697000503540039,
+      "learning_rate": 1.8994931272945364e-05,
+      "loss": 0.3043,
+      "step": 3850
+    },
+    {
+      "epoch": 0.14412225155968877,
+      "grad_norm": 0.4875340461730957,
+      "learning_rate": 1.8992363789083534e-05,
+      "loss": 0.2718,
+      "step": 3855
+    },
+    {
+      "epoch": 0.14430918055003855,
+      "grad_norm": 0.2768021821975708,
+      "learning_rate": 1.89897932040131e-05,
+      "loss": 0.2831,
+      "step": 3860
+    },
+    {
+      "epoch": 0.14449610954038836,
+      "grad_norm": 0.3775290548801422,
+      "learning_rate": 1.8987219518620573e-05,
+      "loss": 0.3754,
+      "step": 3865
+    },
+    {
+      "epoch": 0.14468303853073813,
+      "grad_norm": 0.4714037775993347,
+      "learning_rate": 1.8984642733793556e-05,
+      "loss": 0.3069,
+      "step": 3870
+    },
+    {
+      "epoch": 0.14486996752108794,
+      "grad_norm": 0.3473433554172516,
+      "learning_rate": 1.8982062850420705e-05,
+      "loss": 0.2979,
+      "step": 3875
+    },
+    {
+      "epoch": 0.14505689651143772,
+      "grad_norm": 0.4559517204761505,
+      "learning_rate": 1.897947986939174e-05,
+      "loss": 0.3666,
+      "step": 3880
+    },
+    {
+      "epoch": 0.14524382550178752,
+      "grad_norm": 0.5417743921279907,
+      "learning_rate": 1.8976893791597465e-05,
+      "loss": 0.3124,
+      "step": 3885
+    },
+    {
+      "epoch": 0.1454307544921373,
+      "grad_norm": 0.2836984097957611,
+      "learning_rate": 1.8974304617929746e-05,
+      "loss": 0.2611,
+      "step": 3890
+    },
+    {
+      "epoch": 0.1456176834824871,
+      "grad_norm": 0.3992460370063782,
+      "learning_rate": 1.8971712349281506e-05,
+      "loss": 0.3606,
+      "step": 3895
+    },
+    {
+      "epoch": 0.14580461247283688,
+      "grad_norm": 0.40167370438575745,
+      "learning_rate": 1.896911698654675e-05,
+      "loss": 0.3131,
+      "step": 3900
+    },
+    {
+      "epoch": 0.14599154146318666,
+      "grad_norm": 0.45140784978866577,
+      "learning_rate": 1.8966518530620542e-05,
+      "loss": 0.3149,
+      "step": 3905
+    },
+    {
+      "epoch": 0.14617847045353646,
+      "grad_norm": 0.5974220633506775,
+      "learning_rate": 1.8963916982399014e-05,
+      "loss": 0.2877,
+      "step": 3910
+    },
+    {
+      "epoch": 0.14636539944388624,
+      "grad_norm": 0.2815289795398712,
+      "learning_rate": 1.8961312342779374e-05,
+      "loss": 0.3502,
+      "step": 3915
+    },
+    {
+      "epoch": 0.14655232843423605,
+      "grad_norm": 0.3491719365119934,
+      "learning_rate": 1.8958704612659876e-05,
+      "loss": 0.2682,
+      "step": 3920
+    },
+    {
+      "epoch": 0.14673925742458582,
+      "grad_norm": 0.28988122940063477,
+      "learning_rate": 1.8956093792939855e-05,
+      "loss": 0.369,
+      "step": 3925
+    },
+    {
+      "epoch": 0.14692618641493563,
+      "grad_norm": 0.6046833992004395,
+      "learning_rate": 1.895347988451971e-05,
+      "loss": 0.4345,
+      "step": 3930
+    },
+    {
+      "epoch": 0.1471131154052854,
+      "grad_norm": 0.5748773813247681,
+      "learning_rate": 1.895086288830091e-05,
+      "loss": 0.3157,
+      "step": 3935
+    },
+    {
+      "epoch": 0.1473000443956352,
+      "grad_norm": 2.4972338676452637,
+      "learning_rate": 1.8948242805185966e-05,
+      "loss": 0.4136,
+      "step": 3940
+    },
+    {
+      "epoch": 0.147486973385985,
+      "grad_norm": 0.41845616698265076,
+      "learning_rate": 1.8945619636078483e-05,
+      "loss": 0.2917,
+      "step": 3945
+    },
+    {
+      "epoch": 0.1476739023763348,
+      "grad_norm": 0.42868664860725403,
+      "learning_rate": 1.894299338188311e-05,
+      "loss": 0.2979,
+      "step": 3950
+    },
+    {
+      "epoch": 0.14786083136668457,
+      "grad_norm": 0.5431304574012756,
+      "learning_rate": 1.8940364043505568e-05,
+      "loss": 0.3204,
+      "step": 3955
+    },
+    {
+      "epoch": 0.14804776035703437,
+      "grad_norm": 0.40625184774398804,
+      "learning_rate": 1.893773162185264e-05,
+      "loss": 0.2749,
+      "step": 3960
+    },
+    {
+      "epoch": 0.14823468934738415,
+      "grad_norm": 0.4661557078361511,
+      "learning_rate": 1.893509611783218e-05,
+      "loss": 0.326,
+      "step": 3965
+    },
+    {
+      "epoch": 0.14842161833773396,
+      "grad_norm": 0.27836883068084717,
+      "learning_rate": 1.8932457532353087e-05,
+      "loss": 0.2812,
+      "step": 3970
+    },
+    {
+      "epoch": 0.14860854732808373,
+      "grad_norm": 0.3862547278404236,
+      "learning_rate": 1.892981586632534e-05,
+      "loss": 0.3615,
+      "step": 3975
+    },
+    {
+      "epoch": 0.14879547631843354,
+      "grad_norm": 0.46969127655029297,
+      "learning_rate": 1.892717112065997e-05,
+      "loss": 0.3065,
+      "step": 3980
+    },
+    {
+      "epoch": 0.14898240530878332,
+      "grad_norm": 0.4676092565059662,
+      "learning_rate": 1.8924523296269077e-05,
+      "loss": 0.3214,
+      "step": 3985
+    },
+    {
+      "epoch": 0.14916933429913312,
+      "grad_norm": 0.5357167720794678,
+      "learning_rate": 1.8921872394065822e-05,
+      "loss": 0.3075,
+      "step": 3990
+    },
+    {
+      "epoch": 0.1493562632894829,
+      "grad_norm": 0.5506564974784851,
+      "learning_rate": 1.891921841496442e-05,
+      "loss": 0.2875,
+      "step": 3995
+    },
+    {
+      "epoch": 0.1495431922798327,
+      "grad_norm": 0.2982819080352783,
+      "learning_rate": 1.8916561359880153e-05,
+      "loss": 0.3099,
+      "step": 4000
+    },
+    {
+      "epoch": 0.14973012127018248,
+      "grad_norm": 1.2449369430541992,
+      "learning_rate": 1.8913901229729367e-05,
+      "loss": 0.385,
+      "step": 4005
+    },
+    {
+      "epoch": 0.1499170502605323,
+      "grad_norm": 0.7360920310020447,
+      "learning_rate": 1.8911238025429464e-05,
+      "loss": 0.3481,
+      "step": 4010
+    },
+    {
+      "epoch": 0.15010397925088206,
+      "grad_norm": 0.37266823649406433,
+      "learning_rate": 1.8908571747898902e-05,
+      "loss": 0.262,
+      "step": 4015
+    },
+    {
+      "epoch": 0.15029090824123187,
+      "grad_norm": 0.5776626467704773,
+      "learning_rate": 1.8905902398057208e-05,
+      "loss": 0.3394,
+      "step": 4020
+    },
+    {
+      "epoch": 0.15047783723158165,
+      "grad_norm": 0.6980562210083008,
+      "learning_rate": 1.8903229976824963e-05,
+      "loss": 0.3497,
+      "step": 4025
+    },
+    {
+      "epoch": 0.15066476622193145,
+      "grad_norm": 0.46382442116737366,
+      "learning_rate": 1.890055448512381e-05,
+      "loss": 0.2935,
+      "step": 4030
+    },
+    {
+      "epoch": 0.15085169521228123,
+      "grad_norm": 0.4823243319988251,
+      "learning_rate": 1.889787592387645e-05,
+      "loss": 0.2739,
+      "step": 4035
+    },
+    {
+      "epoch": 0.15103862420263103,
+      "grad_norm": 0.4817698895931244,
+      "learning_rate": 1.8895194294006635e-05,
+      "loss": 0.2478,
+      "step": 4040
+    },
+    {
+      "epoch": 0.1512255531929808,
+      "grad_norm": 0.6213811039924622,
+      "learning_rate": 1.8892509596439192e-05,
+      "loss": 0.3307,
+      "step": 4045
+    },
+    {
+      "epoch": 0.15141248218333062,
+      "grad_norm": 0.4142370820045471,
+      "learning_rate": 1.8889821832099988e-05,
+      "loss": 0.3091,
+      "step": 4050
+    },
+    {
+      "epoch": 0.1515994111736804,
+      "grad_norm": 0.5155577659606934,
+      "learning_rate": 1.8887131001915964e-05,
+      "loss": 0.3226,
+      "step": 4055
+    },
+    {
+      "epoch": 0.1517863401640302,
+      "grad_norm": 0.6737167239189148,
+      "learning_rate": 1.8884437106815103e-05,
+      "loss": 0.3247,
+      "step": 4060
+    },
+    {
+      "epoch": 0.15197326915437998,
+      "grad_norm": 0.2817344665527344,
+      "learning_rate": 1.8881740147726458e-05,
+      "loss": 0.3737,
+      "step": 4065
+    },
+    {
+      "epoch": 0.15216019814472978,
+      "grad_norm": 0.3459798991680145,
+      "learning_rate": 1.887904012558013e-05,
+      "loss": 0.3037,
+      "step": 4070
+    },
+    {
+      "epoch": 0.15234712713507956,
+      "grad_norm": 0.34796640276908875,
+      "learning_rate": 1.8876337041307275e-05,
+      "loss": 0.2671,
+      "step": 4075
+    },
+    {
+      "epoch": 0.15253405612542936,
+      "grad_norm": 0.6775915622711182,
+      "learning_rate": 1.8873630895840114e-05,
+      "loss": 0.3625,
+      "step": 4080
+    },
+    {
+      "epoch": 0.15272098511577914,
+      "grad_norm": 0.35140863060951233,
+      "learning_rate": 1.887092169011192e-05,
+      "loss": 0.2942,
+      "step": 4085
+    },
+    {
+      "epoch": 0.15290791410612894,
+      "grad_norm": 0.8965099453926086,
+      "learning_rate": 1.8868209425057025e-05,
+      "loss": 0.3787,
+      "step": 4090
+    },
+    {
+      "epoch": 0.15309484309647872,
+      "grad_norm": 0.4382193684577942,
+      "learning_rate": 1.88654941016108e-05,
+      "loss": 0.3455,
+      "step": 4095
+    },
+    {
+      "epoch": 0.15328177208682853,
+      "grad_norm": 0.4126281142234802,
+      "learning_rate": 1.8862775720709686e-05,
+      "loss": 0.2807,
+      "step": 4100
+    },
+    {
+      "epoch": 0.1534687010771783,
+      "grad_norm": 0.31981319189071655,
+      "learning_rate": 1.886005428329118e-05,
+      "loss": 0.3167,
+      "step": 4105
+    },
+    {
+      "epoch": 0.1536556300675281,
+      "grad_norm": 0.41247910261154175,
+      "learning_rate": 1.8857329790293824e-05,
+      "loss": 0.2936,
+      "step": 4110
+    },
+    {
+      "epoch": 0.1538425590578779,
+      "grad_norm": 0.48248907923698425,
+      "learning_rate": 1.885460224265722e-05,
+      "loss": 0.3004,
+      "step": 4115
+    },
+    {
+      "epoch": 0.1540294880482277,
+      "grad_norm": 0.578687846660614,
+      "learning_rate": 1.8851871641322016e-05,
+      "loss": 0.28,
+      "step": 4120
+    },
+    {
+      "epoch": 0.15421641703857747,
+      "grad_norm": 0.20185329020023346,
+      "learning_rate": 1.884913798722992e-05,
+      "loss": 0.334,
+      "step": 4125
+    },
+    {
+      "epoch": 0.15440334602892727,
+      "grad_norm": 0.5341799259185791,
+      "learning_rate": 1.8846401281323693e-05,
+      "loss": 0.3087,
+      "step": 4130
+    },
+    {
+      "epoch": 0.15459027501927705,
+      "grad_norm": 0.2198217362165451,
+      "learning_rate": 1.884366152454715e-05,
+      "loss": 0.2779,
+      "step": 4135
+    },
+    {
+      "epoch": 0.15477720400962686,
+      "grad_norm": 0.4372238218784332,
+      "learning_rate": 1.8840918717845146e-05,
+      "loss": 0.3726,
+      "step": 4140
+    },
+    {
+      "epoch": 0.15496413299997663,
+      "grad_norm": 0.48354169726371765,
+      "learning_rate": 1.88381728621636e-05,
+      "loss": 0.3891,
+      "step": 4145
+    },
+    {
+      "epoch": 0.1551510619903264,
+      "grad_norm": 0.5194223523139954,
+      "learning_rate": 1.883542395844948e-05,
+      "loss": 0.3032,
+      "step": 4150
+    },
+    {
+      "epoch": 0.15533799098067622,
+      "grad_norm": 0.23291267454624176,
+      "learning_rate": 1.8832672007650805e-05,
+      "loss": 0.2777,
+      "step": 4155
+    },
+    {
+      "epoch": 0.155524919971026,
+      "grad_norm": 0.8497887253761292,
+      "learning_rate": 1.882991701071664e-05,
+      "loss": 0.3094,
+      "step": 4160
+    },
+    {
+      "epoch": 0.1557118489613758,
+      "grad_norm": 0.4399624466896057,
+      "learning_rate": 1.8827158968597113e-05,
+      "loss": 0.3451,
+      "step": 4165
+    },
+    {
+      "epoch": 0.15589877795172558,
+      "grad_norm": 0.3427572250366211,
+      "learning_rate": 1.8824397882243382e-05,
+      "loss": 0.2513,
+      "step": 4170
+    },
+    {
+      "epoch": 0.15608570694207538,
+      "grad_norm": 0.5564326643943787,
+      "learning_rate": 1.8821633752607672e-05,
+      "loss": 0.3496,
+      "step": 4175
+    },
+    {
+      "epoch": 0.15627263593242516,
+      "grad_norm": 0.4042254090309143,
+      "learning_rate": 1.8818866580643254e-05,
+      "loss": 0.3207,
+      "step": 4180
+    },
+    {
+      "epoch": 0.15645956492277496,
+      "grad_norm": 0.25726428627967834,
+      "learning_rate": 1.8816096367304447e-05,
+      "loss": 0.2839,
+      "step": 4185
+    },
+    {
+      "epoch": 0.15664649391312474,
+      "grad_norm": 0.6856520175933838,
+      "learning_rate": 1.8813323113546614e-05,
+      "loss": 0.3228,
+      "step": 4190
+    },
+    {
+      "epoch": 0.15683342290347455,
+      "grad_norm": 0.27812889218330383,
+      "learning_rate": 1.8810546820326173e-05,
+      "loss": 0.2724,
+      "step": 4195
+    },
+    {
+      "epoch": 0.15702035189382432,
+      "grad_norm": 0.48270055651664734,
+      "learning_rate": 1.880776748860059e-05,
+      "loss": 0.4022,
+      "step": 4200
+    },
+    {
+      "epoch": 0.15720728088417413,
+      "grad_norm": 0.8398357033729553,
+      "learning_rate": 1.8804985119328375e-05,
+      "loss": 0.3086,
+      "step": 4205
+    },
+    {
+      "epoch": 0.1573942098745239,
+      "grad_norm": 0.34941208362579346,
+      "learning_rate": 1.8802199713469084e-05,
+      "loss": 0.3078,
+      "step": 4210
+    },
+    {
+      "epoch": 0.1575811388648737,
+      "grad_norm": 0.3891669511795044,
+      "learning_rate": 1.8799411271983325e-05,
+      "loss": 0.2914,
+      "step": 4215
+    },
+    {
+      "epoch": 0.1577680678552235,
+      "grad_norm": 0.4491298794746399,
+      "learning_rate": 1.8796619795832758e-05,
+      "loss": 0.3676,
+      "step": 4220
+    },
+    {
+      "epoch": 0.1579549968455733,
+      "grad_norm": 0.3361729383468628,
+      "learning_rate": 1.8793825285980076e-05,
+      "loss": 0.3815,
+      "step": 4225
+    },
+    {
+      "epoch": 0.15814192583592307,
+      "grad_norm": 0.3507533371448517,
+      "learning_rate": 1.879102774338903e-05,
+      "loss": 0.3435,
+      "step": 4230
+    },
+    {
+      "epoch": 0.15832885482627287,
+      "grad_norm": 0.44350340962409973,
+      "learning_rate": 1.8788227169024406e-05,
+      "loss": 0.2433,
+      "step": 4235
+    },
+    {
+      "epoch": 0.15851578381662265,
+      "grad_norm": 0.4336090683937073,
+      "learning_rate": 1.878542356385205e-05,
+      "loss": 0.3624,
+      "step": 4240
+    },
+    {
+      "epoch": 0.15870271280697246,
+      "grad_norm": 0.5720579028129578,
+      "learning_rate": 1.878261692883884e-05,
+      "loss": 0.3331,
+      "step": 4245
+    },
+    {
+      "epoch": 0.15888964179732223,
+      "grad_norm": 0.5189465880393982,
+      "learning_rate": 1.8779807264952704e-05,
+      "loss": 0.3359,
+      "step": 4250
+    },
+    {
+      "epoch": 0.15907657078767204,
+      "grad_norm": 0.2319631725549698,
+      "learning_rate": 1.8776994573162615e-05,
+      "loss": 0.3025,
+      "step": 4255
+    },
+    {
+      "epoch": 0.15926349977802182,
+      "grad_norm": 0.25383996963500977,
+      "learning_rate": 1.877417885443859e-05,
+      "loss": 0.332,
+      "step": 4260
+    },
+    {
+      "epoch": 0.15945042876837162,
+      "grad_norm": 0.4644213020801544,
+      "learning_rate": 1.8771360109751694e-05,
+      "loss": 0.4095,
+      "step": 4265
+    },
+    {
+      "epoch": 0.1596373577587214,
+      "grad_norm": 0.37912002205848694,
+      "learning_rate": 1.8768538340074024e-05,
+      "loss": 0.3924,
+      "step": 4270
+    },
+    {
+      "epoch": 0.1598242867490712,
+      "grad_norm": 0.3221535086631775,
+      "learning_rate": 1.8765713546378733e-05,
+      "loss": 0.3375,
+      "step": 4275
+    },
+    {
+      "epoch": 0.16001121573942098,
+      "grad_norm": 0.3765072822570801,
+      "learning_rate": 1.8762885729640007e-05,
+      "loss": 0.3454,
+      "step": 4280
+    },
+    {
+      "epoch": 0.16019814472977079,
+      "grad_norm": 0.43610912561416626,
+      "learning_rate": 1.8760054890833083e-05,
+      "loss": 0.2819,
+      "step": 4285
+    },
+    {
+      "epoch": 0.16038507372012056,
+      "grad_norm": 0.5509352684020996,
+      "learning_rate": 1.8757221030934234e-05,
+      "loss": 0.3796,
+      "step": 4290
+    },
+    {
+      "epoch": 0.16057200271047037,
+      "grad_norm": 0.43372535705566406,
+      "learning_rate": 1.8754384150920777e-05,
+      "loss": 0.3896,
+      "step": 4295
+    },
+    {
+      "epoch": 0.16075893170082015,
+      "grad_norm": 0.35043320059776306,
+      "learning_rate": 1.8751544251771072e-05,
+      "loss": 0.3586,
+      "step": 4300
+    },
+    {
+      "epoch": 0.16094586069116995,
+      "grad_norm": 0.6886361837387085,
+      "learning_rate": 1.874870133446452e-05,
+      "loss": 0.3648,
+      "step": 4305
+    },
+    {
+      "epoch": 0.16113278968151973,
+      "grad_norm": 0.6359232068061829,
+      "learning_rate": 1.8745855399981555e-05,
+      "loss": 0.2569,
+      "step": 4310
+    },
+    {
+      "epoch": 0.16131971867186953,
+      "grad_norm": 0.4465806782245636,
+      "learning_rate": 1.8743006449303663e-05,
+      "loss": 0.3305,
+      "step": 4315
+    },
+    {
+      "epoch": 0.1615066476622193,
+      "grad_norm": 0.6533617973327637,
+      "learning_rate": 1.874015448341337e-05,
+      "loss": 0.3634,
+      "step": 4320
+    },
+    {
+      "epoch": 0.16169357665256912,
+      "grad_norm": 0.4724544286727905,
+      "learning_rate": 1.8737299503294233e-05,
+      "loss": 0.2838,
+      "step": 4325
+    },
+    {
+      "epoch": 0.1618805056429189,
+      "grad_norm": 0.4181117117404938,
+      "learning_rate": 1.873444150993085e-05,
+      "loss": 0.2875,
+      "step": 4330
+    },
+    {
+      "epoch": 0.1620674346332687,
+      "grad_norm": 0.2569214105606079,
+      "learning_rate": 1.8731580504308865e-05,
+      "loss": 0.4176,
+      "step": 4335
+    },
+    {
+      "epoch": 0.16225436362361847,
+      "grad_norm": 1.067435622215271,
+      "learning_rate": 1.872871648741495e-05,
+      "loss": 0.3486,
+      "step": 4340
+    },
+    {
+      "epoch": 0.16244129261396828,
+      "grad_norm": 0.20301935076713562,
+      "learning_rate": 1.8725849460236833e-05,
+      "loss": 0.2639,
+      "step": 4345
+    },
+    {
+      "epoch": 0.16262822160431806,
+      "grad_norm": 0.3194063901901245,
+      "learning_rate": 1.8722979423763264e-05,
+      "loss": 0.2772,
+      "step": 4350
+    },
+    {
+      "epoch": 0.16281515059466786,
+      "grad_norm": 0.5217241048812866,
+      "learning_rate": 1.872010637898404e-05,
+      "loss": 0.3767,
+      "step": 4355
+    },
+    {
+      "epoch": 0.16300207958501764,
+      "grad_norm": 0.658311128616333,
+      "learning_rate": 1.8717230326889984e-05,
+      "loss": 0.4492,
+      "step": 4360
+    },
+    {
+      "epoch": 0.16318900857536744,
+      "grad_norm": 0.26515570282936096,
+      "learning_rate": 1.871435126847297e-05,
+      "loss": 0.2921,
+      "step": 4365
+    },
+    {
+      "epoch": 0.16337593756571722,
+      "grad_norm": 0.3176003396511078,
+      "learning_rate": 1.87114692047259e-05,
+      "loss": 0.3403,
+      "step": 4370
+    },
+    {
+      "epoch": 0.16356286655606703,
+      "grad_norm": 0.5091459155082703,
+      "learning_rate": 1.8708584136642717e-05,
+      "loss": 0.3209,
+      "step": 4375
+    },
+    {
+      "epoch": 0.1637497955464168,
+      "grad_norm": 0.35556361079216003,
+      "learning_rate": 1.8705696065218398e-05,
+      "loss": 0.311,
+      "step": 4380
+    },
+    {
+      "epoch": 0.1639367245367666,
+      "grad_norm": 0.43555188179016113,
+      "learning_rate": 1.8702804991448955e-05,
+      "loss": 0.3309,
+      "step": 4385
+    },
+    {
+      "epoch": 0.1641236535271164,
+      "grad_norm": 0.2813642919063568,
+      "learning_rate": 1.8699910916331438e-05,
+      "loss": 0.2802,
+      "step": 4390
+    },
+    {
+      "epoch": 0.16431058251746616,
+      "grad_norm": 0.32983142137527466,
+      "learning_rate": 1.869701384086393e-05,
+      "loss": 0.3307,
+      "step": 4395
+    },
+    {
+      "epoch": 0.16449751150781597,
+      "grad_norm": 0.8605227470397949,
+      "learning_rate": 1.8694113766045552e-05,
+      "loss": 0.2888,
+      "step": 4400
+    },
+    {
+      "epoch": 0.16468444049816575,
+      "grad_norm": 0.4275653660297394,
+      "learning_rate": 1.869121069287645e-05,
+      "loss": 0.3053,
+      "step": 4405
+    },
+    {
+      "epoch": 0.16487136948851555,
+      "grad_norm": 0.32885563373565674,
+      "learning_rate": 1.8688304622357817e-05,
+      "loss": 0.3125,
+      "step": 4410
+    },
+    {
+      "epoch": 0.16505829847886533,
+      "grad_norm": 0.38488566875457764,
+      "learning_rate": 1.868539555549187e-05,
+      "loss": 0.2482,
+      "step": 4415
+    },
+    {
+      "epoch": 0.16524522746921513,
+      "grad_norm": 0.3290596902370453,
+      "learning_rate": 1.8682483493281864e-05,
+      "loss": 0.3019,
+      "step": 4420
+    },
+    {
+      "epoch": 0.1654321564595649,
+      "grad_norm": 0.9788287878036499,
+      "learning_rate": 1.8679568436732084e-05,
+      "loss": 0.3028,
+      "step": 4425
+    },
+    {
+      "epoch": 0.16561908544991472,
+      "grad_norm": 0.728779137134552,
+      "learning_rate": 1.8676650386847855e-05,
+      "loss": 0.3395,
+      "step": 4430
+    },
+    {
+      "epoch": 0.1658060144402645,
+      "grad_norm": 0.40499404072761536,
+      "learning_rate": 1.8673729344635524e-05,
+      "loss": 0.4077,
+      "step": 4435
+    },
+    {
+      "epoch": 0.1659929434306143,
+      "grad_norm": 1.272477626800537,
+      "learning_rate": 1.8670805311102477e-05,
+      "loss": 0.3258,
+      "step": 4440
+    },
+    {
+      "epoch": 0.16617987242096408,
+      "grad_norm": 0.32982587814331055,
+      "learning_rate": 1.8667878287257124e-05,
+      "loss": 0.283,
+      "step": 4445
+    },
+    {
+      "epoch": 0.16636680141131388,
+      "grad_norm": 0.4656912088394165,
+      "learning_rate": 1.8664948274108918e-05,
+      "loss": 0.2622,
+      "step": 4450
+    },
+    {
+      "epoch": 0.16655373040166366,
+      "grad_norm": 0.3891690671443939,
+      "learning_rate": 1.866201527266834e-05,
+      "loss": 0.3216,
+      "step": 4455
+    },
+    {
+      "epoch": 0.16674065939201346,
+      "grad_norm": 0.40880194306373596,
+      "learning_rate": 1.8659079283946882e-05,
+      "loss": 0.2894,
+      "step": 4460
+    },
+    {
+      "epoch": 0.16692758838236324,
+      "grad_norm": 1.5993868112564087,
+      "learning_rate": 1.86561403089571e-05,
+      "loss": 0.3405,
+      "step": 4465
+    },
+    {
+      "epoch": 0.16711451737271305,
+      "grad_norm": 0.5294637084007263,
+      "learning_rate": 1.8653198348712552e-05,
+      "loss": 0.2742,
+      "step": 4470
+    },
+    {
+      "epoch": 0.16730144636306282,
+      "grad_norm": 0.47895947098731995,
+      "learning_rate": 1.865025340422784e-05,
+      "loss": 0.271,
+      "step": 4475
+    },
+    {
+      "epoch": 0.16748837535341263,
+      "grad_norm": 0.17551173269748688,
+      "learning_rate": 1.864730547651859e-05,
+      "loss": 0.2728,
+      "step": 4480
+    },
+    {
+      "epoch": 0.1676753043437624,
+      "grad_norm": 0.5507614612579346,
+      "learning_rate": 1.8644354566601458e-05,
+      "loss": 0.3346,
+      "step": 4485
+    },
+    {
+      "epoch": 0.1678622333341122,
+      "grad_norm": 0.4217040240764618,
+      "learning_rate": 1.864140067549413e-05,
+      "loss": 0.3026,
+      "step": 4490
+    },
+    {
+      "epoch": 0.168049162324462,
+      "grad_norm": 0.31403791904449463,
+      "learning_rate": 1.8638443804215315e-05,
+      "loss": 0.2778,
+      "step": 4495
+    },
+    {
+      "epoch": 0.1682360913148118,
+      "grad_norm": 0.3219631016254425,
+      "learning_rate": 1.8635483953784755e-05,
+      "loss": 0.3311,
+      "step": 4500
+    },
+    {
+      "epoch": 0.16842302030516157,
+      "grad_norm": 0.708217442035675,
+      "learning_rate": 1.8632521125223215e-05,
+      "loss": 0.4032,
+      "step": 4505
+    },
+    {
+      "epoch": 0.16860994929551137,
+      "grad_norm": 1.9042284488677979,
+      "learning_rate": 1.8629555319552492e-05,
+      "loss": 0.3184,
+      "step": 4510
+    },
+    {
+      "epoch": 0.16879687828586115,
+      "grad_norm": 0.42551693320274353,
+      "learning_rate": 1.862658653779541e-05,
+      "loss": 0.2671,
+      "step": 4515
+    },
+    {
+      "epoch": 0.16898380727621096,
+      "grad_norm": 0.22539173066616058,
+      "learning_rate": 1.8623614780975813e-05,
+      "loss": 0.3398,
+      "step": 4520
+    },
+    {
+      "epoch": 0.16917073626656073,
+      "grad_norm": 0.353363037109375,
+      "learning_rate": 1.862064005011858e-05,
+      "loss": 0.2993,
+      "step": 4525
+    },
+    {
+      "epoch": 0.16935766525691054,
+      "grad_norm": 0.25013571977615356,
+      "learning_rate": 1.86176623462496e-05,
+      "loss": 0.2873,
+      "step": 4530
+    },
+    {
+      "epoch": 0.16954459424726032,
+      "grad_norm": 0.28396397829055786,
+      "learning_rate": 1.8614681670395808e-05,
+      "loss": 0.326,
+      "step": 4535
+    },
+    {
+      "epoch": 0.16973152323761012,
+      "grad_norm": 0.33238470554351807,
+      "learning_rate": 1.8611698023585146e-05,
+      "loss": 0.3196,
+      "step": 4540
+    },
+    {
+      "epoch": 0.1699184522279599,
+      "grad_norm": 0.4479984939098358,
+      "learning_rate": 1.8608711406846595e-05,
+      "loss": 0.3363,
+      "step": 4545
+    },
+    {
+      "epoch": 0.1701053812183097,
+      "grad_norm": 0.4897647500038147,
+      "learning_rate": 1.8605721821210146e-05,
+      "loss": 0.3515,
+      "step": 4550
+    },
+    {
+      "epoch": 0.17029231020865948,
+      "grad_norm": 0.33985239267349243,
+      "learning_rate": 1.8602729267706833e-05,
+      "loss": 0.3559,
+      "step": 4555
+    },
+    {
+      "epoch": 0.17047923919900929,
+      "grad_norm": 0.4013058543205261,
+      "learning_rate": 1.859973374736869e-05,
+      "loss": 0.2976,
+      "step": 4560
+    },
+    {
+      "epoch": 0.17066616818935906,
+      "grad_norm": 0.3209231197834015,
+      "learning_rate": 1.8596735261228793e-05,
+      "loss": 0.3168,
+      "step": 4565
+    },
+    {
+      "epoch": 0.17085309717970887,
+      "grad_norm": 0.5259041786193848,
+      "learning_rate": 1.859373381032123e-05,
+      "loss": 0.3052,
+      "step": 4570
+    },
+    {
+      "epoch": 0.17104002617005865,
+      "grad_norm": 0.34932300448417664,
+      "learning_rate": 1.8590729395681122e-05,
+      "loss": 0.3515,
+      "step": 4575
+    },
+    {
+      "epoch": 0.17122695516040845,
+      "grad_norm": 0.615368127822876,
+      "learning_rate": 1.8587722018344598e-05,
+      "loss": 0.2997,
+      "step": 4580
+    },
+    {
+      "epoch": 0.17141388415075823,
+      "grad_norm": 0.36148715019226074,
+      "learning_rate": 1.8584711679348818e-05,
+      "loss": 0.2574,
+      "step": 4585
+    },
+    {
+      "epoch": 0.17160081314110803,
+      "grad_norm": 0.23802785575389862,
+      "learning_rate": 1.8581698379731965e-05,
+      "loss": 0.2897,
+      "step": 4590
+    },
+    {
+      "epoch": 0.1717877421314578,
+      "grad_norm": 0.3403271436691284,
+      "learning_rate": 1.857868212053324e-05,
+      "loss": 0.3736,
+      "step": 4595
+    },
+    {
+      "epoch": 0.17197467112180762,
+      "grad_norm": 0.2740768492221832,
+      "learning_rate": 1.8575662902792854e-05,
+      "loss": 0.3326,
+      "step": 4600
+    },
+    {
+      "epoch": 0.1721616001121574,
+      "grad_norm": 0.32492953538894653,
+      "learning_rate": 1.8572640727552064e-05,
+      "loss": 0.3538,
+      "step": 4605
+    },
+    {
+      "epoch": 0.1723485291025072,
+      "grad_norm": 0.25937604904174805,
+      "learning_rate": 1.856961559585312e-05,
+      "loss": 0.268,
+      "step": 4610
+    },
+    {
+      "epoch": 0.17253545809285697,
+      "grad_norm": 0.28443995118141174,
+      "learning_rate": 1.8566587508739312e-05,
+      "loss": 0.278,
+      "step": 4615
+    },
+    {
+      "epoch": 0.17272238708320678,
+      "grad_norm": 0.31061285734176636,
+      "learning_rate": 1.856355646725493e-05,
+      "loss": 0.3424,
+      "step": 4620
+    },
+    {
+      "epoch": 0.17290931607355656,
+      "grad_norm": 0.29571932554244995,
+      "learning_rate": 1.8560522472445304e-05,
+      "loss": 0.3045,
+      "step": 4625
+    },
+    {
+      "epoch": 0.17309624506390636,
+      "grad_norm": 0.38453951478004456,
+      "learning_rate": 1.8557485525356765e-05,
+      "loss": 0.3163,
+      "step": 4630
+    },
+    {
+      "epoch": 0.17328317405425614,
+      "grad_norm": 0.21908405423164368,
+      "learning_rate": 1.855444562703667e-05,
+      "loss": 0.2375,
+      "step": 4635
+    },
+    {
+      "epoch": 0.17347010304460592,
+      "grad_norm": 0.3414257764816284,
+      "learning_rate": 1.85514027785334e-05,
+      "loss": 0.3415,
+      "step": 4640
+    },
+    {
+      "epoch": 0.17365703203495572,
+      "grad_norm": 0.6744435429573059,
+      "learning_rate": 1.8548356980896337e-05,
+      "loss": 0.3277,
+      "step": 4645
+    },
+    {
+      "epoch": 0.1738439610253055,
+      "grad_norm": 0.6530652046203613,
+      "learning_rate": 1.854530823517589e-05,
+      "loss": 0.3593,
+      "step": 4650
+    },
+    {
+      "epoch": 0.1740308900156553,
+      "grad_norm": 0.5472162961959839,
+      "learning_rate": 1.854225654242349e-05,
+      "loss": 0.295,
+      "step": 4655
+    },
+    {
+      "epoch": 0.17421781900600508,
+      "grad_norm": 0.5321584939956665,
+      "learning_rate": 1.8539201903691574e-05,
+      "loss": 0.3521,
+      "step": 4660
+    },
+    {
+      "epoch": 0.1744047479963549,
+      "grad_norm": 0.4973386824131012,
+      "learning_rate": 1.8536144320033602e-05,
+      "loss": 0.3237,
+      "step": 4665
+    },
+    {
+      "epoch": 0.17459167698670466,
+      "grad_norm": 0.37571483850479126,
+      "learning_rate": 1.8533083792504043e-05,
+      "loss": 0.3179,
+      "step": 4670
+    },
+    {
+      "epoch": 0.17477860597705447,
+      "grad_norm": 0.31560763716697693,
+      "learning_rate": 1.8530020322158392e-05,
+      "loss": 0.3937,
+      "step": 4675
+    },
+    {
+      "epoch": 0.17496553496740425,
+      "grad_norm": 0.3787771165370941,
+      "learning_rate": 1.8526953910053143e-05,
+      "loss": 0.2801,
+      "step": 4680
+    },
+    {
+      "epoch": 0.17515246395775405,
+      "grad_norm": 0.2740449905395508,
+      "learning_rate": 1.852388455724582e-05,
+      "loss": 0.3061,
+      "step": 4685
+    },
+    {
+      "epoch": 0.17533939294810383,
+      "grad_norm": 0.43204188346862793,
+      "learning_rate": 1.8520812264794954e-05,
+      "loss": 0.3166,
+      "step": 4690
+    },
+    {
+      "epoch": 0.17552632193845363,
+      "grad_norm": 0.2784161865711212,
+      "learning_rate": 1.851773703376009e-05,
+      "loss": 0.2652,
+      "step": 4695
+    },
+    {
+      "epoch": 0.1757132509288034,
+      "grad_norm": 0.45802968740463257,
+      "learning_rate": 1.8514658865201786e-05,
+      "loss": 0.3275,
+      "step": 4700
+    },
+    {
+      "epoch": 0.17590017991915322,
+      "grad_norm": 0.30260100960731506,
+      "learning_rate": 1.8511577760181615e-05,
+      "loss": 0.3034,
+      "step": 4705
+    },
+    {
+      "epoch": 0.176087108909503,
+      "grad_norm": 0.44082552194595337,
+      "learning_rate": 1.8508493719762162e-05,
+      "loss": 0.3504,
+      "step": 4710
+    },
+    {
+      "epoch": 0.1762740378998528,
+      "grad_norm": 0.3765546977519989,
+      "learning_rate": 1.850540674500703e-05,
+      "loss": 0.3524,
+      "step": 4715
+    },
+    {
+      "epoch": 0.17646096689020258,
+      "grad_norm": 0.30463677644729614,
+      "learning_rate": 1.8502316836980815e-05,
+      "loss": 0.3177,
+      "step": 4720
+    },
+    {
+      "epoch": 0.17664789588055238,
+      "grad_norm": 0.39802396297454834,
+      "learning_rate": 1.8499223996749148e-05,
+      "loss": 0.2554,
+      "step": 4725
+    },
+    {
+      "epoch": 0.17683482487090216,
+      "grad_norm": 0.44730278849601746,
+      "learning_rate": 1.849612822537866e-05,
+      "loss": 0.3275,
+      "step": 4730
+    },
+    {
+      "epoch": 0.17702175386125196,
+      "grad_norm": 0.20249329507350922,
+      "learning_rate": 1.849302952393699e-05,
+      "loss": 0.3258,
+      "step": 4735
+    },
+    {
+      "epoch": 0.17720868285160174,
+      "grad_norm": 0.4662615656852722,
+      "learning_rate": 1.8489927893492794e-05,
+      "loss": 0.3144,
+      "step": 4740
+    },
+    {
+      "epoch": 0.17739561184195154,
+      "grad_norm": 0.43017593026161194,
+      "learning_rate": 1.8486823335115735e-05,
+      "loss": 0.3936,
+      "step": 4745
+    },
+    {
+      "epoch": 0.17758254083230132,
+      "grad_norm": 0.3419100344181061,
+      "learning_rate": 1.8483715849876486e-05,
+      "loss": 0.4101,
+      "step": 4750
+    },
+    {
+      "epoch": 0.17776946982265113,
+      "grad_norm": 0.08327355980873108,
+      "learning_rate": 1.8480605438846724e-05,
+      "loss": 0.3454,
+      "step": 4755
+    },
+    {
+      "epoch": 0.1779563988130009,
+      "grad_norm": 0.386849045753479,
+      "learning_rate": 1.847749210309915e-05,
+      "loss": 0.2913,
+      "step": 4760
+    },
+    {
+      "epoch": 0.1781433278033507,
+      "grad_norm": 0.7153798937797546,
+      "learning_rate": 1.8474375843707464e-05,
+      "loss": 0.2839,
+      "step": 4765
+    },
+    {
+      "epoch": 0.1783302567937005,
+      "grad_norm": 0.3605651259422302,
+      "learning_rate": 1.8471256661746367e-05,
+      "loss": 0.3087,
+      "step": 4770
+    },
+    {
+      "epoch": 0.1785171857840503,
+      "grad_norm": 0.3076781630516052,
+      "learning_rate": 1.8468134558291582e-05,
+      "loss": 0.3429,
+      "step": 4775
+    },
+    {
+      "epoch": 0.17870411477440007,
+      "grad_norm": 0.29881712794303894,
+      "learning_rate": 1.846500953441983e-05,
+      "loss": 0.3679,
+      "step": 4780
+    },
+    {
+      "epoch": 0.17889104376474987,
+      "grad_norm": 0.39074286818504333,
+      "learning_rate": 1.8461881591208843e-05,
+      "loss": 0.4408,
+      "step": 4785
+    },
+    {
+      "epoch": 0.17907797275509965,
+      "grad_norm": 0.5738774538040161,
+      "learning_rate": 1.8458750729737356e-05,
+      "loss": 0.3933,
+      "step": 4790
+    },
+    {
+      "epoch": 0.17926490174544946,
+      "grad_norm": 0.6319369077682495,
+      "learning_rate": 1.8455616951085118e-05,
+      "loss": 0.3198,
+      "step": 4795
+    },
+    {
+      "epoch": 0.17945183073579923,
+      "grad_norm": 0.4841550886631012,
+      "learning_rate": 1.845248025633288e-05,
+      "loss": 0.389,
+      "step": 4800
+    },
+    {
+      "epoch": 0.17963875972614904,
+      "grad_norm": 0.31131798028945923,
+      "learning_rate": 1.8449340646562396e-05,
+      "loss": 0.2501,
+      "step": 4805
+    },
+    {
+      "epoch": 0.17982568871649882,
+      "grad_norm": 0.24542942643165588,
+      "learning_rate": 1.844619812285642e-05,
+      "loss": 0.3488,
+      "step": 4810
+    },
+    {
+      "epoch": 0.18001261770684862,
+      "grad_norm": 0.31681129336357117,
+      "learning_rate": 1.8443052686298733e-05,
+      "loss": 0.331,
+      "step": 4815
+    },
+    {
+      "epoch": 0.1801995466971984,
+      "grad_norm": 0.22183604538440704,
+      "learning_rate": 1.8439904337974095e-05,
+      "loss": 0.259,
+      "step": 4820
+    },
+    {
+      "epoch": 0.1803864756875482,
+      "grad_norm": 1.1499030590057373,
+      "learning_rate": 1.843675307896829e-05,
+      "loss": 0.2915,
+      "step": 4825
+    },
+    {
+      "epoch": 0.18057340467789798,
+      "grad_norm": 0.2945778965950012,
+      "learning_rate": 1.8433598910368085e-05,
+      "loss": 0.3305,
+      "step": 4830
+    },
+    {
+      "epoch": 0.18076033366824779,
+      "grad_norm": 0.3298734724521637,
+      "learning_rate": 1.8430441833261273e-05,
+      "loss": 0.2948,
+      "step": 4835
+    },
+    {
+      "epoch": 0.18094726265859756,
+      "grad_norm": 0.46152833104133606,
+      "learning_rate": 1.8427281848736632e-05,
+      "loss": 0.3241,
+      "step": 4840
+    },
+    {
+      "epoch": 0.18113419164894737,
+      "grad_norm": 0.37429046630859375,
+      "learning_rate": 1.8424118957883954e-05,
+      "loss": 0.4391,
+      "step": 4845
+    },
+    {
+      "epoch": 0.18132112063929715,
+      "grad_norm": 0.44875389337539673,
+      "learning_rate": 1.8420953161794033e-05,
+      "loss": 0.2826,
+      "step": 4850
+    },
+    {
+      "epoch": 0.18150804962964695,
+      "grad_norm": 0.287893146276474,
+      "learning_rate": 1.8417784461558656e-05,
+      "loss": 0.2415,
+      "step": 4855
+    },
+    {
+      "epoch": 0.18169497861999673,
+      "grad_norm": 0.36352619528770447,
+      "learning_rate": 1.8414612858270616e-05,
+      "loss": 0.2222,
+      "step": 4860
+    },
+    {
+      "epoch": 0.18188190761034653,
+      "grad_norm": 0.35056108236312866,
+      "learning_rate": 1.841143835302371e-05,
+      "loss": 0.2468,
+      "step": 4865
+    },
+    {
+      "epoch": 0.1820688366006963,
+      "grad_norm": 0.384459912776947,
+      "learning_rate": 1.8408260946912736e-05,
+      "loss": 0.2602,
+      "step": 4870
+    },
+    {
+      "epoch": 0.18225576559104611,
+      "grad_norm": 0.38520097732543945,
+      "learning_rate": 1.8405080641033487e-05,
+      "loss": 0.2951,
+      "step": 4875
+    },
+    {
+      "epoch": 0.1824426945813959,
+      "grad_norm": 0.2897312343120575,
+      "learning_rate": 1.8401897436482762e-05,
+      "loss": 0.3556,
+      "step": 4880
+    },
+    {
+      "epoch": 0.18262962357174567,
+      "grad_norm": 0.5431839823722839,
+      "learning_rate": 1.8398711334358355e-05,
+      "loss": 0.2967,
+      "step": 4885
+    },
+    {
+      "epoch": 0.18281655256209547,
+      "grad_norm": 0.6530733108520508,
+      "learning_rate": 1.8395522335759067e-05,
+      "loss": 0.296,
+      "step": 4890
+    },
+    {
+      "epoch": 0.18300348155244525,
+      "grad_norm": 0.5999047160148621,
+      "learning_rate": 1.8392330441784683e-05,
+      "loss": 0.3362,
+      "step": 4895
+    },
+    {
+      "epoch": 0.18319041054279506,
+      "grad_norm": 0.5073540806770325,
+      "learning_rate": 1.8389135653535998e-05,
+      "loss": 0.2777,
+      "step": 4900
+    },
+    {
+      "epoch": 0.18337733953314483,
+      "grad_norm": 0.48142045736312866,
+      "learning_rate": 1.838593797211481e-05,
+      "loss": 0.2659,
+      "step": 4905
+    },
+    {
+      "epoch": 0.18356426852349464,
+      "grad_norm": 0.2865387201309204,
+      "learning_rate": 1.8382737398623904e-05,
+      "loss": 0.2702,
+      "step": 4910
+    },
+    {
+      "epoch": 0.18375119751384442,
+      "grad_norm": 0.6015134453773499,
+      "learning_rate": 1.8379533934167063e-05,
+      "loss": 0.3637,
+      "step": 4915
+    },
+    {
+      "epoch": 0.18393812650419422,
+      "grad_norm": 0.33356979489326477,
+      "learning_rate": 1.8376327579849068e-05,
+      "loss": 0.3628,
+      "step": 4920
+    },
+    {
+      "epoch": 0.184125055494544,
+      "grad_norm": 0.7000239491462708,
+      "learning_rate": 1.8373118336775707e-05,
+      "loss": 0.3346,
+      "step": 4925
+    },
+    {
+      "epoch": 0.1843119844848938,
+      "grad_norm": 0.4333287477493286,
+      "learning_rate": 1.8369906206053753e-05,
+      "loss": 0.2867,
+      "step": 4930
+    },
+    {
+      "epoch": 0.18449891347524358,
+      "grad_norm": 0.4251829981803894,
+      "learning_rate": 1.8366691188790976e-05,
+      "loss": 0.2435,
+      "step": 4935
+    },
+    {
+      "epoch": 0.1846858424655934,
+      "grad_norm": 0.41755521297454834,
+      "learning_rate": 1.836347328609614e-05,
+      "loss": 0.3786,
+      "step": 4940
+    },
+    {
+      "epoch": 0.18487277145594316,
+      "grad_norm": 0.40600693225860596,
+      "learning_rate": 1.8360252499079015e-05,
+      "loss": 0.2621,
+      "step": 4945
+    },
+    {
+      "epoch": 0.18505970044629297,
+      "grad_norm": 0.5500521063804626,
+      "learning_rate": 1.8357028828850356e-05,
+      "loss": 0.2544,
+      "step": 4950
+    },
+    {
+      "epoch": 0.18524662943664275,
+      "grad_norm": 0.5678396224975586,
+      "learning_rate": 1.8353802276521908e-05,
+      "loss": 0.3178,
+      "step": 4955
+    },
+    {
+      "epoch": 0.18543355842699255,
+      "grad_norm": 0.6940974593162537,
+      "learning_rate": 1.8350572843206425e-05,
+      "loss": 0.3536,
+      "step": 4960
+    },
+    {
+      "epoch": 0.18562048741734233,
+      "grad_norm": 0.2244710922241211,
+      "learning_rate": 1.8347340530017642e-05,
+      "loss": 0.3592,
+      "step": 4965
+    },
+    {
+      "epoch": 0.18580741640769213,
+      "grad_norm": 0.3188341557979584,
+      "learning_rate": 1.8344105338070294e-05,
+      "loss": 0.35,
+      "step": 4970
+    },
+    {
+      "epoch": 0.1859943453980419,
+      "grad_norm": 0.33533531427383423,
+      "learning_rate": 1.8340867268480102e-05,
+      "loss": 0.3611,
+      "step": 4975
+    },
+    {
+      "epoch": 0.18618127438839172,
+      "grad_norm": 0.3602057099342346,
+      "learning_rate": 1.8337626322363782e-05,
+      "loss": 0.3204,
+      "step": 4980
+    },
+    {
+      "epoch": 0.1863682033787415,
+      "grad_norm": 0.4599422216415405,
+      "learning_rate": 1.833438250083905e-05,
+      "loss": 0.3248,
+      "step": 4985
+    },
+    {
+      "epoch": 0.1865551323690913,
+      "grad_norm": 0.24511076509952545,
+      "learning_rate": 1.8331135805024606e-05,
+      "loss": 0.3311,
+      "step": 4990
+    },
+    {
+      "epoch": 0.18674206135944107,
+      "grad_norm": 0.6302551031112671,
+      "learning_rate": 1.8327886236040137e-05,
+      "loss": 0.3417,
+      "step": 4995
+    },
+    {
+      "epoch": 0.18692899034979088,
+      "grad_norm": 0.3400326073169708,
+      "learning_rate": 1.832463379500633e-05,
+      "loss": 0.2946,
+      "step": 5000
+    },
+    {
+      "epoch": 0.18711591934014066,
+      "grad_norm": 0.2726198732852936,
+      "learning_rate": 1.8321378483044855e-05,
+      "loss": 0.317,
+      "step": 5005
+    },
+    {
+      "epoch": 0.18730284833049046,
+      "grad_norm": 0.28910648822784424,
+      "learning_rate": 1.8318120301278382e-05,
+      "loss": 0.4624,
+      "step": 5010
+    },
+    {
+      "epoch": 0.18748977732084024,
+      "grad_norm": 0.47752344608306885,
+      "learning_rate": 1.831485925083056e-05,
+      "loss": 0.3485,
+      "step": 5015
+    },
+    {
+      "epoch": 0.18767670631119004,
+      "grad_norm": 0.4108814597129822,
+      "learning_rate": 1.8311595332826034e-05,
+      "loss": 0.2964,
+      "step": 5020
+    },
+    {
+      "epoch": 0.18786363530153982,
+      "grad_norm": 0.38854309916496277,
+      "learning_rate": 1.8308328548390437e-05,
+      "loss": 0.3555,
+      "step": 5025
+    },
+    {
+      "epoch": 0.18805056429188963,
+      "grad_norm": 0.3640444576740265,
+      "learning_rate": 1.8305058898650387e-05,
+      "loss": 0.2959,
+      "step": 5030
+    },
+    {
+      "epoch": 0.1882374932822394,
+      "grad_norm": 0.5036413073539734,
+      "learning_rate": 1.830178638473349e-05,
+      "loss": 0.2471,
+      "step": 5035
+    },
+    {
+      "epoch": 0.1884244222725892,
+      "grad_norm": 0.8091880083084106,
+      "learning_rate": 1.8298511007768347e-05,
+      "loss": 0.3366,
+      "step": 5040
+    },
+    {
+      "epoch": 0.188611351262939,
+      "grad_norm": 0.43067753314971924,
+      "learning_rate": 1.829523276888454e-05,
+      "loss": 0.2554,
+      "step": 5045
+    },
+    {
+      "epoch": 0.1887982802532888,
+      "grad_norm": 0.1925475299358368,
+      "learning_rate": 1.8291951669212637e-05,
+      "loss": 0.2483,
+      "step": 5050
+    },
+    {
+      "epoch": 0.18898520924363857,
+      "grad_norm": 0.5737656354904175,
+      "learning_rate": 1.82886677098842e-05,
+      "loss": 0.3381,
+      "step": 5055
+    },
+    {
+      "epoch": 0.18917213823398837,
+      "grad_norm": 0.6750152707099915,
+      "learning_rate": 1.828538089203177e-05,
+      "loss": 0.3419,
+      "step": 5060
+    },
+    {
+      "epoch": 0.18935906722433815,
+      "grad_norm": 0.4922321140766144,
+      "learning_rate": 1.828209121678888e-05,
+      "loss": 0.268,
+      "step": 5065
+    },
+    {
+      "epoch": 0.18954599621468796,
+      "grad_norm": 0.557983934879303,
+      "learning_rate": 1.8278798685290037e-05,
+      "loss": 0.2987,
+      "step": 5070
+    },
+    {
+      "epoch": 0.18973292520503773,
+      "grad_norm": 0.4714908301830292,
+      "learning_rate": 1.8275503298670742e-05,
+      "loss": 0.3768,
+      "step": 5075
+    },
+    {
+      "epoch": 0.18991985419538754,
+      "grad_norm": 0.38415268063545227,
+      "learning_rate": 1.8272205058067488e-05,
+      "loss": 0.2899,
+      "step": 5080
+    },
+    {
+      "epoch": 0.19010678318573732,
+      "grad_norm": 0.640612006187439,
+      "learning_rate": 1.8268903964617738e-05,
+      "loss": 0.3488,
+      "step": 5085
+    },
+    {
+      "epoch": 0.19029371217608712,
+      "grad_norm": 0.3866754472255707,
+      "learning_rate": 1.826560001945994e-05,
+      "loss": 0.2575,
+      "step": 5090
+    },
+    {
+      "epoch": 0.1904806411664369,
+      "grad_norm": 0.6181028485298157,
+      "learning_rate": 1.826229322373354e-05,
+      "loss": 0.3078,
+      "step": 5095
+    },
+    {
+      "epoch": 0.1906675701567867,
+      "grad_norm": 0.5132717490196228,
+      "learning_rate": 1.825898357857895e-05,
+      "loss": 0.3201,
+      "step": 5100
+    },
+    {
+      "epoch": 0.19085449914713648,
+      "grad_norm": 0.3530460298061371,
+      "learning_rate": 1.825567108513757e-05,
+      "loss": 0.27,
+      "step": 5105
+    },
+    {
+      "epoch": 0.19104142813748629,
+      "grad_norm": 0.6483767628669739,
+      "learning_rate": 1.825235574455179e-05,
+      "loss": 0.3382,
+      "step": 5110
+    },
+    {
+      "epoch": 0.19122835712783606,
+      "grad_norm": 0.6425184607505798,
+      "learning_rate": 1.8249037557964975e-05,
+      "loss": 0.2799,
+      "step": 5115
+    },
+    {
+      "epoch": 0.19141528611818587,
+      "grad_norm": 0.5412061810493469,
+      "learning_rate": 1.8245716526521475e-05,
+      "loss": 0.3593,
+      "step": 5120
+    },
+    {
+      "epoch": 0.19160221510853565,
+      "grad_norm": 0.4309811592102051,
+      "learning_rate": 1.8242392651366607e-05,
+      "loss": 0.3289,
+      "step": 5125
+    },
+    {
+      "epoch": 0.19178914409888542,
+      "grad_norm": 0.48865464329719543,
+      "learning_rate": 1.823906593364669e-05,
+      "loss": 0.2389,
+      "step": 5130
+    },
+    {
+      "epoch": 0.19197607308923523,
+      "grad_norm": 0.4698772132396698,
+      "learning_rate": 1.8235736374509015e-05,
+      "loss": 0.3453,
+      "step": 5135
+    },
+    {
+      "epoch": 0.192163002079585,
+      "grad_norm": 0.39952799677848816,
+      "learning_rate": 1.8232403975101845e-05,
+      "loss": 0.2537,
+      "step": 5140
+    },
+    {
+      "epoch": 0.1923499310699348,
+      "grad_norm": 0.3500300347805023,
+      "learning_rate": 1.8229068736574434e-05,
+      "loss": 0.3181,
+      "step": 5145
+    },
+    {
+      "epoch": 0.1925368600602846,
+      "grad_norm": 0.3574506342411041,
+      "learning_rate": 1.8225730660077007e-05,
+      "loss": 0.2995,
+      "step": 5150
+    },
+    {
+      "epoch": 0.1927237890506344,
+      "grad_norm": 0.5655731558799744,
+      "learning_rate": 1.8222389746760774e-05,
+      "loss": 0.2967,
+      "step": 5155
+    },
+    {
+      "epoch": 0.19291071804098417,
+      "grad_norm": 0.5786092877388,
+      "learning_rate": 1.8219045997777916e-05,
+      "loss": 0.265,
+      "step": 5160
+    },
+    {
+      "epoch": 0.19309764703133397,
+      "grad_norm": 0.30824795365333557,
+      "learning_rate": 1.8215699414281602e-05,
+      "loss": 0.3385,
+      "step": 5165
+    },
+    {
+      "epoch": 0.19328457602168375,
+      "grad_norm": 0.379058301448822,
+      "learning_rate": 1.8212349997425967e-05,
+      "loss": 0.3196,
+      "step": 5170
+    },
+    {
+      "epoch": 0.19347150501203356,
+      "grad_norm": 0.2118494063615799,
+      "learning_rate": 1.820899774836613e-05,
+      "loss": 0.2843,
+      "step": 5175
+    },
+    {
+      "epoch": 0.19365843400238333,
+      "grad_norm": 0.49254727363586426,
+      "learning_rate": 1.820564266825819e-05,
+      "loss": 0.3449,
+      "step": 5180
+    },
+    {
+      "epoch": 0.19384536299273314,
+      "grad_norm": 0.45277997851371765,
+      "learning_rate": 1.8202284758259215e-05,
+      "loss": 0.3104,
+      "step": 5185
+    },
+    {
+      "epoch": 0.19403229198308292,
+      "grad_norm": 0.45387160778045654,
+      "learning_rate": 1.8198924019527252e-05,
+      "loss": 0.3288,
+      "step": 5190
+    },
+    {
+      "epoch": 0.19421922097343272,
+      "grad_norm": 0.30750924348831177,
+      "learning_rate": 1.8195560453221322e-05,
+      "loss": 0.2391,
+      "step": 5195
+    },
+    {
+      "epoch": 0.1944061499637825,
+      "grad_norm": 0.2546007037162781,
+      "learning_rate": 1.8192194060501428e-05,
+      "loss": 0.3273,
+      "step": 5200
+    },
+    {
+      "epoch": 0.1945930789541323,
+      "grad_norm": 0.37746095657348633,
+      "learning_rate": 1.8188824842528535e-05,
+      "loss": 0.401,
+      "step": 5205
+    },
+    {
+      "epoch": 0.19478000794448208,
+      "grad_norm": 0.3548900783061981,
+      "learning_rate": 1.8185452800464593e-05,
+      "loss": 0.2809,
+      "step": 5210
+    },
+    {
+      "epoch": 0.19496693693483189,
+      "grad_norm": 0.21679793298244476,
+      "learning_rate": 1.8182077935472525e-05,
+      "loss": 0.337,
+      "step": 5215
+    },
+    {
+      "epoch": 0.19515386592518166,
+      "grad_norm": 0.6175905466079712,
+      "learning_rate": 1.8178700248716225e-05,
+      "loss": 0.3285,
+      "step": 5220
+    },
+    {
+      "epoch": 0.19534079491553147,
+      "grad_norm": 0.42689886689186096,
+      "learning_rate": 1.8175319741360553e-05,
+      "loss": 0.3127,
+      "step": 5225
+    },
+    {
+      "epoch": 0.19552772390588125,
+      "grad_norm": 0.26765209436416626,
+      "learning_rate": 1.8171936414571358e-05,
+      "loss": 0.2753,
+      "step": 5230
+    },
+    {
+      "epoch": 0.19571465289623105,
+      "grad_norm": 0.2162155956029892,
+      "learning_rate": 1.816855026951545e-05,
+      "loss": 0.3051,
+      "step": 5235
+    },
+    {
+      "epoch": 0.19590158188658083,
+      "grad_norm": 0.4493172764778137,
+      "learning_rate": 1.8165161307360613e-05,
+      "loss": 0.2757,
+      "step": 5240
+    },
+    {
+      "epoch": 0.19608851087693063,
+      "grad_norm": 0.3421705961227417,
+      "learning_rate": 1.81617695292756e-05,
+      "loss": 0.3372,
+      "step": 5245
+    },
+    {
+      "epoch": 0.1962754398672804,
+      "grad_norm": 0.48634567856788635,
+      "learning_rate": 1.8158374936430144e-05,
+      "loss": 0.3392,
+      "step": 5250
+    },
+    {
+      "epoch": 0.19646236885763022,
+      "grad_norm": 0.3790494203567505,
+      "learning_rate": 1.8154977529994934e-05,
+      "loss": 0.2742,
+      "step": 5255
+    },
+    {
+      "epoch": 0.19664929784798,
+      "grad_norm": 0.48067718744277954,
+      "learning_rate": 1.8151577311141647e-05,
+      "loss": 0.304,
+      "step": 5260
+    },
+    {
+      "epoch": 0.1968362268383298,
+      "grad_norm": 0.374163419008255,
+      "learning_rate": 1.814817428104292e-05,
+      "loss": 0.2721,
+      "step": 5265
+    },
+    {
+      "epoch": 0.19702315582867957,
+      "grad_norm": 0.3786110281944275,
+      "learning_rate": 1.8144768440872353e-05,
+      "loss": 0.3316,
+      "step": 5270
+    },
+    {
+      "epoch": 0.19721008481902938,
+      "grad_norm": 0.4367314875125885,
+      "learning_rate": 1.8141359791804532e-05,
+      "loss": 0.3282,
+      "step": 5275
+    },
+    {
+      "epoch": 0.19739701380937916,
+      "grad_norm": 0.3415897786617279,
+      "learning_rate": 1.8137948335014998e-05,
+      "loss": 0.2399,
+      "step": 5280
+    },
+    {
+      "epoch": 0.19758394279972896,
+      "grad_norm": 0.46700483560562134,
+      "learning_rate": 1.813453407168026e-05,
+      "loss": 0.2813,
+      "step": 5285
+    },
+    {
+      "epoch": 0.19777087179007874,
+      "grad_norm": 0.30502578616142273,
+      "learning_rate": 1.813111700297781e-05,
+      "loss": 0.3091,
+      "step": 5290
+    },
+    {
+      "epoch": 0.19795780078042854,
+      "grad_norm": 0.37599360942840576,
+      "learning_rate": 1.812769713008609e-05,
+      "loss": 0.2241,
+      "step": 5295
+    },
+    {
+      "epoch": 0.19814472977077832,
+      "grad_norm": 0.4802122712135315,
+      "learning_rate": 1.812427445418452e-05,
+      "loss": 0.2691,
+      "step": 5300
+    },
+    {
+      "epoch": 0.19833165876112813,
+      "grad_norm": 0.41680780053138733,
+      "learning_rate": 1.8120848976453475e-05,
+      "loss": 0.4307,
+      "step": 5305
+    },
+    {
+      "epoch": 0.1985185877514779,
+      "grad_norm": 0.4099524915218353,
+      "learning_rate": 1.8117420698074318e-05,
+      "loss": 0.3089,
+      "step": 5310
+    },
+    {
+      "epoch": 0.1987055167418277,
+      "grad_norm": 0.6154845952987671,
+      "learning_rate": 1.811398962022935e-05,
+      "loss": 0.3269,
+      "step": 5315
+    },
+    {
+      "epoch": 0.1988924457321775,
+      "grad_norm": 0.29491856694221497,
+      "learning_rate": 1.811055574410186e-05,
+      "loss": 0.2856,
+      "step": 5320
+    },
+    {
+      "epoch": 0.1990793747225273,
+      "grad_norm": 0.3641104996204376,
+      "learning_rate": 1.810711907087609e-05,
+      "loss": 0.2881,
+      "step": 5325
+    },
+    {
+      "epoch": 0.19926630371287707,
+      "grad_norm": 0.33928000926971436,
+      "learning_rate": 1.8103679601737244e-05,
+      "loss": 0.2707,
+      "step": 5330
+    },
+    {
+      "epoch": 0.19945323270322687,
+      "grad_norm": 0.21978841722011566,
+      "learning_rate": 1.810023733787151e-05,
+      "loss": 0.2939,
+      "step": 5335
+    },
+    {
+      "epoch": 0.19964016169357665,
+      "grad_norm": 0.3131517469882965,
+      "learning_rate": 1.8096792280466016e-05,
+      "loss": 0.2912,
+      "step": 5340
+    },
+    {
+      "epoch": 0.19982709068392646,
+      "grad_norm": 0.42469385266304016,
+      "learning_rate": 1.8093344430708873e-05,
+      "loss": 0.2581,
+      "step": 5345
+    },
+    {
+      "epoch": 0.20001401967427623,
+      "grad_norm": 0.31522685289382935,
+      "learning_rate": 1.8089893789789134e-05,
+      "loss": 0.3042,
+      "step": 5350
+    },
+    {
+      "epoch": 0.20020094866462604,
+      "grad_norm": 0.5416140556335449,
+      "learning_rate": 1.8086440358896834e-05,
+      "loss": 0.2743,
+      "step": 5355
+    },
+    {
+      "epoch": 0.20038787765497582,
+      "grad_norm": 0.424735963344574,
+      "learning_rate": 1.808298413922296e-05,
+      "loss": 0.2687,
+      "step": 5360
+    },
+    {
+      "epoch": 0.20057480664532562,
+      "grad_norm": 0.24916088581085205,
+      "learning_rate": 1.807952513195946e-05,
+      "loss": 0.3091,
+      "step": 5365
+    },
+    {
+      "epoch": 0.2007617356356754,
+      "grad_norm": 0.5970064401626587,
+      "learning_rate": 1.8076063338299254e-05,
+      "loss": 0.3088,
+      "step": 5370
+    },
+    {
+      "epoch": 0.20094866462602518,
+      "grad_norm": 0.5019383430480957,
+      "learning_rate": 1.807259875943621e-05,
+      "loss": 0.2508,
+      "step": 5375
+    },
+    {
+      "epoch": 0.20113559361637498,
+      "grad_norm": 0.39974379539489746,
+      "learning_rate": 1.8069131396565164e-05,
+      "loss": 0.2661,
+      "step": 5380
+    },
+    {
+      "epoch": 0.20132252260672476,
+      "grad_norm": 0.42386549711227417,
+      "learning_rate": 1.8065661250881908e-05,
+      "loss": 0.2875,
+      "step": 5385
+    },
+    {
+      "epoch": 0.20150945159707456,
+      "grad_norm": 0.626395583152771,
+      "learning_rate": 1.8062188323583193e-05,
+      "loss": 0.3221,
+      "step": 5390
+    },
+    {
+      "epoch": 0.20169638058742434,
+      "grad_norm": 0.5648181438446045,
+      "learning_rate": 1.805871261586674e-05,
+      "loss": 0.308,
+      "step": 5395
+    },
+    {
+      "epoch": 0.20188330957777414,
+      "grad_norm": 0.39882054924964905,
+      "learning_rate": 1.8055234128931218e-05,
+      "loss": 0.2746,
+      "step": 5400
+    },
+    {
+      "epoch": 0.20207023856812392,
+      "grad_norm": 0.3538493812084198,
+      "learning_rate": 1.8051752863976257e-05,
+      "loss": 0.364,
+      "step": 5405
+    },
+    {
+      "epoch": 0.20225716755847373,
+      "grad_norm": 0.5525597333908081,
+      "learning_rate": 1.804826882220244e-05,
+      "loss": 0.315,
+      "step": 5410
+    },
+    {
+      "epoch": 0.2024440965488235,
+      "grad_norm": 0.42425400018692017,
+      "learning_rate": 1.8044782004811325e-05,
+      "loss": 0.3296,
+      "step": 5415
+    },
+    {
+      "epoch": 0.2026310255391733,
+      "grad_norm": 0.3802151083946228,
+      "learning_rate": 1.8041292413005406e-05,
+      "loss": 0.2939,
+      "step": 5420
+    },
+    {
+      "epoch": 0.2028179545295231,
+      "grad_norm": 0.6813692450523376,
+      "learning_rate": 1.8037800047988145e-05,
+      "loss": 0.329,
+      "step": 5425
+    },
+    {
+      "epoch": 0.2030048835198729,
+      "grad_norm": 0.4450630247592926,
+      "learning_rate": 1.8034304910963957e-05,
+      "loss": 0.2982,
+      "step": 5430
+    },
+    {
+      "epoch": 0.20319181251022267,
+      "grad_norm": 0.4038054347038269,
+      "learning_rate": 1.8030807003138223e-05,
+      "loss": 0.4678,
+      "step": 5435
+    },
+    {
+      "epoch": 0.20337874150057247,
+      "grad_norm": 0.6570813655853271,
+      "learning_rate": 1.8027306325717263e-05,
+      "loss": 0.2544,
+      "step": 5440
+    },
+    {
+      "epoch": 0.20356567049092225,
+      "grad_norm": 0.3153902292251587,
+      "learning_rate": 1.802380287990836e-05,
+      "loss": 0.2733,
+      "step": 5445
+    },
+    {
+      "epoch": 0.20375259948127206,
+      "grad_norm": 0.3435896635055542,
+      "learning_rate": 1.802029666691976e-05,
+      "loss": 0.2971,
+      "step": 5450
+    },
+    {
+      "epoch": 0.20393952847162183,
+      "grad_norm": 0.24299199879169464,
+      "learning_rate": 1.8016787687960645e-05,
+      "loss": 0.2603,
+      "step": 5455
+    },
+    {
+      "epoch": 0.20412645746197164,
+      "grad_norm": 0.7944544553756714,
+      "learning_rate": 1.801327594424117e-05,
+      "loss": 0.3944,
+      "step": 5460
+    },
+    {
+      "epoch": 0.20431338645232142,
+      "grad_norm": 0.3297552764415741,
+      "learning_rate": 1.8009761436972427e-05,
+      "loss": 0.282,
+      "step": 5465
+    },
+    {
+      "epoch": 0.20450031544267122,
+      "grad_norm": 0.48783180117607117,
+      "learning_rate": 1.8006244167366478e-05,
+      "loss": 0.2335,
+      "step": 5470
+    },
+    {
+      "epoch": 0.204687244433021,
+      "grad_norm": 0.6959665417671204,
+      "learning_rate": 1.800272413663632e-05,
+      "loss": 0.2497,
+      "step": 5475
+    },
+    {
+      "epoch": 0.2048741734233708,
+      "grad_norm": 0.3114238679409027,
+      "learning_rate": 1.7999201345995918e-05,
+      "loss": 0.248,
+      "step": 5480
+    },
+    {
+      "epoch": 0.20506110241372058,
+      "grad_norm": 0.4099029004573822,
+      "learning_rate": 1.7995675796660175e-05,
+      "loss": 0.3176,
+      "step": 5485
+    },
+    {
+      "epoch": 0.20524803140407039,
+      "grad_norm": 0.9323566555976868,
+      "learning_rate": 1.7992147489844956e-05,
+      "loss": 0.3653,
+      "step": 5490
+    },
+    {
+      "epoch": 0.20543496039442016,
+      "grad_norm": 0.5710521340370178,
+      "learning_rate": 1.798861642676707e-05,
+      "loss": 0.2791,
+      "step": 5495
+    },
+    {
+      "epoch": 0.20562188938476997,
+      "grad_norm": 0.39266684651374817,
+      "learning_rate": 1.7985082608644285e-05,
+      "loss": 0.2759,
+      "step": 5500
+    },
+    {
+      "epoch": 0.20580881837511975,
+      "grad_norm": 0.9457874298095703,
+      "learning_rate": 1.7981546036695307e-05,
+      "loss": 0.354,
+      "step": 5505
+    },
+    {
+      "epoch": 0.20599574736546955,
+      "grad_norm": 0.507675290107727,
+      "learning_rate": 1.7978006712139802e-05,
+      "loss": 0.3088,
+      "step": 5510
+    },
+    {
+      "epoch": 0.20618267635581933,
+      "grad_norm": 0.44382691383361816,
+      "learning_rate": 1.797446463619838e-05,
+      "loss": 0.3084,
+      "step": 5515
+    },
+    {
+      "epoch": 0.20636960534616913,
+      "grad_norm": 0.6009101271629333,
+      "learning_rate": 1.7970919810092603e-05,
+      "loss": 0.2258,
+      "step": 5520
+    },
+    {
+      "epoch": 0.2065565343365189,
+      "grad_norm": 0.46515893936157227,
+      "learning_rate": 1.7967372235044975e-05,
+      "loss": 0.3775,
+      "step": 5525
+    },
+    {
+      "epoch": 0.20674346332686871,
+      "grad_norm": 0.3712780475616455,
+      "learning_rate": 1.7963821912278963e-05,
+      "loss": 0.3656,
+      "step": 5530
+    },
+    {
+      "epoch": 0.2069303923172185,
+      "grad_norm": 0.6315010190010071,
+      "learning_rate": 1.7960268843018964e-05,
+      "loss": 0.3175,
+      "step": 5535
+    },
+    {
+      "epoch": 0.2071173213075683,
+      "grad_norm": 0.4107741117477417,
+      "learning_rate": 1.7956713028490332e-05,
+      "loss": 0.2603,
+      "step": 5540
+    },
+    {
+      "epoch": 0.20730425029791807,
+      "grad_norm": 0.2759909927845001,
+      "learning_rate": 1.7953154469919365e-05,
+      "loss": 0.2768,
+      "step": 5545
+    },
+    {
+      "epoch": 0.20749117928826788,
+      "grad_norm": 0.3919317126274109,
+      "learning_rate": 1.7949593168533304e-05,
+      "loss": 0.4107,
+      "step": 5550
+    },
+    {
+      "epoch": 0.20767810827861766,
+      "grad_norm": 0.23572702705860138,
+      "learning_rate": 1.7946029125560352e-05,
+      "loss": 0.3234,
+      "step": 5555
+    },
+    {
+      "epoch": 0.20786503726896746,
+      "grad_norm": 0.40106138586997986,
+      "learning_rate": 1.794246234222963e-05,
+      "loss": 0.3748,
+      "step": 5560
+    },
+    {
+      "epoch": 0.20805196625931724,
+      "grad_norm": 0.5880841016769409,
+      "learning_rate": 1.793889281977123e-05,
+      "loss": 0.2439,
+      "step": 5565
+    },
+    {
+      "epoch": 0.20823889524966704,
+      "grad_norm": 0.2483060657978058,
+      "learning_rate": 1.7935320559416173e-05,
+      "loss": 0.3053,
+      "step": 5570
+    },
+    {
+      "epoch": 0.20842582424001682,
+      "grad_norm": 1.2637639045715332,
+      "learning_rate": 1.7931745562396432e-05,
+      "loss": 0.3467,
+      "step": 5575
+    },
+    {
+      "epoch": 0.20861275323036663,
+      "grad_norm": 0.26382896304130554,
+      "learning_rate": 1.7928167829944917e-05,
+      "loss": 0.2091,
+      "step": 5580
+    },
+    {
+      "epoch": 0.2087996822207164,
+      "grad_norm": 0.3013812005519867,
+      "learning_rate": 1.7924587363295493e-05,
+      "loss": 0.2777,
+      "step": 5585
+    },
+    {
+      "epoch": 0.2089866112110662,
+      "grad_norm": 0.5539411306381226,
+      "learning_rate": 1.792100416368295e-05,
+      "loss": 0.3764,
+      "step": 5590
+    },
+    {
+      "epoch": 0.209173540201416,
+      "grad_norm": 0.40408313274383545,
+      "learning_rate": 1.791741823234304e-05,
+      "loss": 0.3294,
+      "step": 5595
+    },
+    {
+      "epoch": 0.2093604691917658,
+      "grad_norm": 0.7178158164024353,
+      "learning_rate": 1.7913829570512445e-05,
+      "loss": 0.3448,
+      "step": 5600
+    },
+    {
+      "epoch": 0.20954739818211557,
+      "grad_norm": 0.31547850370407104,
+      "learning_rate": 1.791023817942879e-05,
+      "loss": 0.3079,
+      "step": 5605
+    },
+    {
+      "epoch": 0.20973432717246537,
+      "grad_norm": 0.4582591950893402,
+      "learning_rate": 1.7906644060330646e-05,
+      "loss": 0.2815,
+      "step": 5610
+    },
+    {
+      "epoch": 0.20992125616281515,
+      "grad_norm": 0.7002764344215393,
+      "learning_rate": 1.7903047214457517e-05,
+      "loss": 0.2797,
+      "step": 5615
+    },
+    {
+      "epoch": 0.21010818515316493,
+      "grad_norm": 0.4524244964122772,
+      "learning_rate": 1.7899447643049855e-05,
+      "loss": 0.402,
+      "step": 5620
+    },
+    {
+      "epoch": 0.21029511414351473,
+      "grad_norm": 0.40630853176116943,
+      "learning_rate": 1.7895845347349047e-05,
+      "loss": 0.3465,
+      "step": 5625
+    },
+    {
+      "epoch": 0.2104820431338645,
+      "grad_norm": 0.21824871003627777,
+      "learning_rate": 1.7892240328597427e-05,
+      "loss": 0.2492,
+      "step": 5630
+    },
+    {
+      "epoch": 0.21066897212421432,
+      "grad_norm": 0.3022218644618988,
+      "learning_rate": 1.7888632588038256e-05,
+      "loss": 0.2882,
+      "step": 5635
+    },
+    {
+      "epoch": 0.2108559011145641,
+      "grad_norm": 0.761012077331543,
+      "learning_rate": 1.7885022126915743e-05,
+      "loss": 0.3003,
+      "step": 5640
+    },
+    {
+      "epoch": 0.2110428301049139,
+      "grad_norm": 0.39378008246421814,
+      "learning_rate": 1.7881408946475035e-05,
+      "loss": 0.296,
+      "step": 5645
+    },
+    {
+      "epoch": 0.21122975909526367,
+      "grad_norm": 0.6320983171463013,
+      "learning_rate": 1.787779304796221e-05,
+      "loss": 0.2578,
+      "step": 5650
+    },
+    {
+      "epoch": 0.21141668808561348,
+      "grad_norm": 0.3272438943386078,
+      "learning_rate": 1.787417443262429e-05,
+      "loss": 0.297,
+      "step": 5655
+    },
+    {
+      "epoch": 0.21160361707596326,
+      "grad_norm": 0.4508860409259796,
+      "learning_rate": 1.7870553101709232e-05,
+      "loss": 0.2263,
+      "step": 5660
+    },
+    {
+      "epoch": 0.21179054606631306,
+      "grad_norm": 0.3349319398403168,
+      "learning_rate": 1.7866929056465933e-05,
+      "loss": 0.3182,
+      "step": 5665
+    },
+    {
+      "epoch": 0.21197747505666284,
+      "grad_norm": 0.7591943740844727,
+      "learning_rate": 1.7863302298144218e-05,
+      "loss": 0.2884,
+      "step": 5670
+    },
+    {
+      "epoch": 0.21216440404701264,
+      "grad_norm": 0.35864681005477905,
+      "learning_rate": 1.785967282799485e-05,
+      "loss": 0.2924,
+      "step": 5675
+    },
+    {
+      "epoch": 0.21235133303736242,
+      "grad_norm": 0.42692896723747253,
+      "learning_rate": 1.785604064726953e-05,
+      "loss": 0.2845,
+      "step": 5680
+    },
+    {
+      "epoch": 0.21253826202771223,
+      "grad_norm": 0.5757957100868225,
+      "learning_rate": 1.7852405757220898e-05,
+      "loss": 0.3391,
+      "step": 5685
+    },
+    {
+      "epoch": 0.212725191018062,
+      "grad_norm": 0.48782822489738464,
+      "learning_rate": 1.7848768159102522e-05,
+      "loss": 0.3037,
+      "step": 5690
+    },
+    {
+      "epoch": 0.2129121200084118,
+      "grad_norm": 0.31845298409461975,
+      "learning_rate": 1.78451278541689e-05,
+      "loss": 0.2772,
+      "step": 5695
+    },
+    {
+      "epoch": 0.2130990489987616,
+      "grad_norm": 0.33685699105262756,
+      "learning_rate": 1.7841484843675473e-05,
+      "loss": 0.319,
+      "step": 5700
+    },
+    {
+      "epoch": 0.2132859779891114,
+      "grad_norm": 0.391272634267807,
+      "learning_rate": 1.7837839128878608e-05,
+      "loss": 0.324,
+      "step": 5705
+    },
+    {
+      "epoch": 0.21347290697946117,
+      "grad_norm": 0.42025449872016907,
+      "learning_rate": 1.7834190711035613e-05,
+      "loss": 0.2744,
+      "step": 5710
+    },
+    {
+      "epoch": 0.21365983596981097,
+      "grad_norm": 0.5056503415107727,
+      "learning_rate": 1.7830539591404717e-05,
+      "loss": 0.2977,
+      "step": 5715
+    },
+    {
+      "epoch": 0.21384676496016075,
+      "grad_norm": 0.5261359214782715,
+      "learning_rate": 1.7826885771245094e-05,
+      "loss": 0.3341,
+      "step": 5720
+    },
+    {
+      "epoch": 0.21403369395051056,
+      "grad_norm": 0.44913962483406067,
+      "learning_rate": 1.7823229251816836e-05,
+      "loss": 0.2768,
+      "step": 5725
+    },
+    {
+      "epoch": 0.21422062294086033,
+      "grad_norm": 0.2031090408563614,
+      "learning_rate": 1.7819570034380968e-05,
+      "loss": 0.2962,
+      "step": 5730
+    },
+    {
+      "epoch": 0.21440755193121014,
+      "grad_norm": 0.3239678144454956,
+      "learning_rate": 1.7815908120199462e-05,
+      "loss": 0.2717,
+      "step": 5735
+    },
+    {
+      "epoch": 0.21459448092155992,
+      "grad_norm": 0.25322166085243225,
+      "learning_rate": 1.7812243510535194e-05,
+      "loss": 0.3227,
+      "step": 5740
+    },
+    {
+      "epoch": 0.21478140991190972,
+      "grad_norm": 0.2611919045448303,
+      "learning_rate": 1.780857620665199e-05,
+      "loss": 0.3009,
+      "step": 5745
+    },
+    {
+      "epoch": 0.2149683389022595,
+      "grad_norm": 0.580700695514679,
+      "learning_rate": 1.7804906209814597e-05,
+      "loss": 0.26,
+      "step": 5750
+    },
+    {
+      "epoch": 0.2151552678926093,
+      "grad_norm": 0.7626398205757141,
+      "learning_rate": 1.780123352128869e-05,
+      "loss": 0.3374,
+      "step": 5755
+    },
+    {
+      "epoch": 0.21534219688295908,
+      "grad_norm": 0.34990110993385315,
+      "learning_rate": 1.779755814234087e-05,
+      "loss": 0.2281,
+      "step": 5760
+    },
+    {
+      "epoch": 0.21552912587330889,
+      "grad_norm": 0.5652785301208496,
+      "learning_rate": 1.779388007423868e-05,
+      "loss": 0.324,
+      "step": 5765
+    },
+    {
+      "epoch": 0.21571605486365866,
+      "grad_norm": 0.45265763998031616,
+      "learning_rate": 1.779019931825058e-05,
+      "loss": 0.3418,
+      "step": 5770
+    },
+    {
+      "epoch": 0.21590298385400847,
+      "grad_norm": 0.47476527094841003,
+      "learning_rate": 1.7786515875645945e-05,
+      "loss": 0.2861,
+      "step": 5775
+    },
+    {
+      "epoch": 0.21608991284435825,
+      "grad_norm": 0.4736066162586212,
+      "learning_rate": 1.77828297476951e-05,
+      "loss": 0.3211,
+      "step": 5780
+    },
+    {
+      "epoch": 0.21627684183470805,
+      "grad_norm": 0.4887428879737854,
+      "learning_rate": 1.777914093566928e-05,
+      "loss": 0.3205,
+      "step": 5785
+    },
+    {
+      "epoch": 0.21646377082505783,
+      "grad_norm": 0.7084537148475647,
+      "learning_rate": 1.7775449440840656e-05,
+      "loss": 0.2698,
+      "step": 5790
+    },
+    {
+      "epoch": 0.21665069981540763,
+      "grad_norm": 0.7960801124572754,
+      "learning_rate": 1.777175526448231e-05,
+      "loss": 0.3193,
+      "step": 5795
+    },
+    {
+      "epoch": 0.2168376288057574,
+      "grad_norm": 0.48259130120277405,
+      "learning_rate": 1.776805840786826e-05,
+      "loss": 0.3232,
+      "step": 5800
+    },
+    {
+      "epoch": 0.21702455779610721,
+      "grad_norm": 0.5537059307098389,
+      "learning_rate": 1.7764358872273452e-05,
+      "loss": 0.2775,
+      "step": 5805
+    },
+    {
+      "epoch": 0.217211486786457,
+      "grad_norm": 0.2787805497646332,
+      "learning_rate": 1.7760656658973748e-05,
+      "loss": 0.285,
+      "step": 5810
+    },
+    {
+      "epoch": 0.2173984157768068,
+      "grad_norm": 0.6352003216743469,
+      "learning_rate": 1.7756951769245933e-05,
+      "loss": 0.3008,
+      "step": 5815
+    },
+    {
+      "epoch": 0.21758534476715657,
+      "grad_norm": 0.45176759362220764,
+      "learning_rate": 1.7753244204367713e-05,
+      "loss": 0.3192,
+      "step": 5820
+    },
+    {
+      "epoch": 0.21777227375750638,
+      "grad_norm": 0.38481858372688293,
+      "learning_rate": 1.7749533965617726e-05,
+      "loss": 0.3189,
+      "step": 5825
+    },
+    {
+      "epoch": 0.21795920274785616,
+      "grad_norm": 0.2588234841823578,
+      "learning_rate": 1.7745821054275533e-05,
+      "loss": 0.3369,
+      "step": 5830
+    },
+    {
+      "epoch": 0.21814613173820596,
+      "grad_norm": 0.4388332664966583,
+      "learning_rate": 1.7742105471621597e-05,
+      "loss": 0.3092,
+      "step": 5835
+    },
+    {
+      "epoch": 0.21833306072855574,
+      "grad_norm": 0.5704757571220398,
+      "learning_rate": 1.7738387218937326e-05,
+      "loss": 0.3485,
+      "step": 5840
+    },
+    {
+      "epoch": 0.21851998971890554,
+      "grad_norm": 0.23019002377986908,
+      "learning_rate": 1.7734666297505034e-05,
+      "loss": 0.3094,
+      "step": 5845
+    },
+    {
+      "epoch": 0.21870691870925532,
+      "grad_norm": 0.582446277141571,
+      "learning_rate": 1.7730942708607965e-05,
+      "loss": 0.3464,
+      "step": 5850
+    },
+    {
+      "epoch": 0.21889384769960513,
+      "grad_norm": 0.621008574962616,
+      "learning_rate": 1.7727216453530275e-05,
+      "loss": 0.3706,
+      "step": 5855
+    },
+    {
+      "epoch": 0.2190807766899549,
+      "grad_norm": 0.4594647288322449,
+      "learning_rate": 1.7723487533557043e-05,
+      "loss": 0.3341,
+      "step": 5860
+    },
+    {
+      "epoch": 0.21926770568030468,
+      "grad_norm": 0.3757644593715668,
+      "learning_rate": 1.7719755949974264e-05,
+      "loss": 0.2758,
+      "step": 5865
+    },
+    {
+      "epoch": 0.21945463467065449,
+      "grad_norm": 0.38085031509399414,
+      "learning_rate": 1.7716021704068858e-05,
+      "loss": 0.248,
+      "step": 5870
+    },
+    {
+      "epoch": 0.21964156366100426,
+      "grad_norm": 0.22712992131710052,
+      "learning_rate": 1.771228479712866e-05,
+      "loss": 0.2829,
+      "step": 5875
+    },
+    {
+      "epoch": 0.21982849265135407,
+      "grad_norm": 0.7901753783226013,
+      "learning_rate": 1.770854523044242e-05,
+      "loss": 0.3236,
+      "step": 5880
+    },
+    {
+      "epoch": 0.22001542164170385,
+      "grad_norm": 0.3070071041584015,
+      "learning_rate": 1.7704803005299806e-05,
+      "loss": 0.2932,
+      "step": 5885
+    },
+    {
+      "epoch": 0.22020235063205365,
+      "grad_norm": 0.6579908132553101,
+      "learning_rate": 1.7701058122991407e-05,
+      "loss": 0.3087,
+      "step": 5890
+    },
+    {
+      "epoch": 0.22038927962240343,
+      "grad_norm": 0.29225432872772217,
+      "learning_rate": 1.7697310584808726e-05,
+      "loss": 0.2783,
+      "step": 5895
+    },
+    {
+      "epoch": 0.22057620861275323,
+      "grad_norm": 0.4828365743160248,
+      "learning_rate": 1.769356039204418e-05,
+      "loss": 0.3803,
+      "step": 5900
+    },
+    {
+      "epoch": 0.220763137603103,
+      "grad_norm": 0.47344258427619934,
+      "learning_rate": 1.76898075459911e-05,
+      "loss": 0.3296,
+      "step": 5905
+    },
+    {
+      "epoch": 0.22095006659345282,
+      "grad_norm": 0.335570752620697,
+      "learning_rate": 1.768605204794374e-05,
+      "loss": 0.4109,
+      "step": 5910
+    },
+    {
+      "epoch": 0.2211369955838026,
+      "grad_norm": 0.4336024820804596,
+      "learning_rate": 1.7682293899197264e-05,
+      "loss": 0.3487,
+      "step": 5915
+    },
+    {
+      "epoch": 0.2213239245741524,
+      "grad_norm": 0.43026483058929443,
+      "learning_rate": 1.7678533101047745e-05,
+      "loss": 0.4125,
+      "step": 5920
+    },
+    {
+      "epoch": 0.22151085356450217,
+      "grad_norm": 0.3426781892776489,
+      "learning_rate": 1.767476965479218e-05,
+      "loss": 0.274,
+      "step": 5925
+    },
+    {
+      "epoch": 0.22169778255485198,
+      "grad_norm": 0.26045331358909607,
+      "learning_rate": 1.7671003561728468e-05,
+      "loss": 0.2431,
+      "step": 5930
+    },
+    {
+      "epoch": 0.22188471154520176,
+      "grad_norm": 0.3758314251899719,
+      "learning_rate": 1.766723482315543e-05,
+      "loss": 0.3625,
+      "step": 5935
+    },
+    {
+      "epoch": 0.22207164053555156,
+      "grad_norm": 0.5306219458580017,
+      "learning_rate": 1.7663463440372795e-05,
+      "loss": 0.3293,
+      "step": 5940
+    },
+    {
+      "epoch": 0.22225856952590134,
+      "grad_norm": 0.4184575080871582,
+      "learning_rate": 1.7659689414681208e-05,
+      "loss": 0.2876,
+      "step": 5945
+    },
+    {
+      "epoch": 0.22244549851625114,
+      "grad_norm": 0.4982224702835083,
+      "learning_rate": 1.765591274738222e-05,
+      "loss": 0.4022,
+      "step": 5950
+    },
+    {
+      "epoch": 0.22263242750660092,
+      "grad_norm": 0.6839665770530701,
+      "learning_rate": 1.765213343977829e-05,
+      "loss": 0.3823,
+      "step": 5955
+    },
+    {
+      "epoch": 0.22281935649695073,
+      "grad_norm": 0.3338620066642761,
+      "learning_rate": 1.7648351493172805e-05,
+      "loss": 0.2864,
+      "step": 5960
+    },
+    {
+      "epoch": 0.2230062854873005,
+      "grad_norm": 0.436450332403183,
+      "learning_rate": 1.764456690887004e-05,
+      "loss": 0.3375,
+      "step": 5965
+    },
+    {
+      "epoch": 0.2231932144776503,
+      "grad_norm": 0.4416438341140747,
+      "learning_rate": 1.7640779688175192e-05,
+      "loss": 0.3351,
+      "step": 5970
+    },
+    {
+      "epoch": 0.2233801434680001,
+      "grad_norm": 2.5495193004608154,
+      "learning_rate": 1.7636989832394365e-05,
+      "loss": 0.3343,
+      "step": 5975
+    },
+    {
+      "epoch": 0.2235670724583499,
+      "grad_norm": 0.6355164647102356,
+      "learning_rate": 1.7633197342834574e-05,
+      "loss": 0.3348,
+      "step": 5980
+    },
+    {
+      "epoch": 0.22375400144869967,
+      "grad_norm": 0.6266055703163147,
+      "learning_rate": 1.762940222080374e-05,
+      "loss": 0.3222,
+      "step": 5985
+    },
+    {
+      "epoch": 0.22394093043904947,
+      "grad_norm": 0.5203749537467957,
+      "learning_rate": 1.7625604467610685e-05,
+      "loss": 0.2935,
+      "step": 5990
+    },
+    {
+      "epoch": 0.22412785942939925,
+      "grad_norm": 0.40464696288108826,
+      "learning_rate": 1.7621804084565153e-05,
+      "loss": 0.3872,
+      "step": 5995
+    },
+    {
+      "epoch": 0.22431478841974906,
+      "grad_norm": 0.43349775671958923,
+      "learning_rate": 1.7618001072977783e-05,
+      "loss": 0.2921,
+      "step": 6000
+    },
+    {
+      "epoch": 0.22450171741009883,
+      "grad_norm": 2.7697973251342773,
+      "learning_rate": 1.7614195434160128e-05,
+      "loss": 0.3545,
+      "step": 6005
+    },
+    {
+      "epoch": 0.22468864640044864,
+      "grad_norm": 1.0596578121185303,
+      "learning_rate": 1.7610387169424643e-05,
+      "loss": 0.3049,
+      "step": 6010
+    },
+    {
+      "epoch": 0.22487557539079842,
+      "grad_norm": 0.39064550399780273,
+      "learning_rate": 1.7606576280084688e-05,
+      "loss": 0.3061,
+      "step": 6015
+    },
+    {
+      "epoch": 0.22506250438114822,
+      "grad_norm": 0.18736739456653595,
+      "learning_rate": 1.760276276745453e-05,
+      "loss": 0.337,
+      "step": 6020
+    },
+    {
+      "epoch": 0.225249433371498,
+      "grad_norm": 0.4019501805305481,
+      "learning_rate": 1.7598946632849338e-05,
+      "loss": 0.2916,
+      "step": 6025
+    },
+    {
+      "epoch": 0.2254363623618478,
+      "grad_norm": 0.5677745938301086,
+      "learning_rate": 1.759512787758519e-05,
+      "loss": 0.3083,
+      "step": 6030
+    },
+    {
+      "epoch": 0.22562329135219758,
+      "grad_norm": 0.6863211393356323,
+      "learning_rate": 1.7591306502979066e-05,
+      "loss": 0.3761,
+      "step": 6035
+    },
+    {
+      "epoch": 0.22581022034254739,
+      "grad_norm": 0.25616636872291565,
+      "learning_rate": 1.7587482510348848e-05,
+      "loss": 0.2742,
+      "step": 6040
+    },
+    {
+      "epoch": 0.22599714933289716,
+      "grad_norm": 0.37293756008148193,
+      "learning_rate": 1.7583655901013323e-05,
+      "loss": 0.3009,
+      "step": 6045
+    },
+    {
+      "epoch": 0.22618407832324697,
+      "grad_norm": 0.33705809712409973,
+      "learning_rate": 1.757982667629217e-05,
+      "loss": 0.2773,
+      "step": 6050
+    },
+    {
+      "epoch": 0.22637100731359674,
+      "grad_norm": 0.43777912855148315,
+      "learning_rate": 1.757599483750599e-05,
+      "loss": 0.3039,
+      "step": 6055
+    },
+    {
+      "epoch": 0.22655793630394655,
+      "grad_norm": 0.4632292091846466,
+      "learning_rate": 1.7572160385976273e-05,
+      "loss": 0.3124,
+      "step": 6060
+    },
+    {
+      "epoch": 0.22674486529429633,
+      "grad_norm": 0.2688206434249878,
+      "learning_rate": 1.756832332302541e-05,
+      "loss": 0.2935,
+      "step": 6065
+    },
+    {
+      "epoch": 0.22693179428464613,
+      "grad_norm": 0.5086357593536377,
+      "learning_rate": 1.7564483649976686e-05,
+      "loss": 0.342,
+      "step": 6070
+    },
+    {
+      "epoch": 0.2271187232749959,
+      "grad_norm": 1.0471845865249634,
+      "learning_rate": 1.7560641368154307e-05,
+      "loss": 0.3716,
+      "step": 6075
+    },
+    {
+      "epoch": 0.22730565226534571,
+      "grad_norm": 0.4194478988647461,
+      "learning_rate": 1.7556796478883355e-05,
+      "loss": 0.2641,
+      "step": 6080
+    },
+    {
+      "epoch": 0.2274925812556955,
+      "grad_norm": 0.6006360650062561,
+      "learning_rate": 1.755294898348983e-05,
+      "loss": 0.3655,
+      "step": 6085
+    },
+    {
+      "epoch": 0.2276795102460453,
+      "grad_norm": 0.48307010531425476,
+      "learning_rate": 1.754909888330062e-05,
+      "loss": 0.2505,
+      "step": 6090
+    },
+    {
+      "epoch": 0.22786643923639507,
+      "grad_norm": 0.3270209729671478,
+      "learning_rate": 1.7545246179643513e-05,
+      "loss": 0.3142,
+      "step": 6095
+    },
+    {
+      "epoch": 0.22805336822674485,
+      "grad_norm": 0.5060667395591736,
+      "learning_rate": 1.7541390873847195e-05,
+      "loss": 0.4389,
+      "step": 6100
+    },
+    {
+      "epoch": 0.22824029721709466,
+      "grad_norm": 0.15000495314598083,
+      "learning_rate": 1.7537532967241255e-05,
+      "loss": 0.3561,
+      "step": 6105
+    },
+    {
+      "epoch": 0.22842722620744443,
+      "grad_norm": 0.6780317425727844,
+      "learning_rate": 1.7533672461156173e-05,
+      "loss": 0.2892,
+      "step": 6110
+    },
+    {
+      "epoch": 0.22861415519779424,
+      "grad_norm": 0.4353383481502533,
+      "learning_rate": 1.7529809356923327e-05,
+      "loss": 0.3026,
+      "step": 6115
+    },
+    {
+      "epoch": 0.22880108418814402,
+      "grad_norm": 0.34928640723228455,
+      "learning_rate": 1.7525943655874987e-05,
+      "loss": 0.3004,
+      "step": 6120
+    },
+    {
+      "epoch": 0.22898801317849382,
+      "grad_norm": 0.538777232170105,
+      "learning_rate": 1.752207535934433e-05,
+      "loss": 0.3053,
+      "step": 6125
+    },
+    {
+      "epoch": 0.2291749421688436,
+      "grad_norm": 0.6690159440040588,
+      "learning_rate": 1.7518204468665415e-05,
+      "loss": 0.3058,
+      "step": 6130
+    },
+    {
+      "epoch": 0.2293618711591934,
+      "grad_norm": 0.29182419180870056,
+      "learning_rate": 1.75143309851732e-05,
+      "loss": 0.2783,
+      "step": 6135
+    },
+    {
+      "epoch": 0.22954880014954318,
+      "grad_norm": 0.4138169586658478,
+      "learning_rate": 1.751045491020354e-05,
+      "loss": 0.363,
+      "step": 6140
+    },
+    {
+      "epoch": 0.22973572913989299,
+      "grad_norm": 0.433714359998703,
+      "learning_rate": 1.750657624509319e-05,
+      "loss": 0.2915,
+      "step": 6145
+    },
+    {
+      "epoch": 0.22992265813024276,
+      "grad_norm": 0.41636890172958374,
+      "learning_rate": 1.750269499117978e-05,
+      "loss": 0.275,
+      "step": 6150
+    },
+    {
+      "epoch": 0.23010958712059257,
+      "grad_norm": 0.38736167550086975,
+      "learning_rate": 1.7498811149801845e-05,
+      "loss": 0.4277,
+      "step": 6155
+    },
+    {
+      "epoch": 0.23029651611094235,
+      "grad_norm": 0.578729510307312,
+      "learning_rate": 1.749492472229881e-05,
+      "loss": 0.3489,
+      "step": 6160
+    },
+    {
+      "epoch": 0.23048344510129215,
+      "grad_norm": 0.6496894359588623,
+      "learning_rate": 1.7491035710011e-05,
+      "loss": 0.2933,
+      "step": 6165
+    },
+    {
+      "epoch": 0.23067037409164193,
+      "grad_norm": 0.2986353039741516,
+      "learning_rate": 1.748714411427962e-05,
+      "loss": 0.2677,
+      "step": 6170
+    },
+    {
+      "epoch": 0.23085730308199173,
+      "grad_norm": 0.38289961218833923,
+      "learning_rate": 1.7483249936446768e-05,
+      "loss": 0.2705,
+      "step": 6175
+    },
+    {
+      "epoch": 0.2310442320723415,
+      "grad_norm": 0.4401322603225708,
+      "learning_rate": 1.7479353177855434e-05,
+      "loss": 0.2505,
+      "step": 6180
+    },
+    {
+      "epoch": 0.23123116106269131,
+      "grad_norm": 0.48002535104751587,
+      "learning_rate": 1.74754538398495e-05,
+      "loss": 0.3417,
+      "step": 6185
+    },
+    {
+      "epoch": 0.2314180900530411,
+      "grad_norm": 0.32335153222084045,
+      "learning_rate": 1.7471551923773732e-05,
+      "loss": 0.3704,
+      "step": 6190
+    },
+    {
+      "epoch": 0.2316050190433909,
+      "grad_norm": 0.34287896752357483,
+      "learning_rate": 1.7467647430973796e-05,
+      "loss": 0.3006,
+      "step": 6195
+    },
+    {
+      "epoch": 0.23179194803374067,
+      "grad_norm": 0.6139170527458191,
+      "learning_rate": 1.7463740362796235e-05,
+      "loss": 0.4078,
+      "step": 6200
+    },
+    {
+      "epoch": 0.23197887702409048,
+      "grad_norm": 0.5172100067138672,
+      "learning_rate": 1.7459830720588486e-05,
+      "loss": 0.4518,
+      "step": 6205
+    },
+    {
+      "epoch": 0.23216580601444026,
+      "grad_norm": 0.5083472728729248,
+      "learning_rate": 1.7455918505698876e-05,
+      "loss": 0.2808,
+      "step": 6210
+    },
+    {
+      "epoch": 0.23235273500479006,
+      "grad_norm": 0.3061027526855469,
+      "learning_rate": 1.745200371947661e-05,
+      "loss": 0.3108,
+      "step": 6215
+    },
+    {
+      "epoch": 0.23253966399513984,
+      "grad_norm": 0.5908476710319519,
+      "learning_rate": 1.7448086363271785e-05,
+      "loss": 0.2467,
+      "step": 6220
+    },
+    {
+      "epoch": 0.23272659298548964,
+      "grad_norm": 0.5854301452636719,
+      "learning_rate": 1.7444166438435392e-05,
+      "loss": 0.3062,
+      "step": 6225
+    },
+    {
+      "epoch": 0.23291352197583942,
+      "grad_norm": 0.6378867030143738,
+      "learning_rate": 1.7440243946319294e-05,
+      "loss": 0.2687,
+      "step": 6230
+    },
+    {
+      "epoch": 0.23310045096618923,
+      "grad_norm": 0.195028617978096,
+      "learning_rate": 1.7436318888276252e-05,
+      "loss": 0.3161,
+      "step": 6235
+    },
+    {
+      "epoch": 0.233287379956539,
+      "grad_norm": 0.22017711400985718,
+      "learning_rate": 1.74323912656599e-05,
+      "loss": 0.381,
+      "step": 6240
+    },
+    {
+      "epoch": 0.2334743089468888,
+      "grad_norm": 0.43266749382019043,
+      "learning_rate": 1.742846107982477e-05,
+      "loss": 0.2814,
+      "step": 6245
+    },
+    {
+      "epoch": 0.2336612379372386,
+      "grad_norm": 0.30291399359703064,
+      "learning_rate": 1.742452833212626e-05,
+      "loss": 0.2848,
+      "step": 6250
+    },
+    {
+      "epoch": 0.2338481669275884,
+      "grad_norm": 0.38583922386169434,
+      "learning_rate": 1.7420593023920673e-05,
+      "loss": 0.4128,
+      "step": 6255
+    },
+    {
+      "epoch": 0.23403509591793817,
+      "grad_norm": 0.34890538454055786,
+      "learning_rate": 1.7416655156565173e-05,
+      "loss": 0.2512,
+      "step": 6260
+    },
+    {
+      "epoch": 0.23422202490828797,
+      "grad_norm": 0.35292932391166687,
+      "learning_rate": 1.741271473141783e-05,
+      "loss": 0.3371,
+      "step": 6265
+    },
+    {
+      "epoch": 0.23440895389863775,
+      "grad_norm": 0.45217952132225037,
+      "learning_rate": 1.7408771749837572e-05,
+      "loss": 0.2664,
+      "step": 6270
+    },
+    {
+      "epoch": 0.23459588288898756,
+      "grad_norm": 0.32959669828414917,
+      "learning_rate": 1.7404826213184228e-05,
+      "loss": 0.392,
+      "step": 6275
+    },
+    {
+      "epoch": 0.23478281187933733,
+      "grad_norm": 0.4836936891078949,
+      "learning_rate": 1.74008781228185e-05,
+      "loss": 0.2778,
+      "step": 6280
+    },
+    {
+      "epoch": 0.23496974086968714,
+      "grad_norm": 0.31659331917762756,
+      "learning_rate": 1.7396927480101968e-05,
+      "loss": 0.2593,
+      "step": 6285
+    },
+    {
+      "epoch": 0.23515666986003692,
+      "grad_norm": 0.3063521981239319,
+      "learning_rate": 1.7392974286397096e-05,
+      "loss": 0.2791,
+      "step": 6290
+    },
+    {
+      "epoch": 0.23534359885038672,
+      "grad_norm": 0.3990836441516876,
+      "learning_rate": 1.7389018543067227e-05,
+      "loss": 0.3199,
+      "step": 6295
+    },
+    {
+      "epoch": 0.2355305278407365,
+      "grad_norm": 0.38259872794151306,
+      "learning_rate": 1.738506025147659e-05,
+      "loss": 0.2996,
+      "step": 6300
+    },
+    {
+      "epoch": 0.2357174568310863,
+      "grad_norm": 0.4150097072124481,
+      "learning_rate": 1.7381099412990276e-05,
+      "loss": 0.338,
+      "step": 6305
+    },
+    {
+      "epoch": 0.23590438582143608,
+      "grad_norm": 0.18883201479911804,
+      "learning_rate": 1.737713602897427e-05,
+      "loss": 0.3198,
+      "step": 6310
+    },
+    {
+      "epoch": 0.23609131481178589,
+      "grad_norm": 0.2726154923439026,
+      "learning_rate": 1.7373170100795433e-05,
+      "loss": 0.2758,
+      "step": 6315
+    },
+    {
+      "epoch": 0.23627824380213566,
+      "grad_norm": 0.3209015727043152,
+      "learning_rate": 1.736920162982149e-05,
+      "loss": 0.3209,
+      "step": 6320
+    },
+    {
+      "epoch": 0.23646517279248547,
+      "grad_norm": 0.32282012701034546,
+      "learning_rate": 1.736523061742107e-05,
+      "loss": 0.3198,
+      "step": 6325
+    },
+    {
+      "epoch": 0.23665210178283524,
+      "grad_norm": 0.6019655466079712,
+      "learning_rate": 1.736125706496364e-05,
+      "loss": 0.3568,
+      "step": 6330
+    },
+    {
+      "epoch": 0.23683903077318505,
+      "grad_norm": 0.42038559913635254,
+      "learning_rate": 1.7357280973819576e-05,
+      "loss": 0.3844,
+      "step": 6335
+    },
+    {
+      "epoch": 0.23702595976353483,
+      "grad_norm": 0.6067512631416321,
+      "learning_rate": 1.7353302345360118e-05,
+      "loss": 0.2941,
+      "step": 6340
+    },
+    {
+      "epoch": 0.2372128887538846,
+      "grad_norm": 0.3025815486907959,
+      "learning_rate": 1.7349321180957382e-05,
+      "loss": 0.2795,
+      "step": 6345
+    },
+    {
+      "epoch": 0.2373998177442344,
+      "grad_norm": 0.30944767594337463,
+      "learning_rate": 1.7345337481984355e-05,
+      "loss": 0.2756,
+      "step": 6350
+    },
+    {
+      "epoch": 0.2375867467345842,
+      "grad_norm": 0.33044394850730896,
+      "learning_rate": 1.73413512498149e-05,
+      "loss": 0.2536,
+      "step": 6355
+    },
+    {
+      "epoch": 0.237773675724934,
+      "grad_norm": 0.45606037974357605,
+      "learning_rate": 1.7337362485823757e-05,
+      "loss": 0.3169,
+      "step": 6360
+    },
+    {
+      "epoch": 0.23796060471528377,
+      "grad_norm": 0.38473767042160034,
+      "learning_rate": 1.7333371191386535e-05,
+      "loss": 0.2228,
+      "step": 6365
+    },
+    {
+      "epoch": 0.23814753370563357,
+      "grad_norm": 0.3136763870716095,
+      "learning_rate": 1.7329377367879715e-05,
+      "loss": 0.2625,
+      "step": 6370
+    },
+    {
+      "epoch": 0.23833446269598335,
+      "grad_norm": 0.33016493916511536,
+      "learning_rate": 1.7325381016680657e-05,
+      "loss": 0.4031,
+      "step": 6375
+    },
+    {
+      "epoch": 0.23852139168633316,
+      "grad_norm": 0.6611918807029724,
+      "learning_rate": 1.7321382139167578e-05,
+      "loss": 0.346,
+      "step": 6380
+    },
+    {
+      "epoch": 0.23870832067668293,
+      "grad_norm": 0.3350640833377838,
+      "learning_rate": 1.7317380736719588e-05,
+      "loss": 0.2905,
+      "step": 6385
+    },
+    {
+      "epoch": 0.23889524966703274,
+      "grad_norm": 0.6378026008605957,
+      "learning_rate": 1.7313376810716654e-05,
+      "loss": 0.2809,
+      "step": 6390
+    },
+    {
+      "epoch": 0.23908217865738252,
+      "grad_norm": 0.5519046783447266,
+      "learning_rate": 1.7309370362539607e-05,
+      "loss": 0.3125,
+      "step": 6395
+    },
+    {
+      "epoch": 0.23926910764773232,
+      "grad_norm": 0.4152654707431793,
+      "learning_rate": 1.7305361393570165e-05,
+      "loss": 0.267,
+      "step": 6400
+    },
+    {
+      "epoch": 0.2394560366380821,
+      "grad_norm": 0.47034916281700134,
+      "learning_rate": 1.7301349905190904e-05,
+      "loss": 0.2861,
+      "step": 6405
+    },
+    {
+      "epoch": 0.2396429656284319,
+      "grad_norm": 1.0928969383239746,
+      "learning_rate": 1.729733589878527e-05,
+      "loss": 0.306,
+      "step": 6410
+    },
+    {
+      "epoch": 0.23982989461878168,
+      "grad_norm": 0.28271111845970154,
+      "learning_rate": 1.729331937573758e-05,
+      "loss": 0.2999,
+      "step": 6415
+    },
+    {
+      "epoch": 0.24001682360913149,
+      "grad_norm": 0.27167975902557373,
+      "learning_rate": 1.7289300337433017e-05,
+      "loss": 0.2157,
+      "step": 6420
+    },
+    {
+      "epoch": 0.24020375259948126,
+      "grad_norm": 0.3268529176712036,
+      "learning_rate": 1.7285278785257633e-05,
+      "loss": 0.2802,
+      "step": 6425
+    },
+    {
+      "epoch": 0.24039068158983107,
+      "grad_norm": 0.6253595352172852,
+      "learning_rate": 1.7281254720598342e-05,
+      "loss": 0.3881,
+      "step": 6430
+    },
+    {
+      "epoch": 0.24057761058018085,
+      "grad_norm": 0.3022212088108063,
+      "learning_rate": 1.727722814484294e-05,
+      "loss": 0.3592,
+      "step": 6435
+    },
+    {
+      "epoch": 0.24076453957053065,
+      "grad_norm": 1.0882967710494995,
+      "learning_rate": 1.7273199059380062e-05,
+      "loss": 0.2548,
+      "step": 6440
+    },
+    {
+      "epoch": 0.24095146856088043,
+      "grad_norm": 0.23813073337078094,
+      "learning_rate": 1.7269167465599236e-05,
+      "loss": 0.2696,
+      "step": 6445
+    },
+    {
+      "epoch": 0.24113839755123023,
+      "grad_norm": 0.4797143042087555,
+      "learning_rate": 1.7265133364890842e-05,
+      "loss": 0.2992,
+      "step": 6450
+    },
+    {
+      "epoch": 0.24132532654158,
+      "grad_norm": 0.3314330577850342,
+      "learning_rate": 1.7261096758646115e-05,
+      "loss": 0.4107,
+      "step": 6455
+    },
+    {
+      "epoch": 0.24151225553192981,
+      "grad_norm": 0.38850629329681396,
+      "learning_rate": 1.7257057648257174e-05,
+      "loss": 0.3134,
+      "step": 6460
+    },
+    {
+      "epoch": 0.2416991845222796,
+      "grad_norm": 0.39222410321235657,
+      "learning_rate": 1.7253016035117e-05,
+      "loss": 0.2126,
+      "step": 6465
+    },
+    {
+      "epoch": 0.2418861135126294,
+      "grad_norm": 0.18110430240631104,
+      "learning_rate": 1.7248971920619413e-05,
+      "loss": 0.2886,
+      "step": 6470
+    },
+    {
+      "epoch": 0.24207304250297917,
+      "grad_norm": 0.47243568301200867,
+      "learning_rate": 1.724492530615912e-05,
+      "loss": 0.3672,
+      "step": 6475
+    },
+    {
+      "epoch": 0.24225997149332898,
+      "grad_norm": 0.3299618661403656,
+      "learning_rate": 1.7240876193131685e-05,
+      "loss": 0.3709,
+      "step": 6480
+    },
+    {
+      "epoch": 0.24244690048367876,
+      "grad_norm": 0.34539106488227844,
+      "learning_rate": 1.7236824582933525e-05,
+      "loss": 0.2648,
+      "step": 6485
+    },
+    {
+      "epoch": 0.24263382947402856,
+      "grad_norm": 0.5265419483184814,
+      "learning_rate": 1.7232770476961932e-05,
+      "loss": 0.2256,
+      "step": 6490
+    },
+    {
+      "epoch": 0.24282075846437834,
+      "grad_norm": 0.5372416377067566,
+      "learning_rate": 1.7228713876615043e-05,
+      "loss": 0.2817,
+      "step": 6495
+    },
+    {
+      "epoch": 0.24300768745472814,
+      "grad_norm": 0.4615037739276886,
+      "learning_rate": 1.7224654783291867e-05,
+      "loss": 0.2583,
+      "step": 6500
+    },
+    {
+      "epoch": 0.24319461644507792,
+      "grad_norm": 0.5872400403022766,
+      "learning_rate": 1.722059319839227e-05,
+      "loss": 0.3394,
+      "step": 6505
+    },
+    {
+      "epoch": 0.24338154543542773,
+      "grad_norm": 0.5134822726249695,
+      "learning_rate": 1.7216529123316975e-05,
+      "loss": 0.2918,
+      "step": 6510
+    },
+    {
+      "epoch": 0.2435684744257775,
+      "grad_norm": 0.3390646278858185,
+      "learning_rate": 1.7212462559467567e-05,
+      "loss": 0.338,
+      "step": 6515
+    },
+    {
+      "epoch": 0.2437554034161273,
+      "grad_norm": 0.24070733785629272,
+      "learning_rate": 1.7208393508246484e-05,
+      "loss": 0.3063,
+      "step": 6520
+    },
+    {
+      "epoch": 0.24394233240647709,
+      "grad_norm": 0.4138699471950531,
+      "learning_rate": 1.7204321971057024e-05,
+      "loss": 0.2647,
+      "step": 6525
+    },
+    {
+      "epoch": 0.2441292613968269,
+      "grad_norm": 0.5086038708686829,
+      "learning_rate": 1.720024794930335e-05,
+      "loss": 0.2562,
+      "step": 6530
+    },
+    {
+      "epoch": 0.24431619038717667,
+      "grad_norm": 0.8470866680145264,
+      "learning_rate": 1.719617144439047e-05,
+      "loss": 0.276,
+      "step": 6535
+    },
+    {
+      "epoch": 0.24450311937752647,
+      "grad_norm": 0.5553832650184631,
+      "learning_rate": 1.7192092457724254e-05,
+      "loss": 0.2874,
+      "step": 6540
+    },
+    {
+      "epoch": 0.24469004836787625,
+      "grad_norm": 0.32335466146469116,
+      "learning_rate": 1.718801099071143e-05,
+      "loss": 0.2903,
+      "step": 6545
+    },
+    {
+      "epoch": 0.24487697735822606,
+      "grad_norm": 0.46024224162101746,
+      "learning_rate": 1.7183927044759576e-05,
+      "loss": 0.3618,
+      "step": 6550
+    },
+    {
+      "epoch": 0.24506390634857583,
+      "grad_norm": 0.29070380330085754,
+      "learning_rate": 1.7179840621277132e-05,
+      "loss": 0.4624,
+      "step": 6555
+    },
+    {
+      "epoch": 0.24525083533892564,
+      "grad_norm": 0.35333648324012756,
+      "learning_rate": 1.7175751721673384e-05,
+      "loss": 0.2641,
+      "step": 6560
+    },
+    {
+      "epoch": 0.24543776432927542,
+      "grad_norm": 0.5118494629859924,
+      "learning_rate": 1.7171660347358482e-05,
+      "loss": 0.289,
+      "step": 6565
+    },
+    {
+      "epoch": 0.24562469331962522,
+      "grad_norm": 0.4456328749656677,
+      "learning_rate": 1.7167566499743417e-05,
+      "loss": 0.2784,
+      "step": 6570
+    },
+    {
+      "epoch": 0.245811622309975,
+      "grad_norm": 0.42578622698783875,
+      "learning_rate": 1.7163470180240047e-05,
+      "loss": 0.242,
+      "step": 6575
+    },
+    {
+      "epoch": 0.2459985513003248,
+      "grad_norm": 0.38927942514419556,
+      "learning_rate": 1.7159371390261067e-05,
+      "loss": 0.3147,
+      "step": 6580
+    },
+    {
+      "epoch": 0.24618548029067458,
+      "grad_norm": 0.4326514005661011,
+      "learning_rate": 1.715527013122004e-05,
+      "loss": 0.3499,
+      "step": 6585
+    },
+    {
+      "epoch": 0.24637240928102436,
+      "grad_norm": 0.4785178601741791,
+      "learning_rate": 1.7151166404531365e-05,
+      "loss": 0.3341,
+      "step": 6590
+    },
+    {
+      "epoch": 0.24655933827137416,
+      "grad_norm": 0.09519144147634506,
+      "learning_rate": 1.7147060211610305e-05,
+      "loss": 0.2694,
+      "step": 6595
+    },
+    {
+      "epoch": 0.24674626726172394,
+      "grad_norm": 0.4483666718006134,
+      "learning_rate": 1.7142951553872968e-05,
+      "loss": 0.4049,
+      "step": 6600
+    },
+    {
+      "epoch": 0.24693319625207374,
+      "grad_norm": 0.2505245804786682,
+      "learning_rate": 1.713884043273631e-05,
+      "loss": 0.3139,
+      "step": 6605
+    },
+    {
+      "epoch": 0.24712012524242352,
+      "grad_norm": 0.5089213848114014,
+      "learning_rate": 1.7134726849618144e-05,
+      "loss": 0.3026,
+      "step": 6610
+    },
+    {
+      "epoch": 0.24730705423277333,
+      "grad_norm": 0.3417368531227112,
+      "learning_rate": 1.7130610805937123e-05,
+      "loss": 0.219,
+      "step": 6615
+    },
+    {
+      "epoch": 0.2474939832231231,
+      "grad_norm": 0.428158164024353,
+      "learning_rate": 1.712649230311275e-05,
+      "loss": 0.3102,
+      "step": 6620
+    },
+    {
+      "epoch": 0.2476809122134729,
+      "grad_norm": 0.4470468759536743,
+      "learning_rate": 1.7122371342565384e-05,
+      "loss": 0.2733,
+      "step": 6625
+    },
+    {
+      "epoch": 0.2478678412038227,
+      "grad_norm": 0.9151923060417175,
+      "learning_rate": 1.7118247925716223e-05,
+      "loss": 0.3076,
+      "step": 6630
+    },
+    {
+      "epoch": 0.2480547701941725,
+      "grad_norm": 0.7877910137176514,
+      "learning_rate": 1.7114122053987318e-05,
+      "loss": 0.3281,
+      "step": 6635
+    },
+    {
+      "epoch": 0.24824169918452227,
+      "grad_norm": 0.47747930884361267,
+      "learning_rate": 1.7109993728801556e-05,
+      "loss": 0.258,
+      "step": 6640
+    },
+    {
+      "epoch": 0.24842862817487207,
+      "grad_norm": 0.3599812388420105,
+      "learning_rate": 1.710586295158269e-05,
+      "loss": 0.2615,
+      "step": 6645
+    },
+    {
+      "epoch": 0.24861555716522185,
+      "grad_norm": 0.34985432028770447,
+      "learning_rate": 1.7101729723755296e-05,
+      "loss": 0.3707,
+      "step": 6650
+    },
+    {
+      "epoch": 0.24880248615557166,
+      "grad_norm": 0.48352763056755066,
+      "learning_rate": 1.7097594046744815e-05,
+      "loss": 0.3285,
+      "step": 6655
+    },
+    {
+      "epoch": 0.24898941514592143,
+      "grad_norm": 0.3733879625797272,
+      "learning_rate": 1.7093455921977516e-05,
+      "loss": 0.3347,
+      "step": 6660
+    },
+    {
+      "epoch": 0.24917634413627124,
+      "grad_norm": 1.183453917503357,
+      "learning_rate": 1.7089315350880525e-05,
+      "loss": 0.395,
+      "step": 6665
+    },
+    {
+      "epoch": 0.24936327312662102,
+      "grad_norm": 0.3995364010334015,
+      "learning_rate": 1.70851723348818e-05,
+      "loss": 0.3719,
+      "step": 6670
+    },
+    {
+      "epoch": 0.24955020211697082,
+      "grad_norm": 0.2912960350513458,
+      "learning_rate": 1.7081026875410156e-05,
+      "loss": 0.2581,
+      "step": 6675
+    },
+    {
+      "epoch": 0.2497371311073206,
+      "grad_norm": 0.32783398032188416,
+      "learning_rate": 1.7076878973895242e-05,
+      "loss": 0.3313,
+      "step": 6680
+    },
+    {
+      "epoch": 0.2499240600976704,
+      "grad_norm": 0.45547330379486084,
+      "learning_rate": 1.7072728631767543e-05,
+      "loss": 0.2585,
+      "step": 6685
+    },
+    {
+      "epoch": 0.2501109890880202,
+      "grad_norm": 0.4544295370578766,
+      "learning_rate": 1.7068575850458402e-05,
+      "loss": 0.2982,
+      "step": 6690
+    },
+    {
+      "epoch": 0.25029791807837,
+      "grad_norm": 0.5787556171417236,
+      "learning_rate": 1.7064420631399986e-05,
+      "loss": 0.2737,
+      "step": 6695
+    },
+    {
+      "epoch": 0.25048484706871976,
+      "grad_norm": 0.4456428587436676,
+      "learning_rate": 1.706026297602532e-05,
+      "loss": 0.2809,
+      "step": 6700
+    },
+    {
+      "epoch": 0.25067177605906954,
+      "grad_norm": 0.27072015404701233,
+      "learning_rate": 1.705610288576825e-05,
+      "loss": 0.2661,
+      "step": 6705
+    },
+    {
+      "epoch": 0.2508587050494194,
+      "grad_norm": 0.2884485721588135,
+      "learning_rate": 1.7051940362063486e-05,
+      "loss": 0.3243,
+      "step": 6710
+    },
+    {
+      "epoch": 0.25104563403976915,
+      "grad_norm": 0.3565753400325775,
+      "learning_rate": 1.7047775406346548e-05,
+      "loss": 0.4407,
+      "step": 6715
+    },
+    {
+      "epoch": 0.2512325630301189,
+      "grad_norm": 0.5307539105415344,
+      "learning_rate": 1.7043608020053823e-05,
+      "loss": 0.3288,
+      "step": 6720
+    },
+    {
+      "epoch": 0.2514194920204687,
+      "grad_norm": 0.2853598892688751,
+      "learning_rate": 1.7039438204622515e-05,
+      "loss": 0.3014,
+      "step": 6725
+    },
+    {
+      "epoch": 0.25160642101081854,
+      "grad_norm": 0.476520836353302,
+      "learning_rate": 1.7035265961490673e-05,
+      "loss": 0.337,
+      "step": 6730
+    },
+    {
+      "epoch": 0.2517933500011683,
+      "grad_norm": 0.34905505180358887,
+      "learning_rate": 1.7031091292097186e-05,
+      "loss": 0.2644,
+      "step": 6735
+    },
+    {
+      "epoch": 0.2519802789915181,
+      "grad_norm": 0.4414178729057312,
+      "learning_rate": 1.702691419788178e-05,
+      "loss": 0.3584,
+      "step": 6740
+    },
+    {
+      "epoch": 0.25216720798186787,
+      "grad_norm": 0.4812869131565094,
+      "learning_rate": 1.7022734680285013e-05,
+      "loss": 0.2968,
+      "step": 6745
+    },
+    {
+      "epoch": 0.2523541369722177,
+      "grad_norm": 0.5577378869056702,
+      "learning_rate": 1.701855274074828e-05,
+      "loss": 0.2407,
+      "step": 6750
+    },
+    {
+      "epoch": 0.2525410659625675,
+      "grad_norm": 0.3004097044467926,
+      "learning_rate": 1.701436838071382e-05,
+      "loss": 0.2893,
+      "step": 6755
+    },
+    {
+      "epoch": 0.25272799495291726,
+      "grad_norm": 0.5560919046401978,
+      "learning_rate": 1.7010181601624687e-05,
+      "loss": 0.305,
+      "step": 6760
+    },
+    {
+      "epoch": 0.25291492394326703,
+      "grad_norm": 0.5027373433113098,
+      "learning_rate": 1.700599240492479e-05,
+      "loss": 0.2415,
+      "step": 6765
+    },
+    {
+      "epoch": 0.25310185293361687,
+      "grad_norm": 0.2853597104549408,
+      "learning_rate": 1.7001800792058856e-05,
+      "loss": 0.236,
+      "step": 6770
+    },
+    {
+      "epoch": 0.25328878192396664,
+      "grad_norm": 0.36475807428359985,
+      "learning_rate": 1.6997606764472457e-05,
+      "loss": 0.2476,
+      "step": 6775
+    },
+    {
+      "epoch": 0.2534757109143164,
+      "grad_norm": 0.2956780791282654,
+      "learning_rate": 1.6993410323611993e-05,
+      "loss": 0.3781,
+      "step": 6780
+    },
+    {
+      "epoch": 0.2536626399046662,
+      "grad_norm": 0.3027787208557129,
+      "learning_rate": 1.6989211470924694e-05,
+      "loss": 0.2951,
+      "step": 6785
+    },
+    {
+      "epoch": 0.25384956889501603,
+      "grad_norm": 0.36557620763778687,
+      "learning_rate": 1.6985010207858624e-05,
+      "loss": 0.2903,
+      "step": 6790
+    },
+    {
+      "epoch": 0.2540364978853658,
+      "grad_norm": 0.2095050811767578,
+      "learning_rate": 1.6980806535862683e-05,
+      "loss": 0.2859,
+      "step": 6795
+    },
+    {
+      "epoch": 0.2542234268757156,
+      "grad_norm": 0.4164871573448181,
+      "learning_rate": 1.6976600456386593e-05,
+      "loss": 0.24,
+      "step": 6800
+    },
+    {
+      "epoch": 0.25441035586606536,
+      "grad_norm": 0.2171694040298462,
+      "learning_rate": 1.6972391970880906e-05,
+      "loss": 0.4062,
+      "step": 6805
+    },
+    {
+      "epoch": 0.2545972848564152,
+      "grad_norm": 0.6664031744003296,
+      "learning_rate": 1.6968181080797012e-05,
+      "loss": 0.2982,
+      "step": 6810
+    },
+    {
+      "epoch": 0.254784213846765,
+      "grad_norm": 0.3303912878036499,
+      "learning_rate": 1.6963967787587133e-05,
+      "loss": 0.2349,
+      "step": 6815
+    },
+    {
+      "epoch": 0.25497114283711475,
+      "grad_norm": 0.42287972569465637,
+      "learning_rate": 1.6959752092704302e-05,
+      "loss": 0.3224,
+      "step": 6820
+    },
+    {
+      "epoch": 0.2551580718274645,
+      "grad_norm": 0.5297410488128662,
+      "learning_rate": 1.69555339976024e-05,
+      "loss": 0.3097,
+      "step": 6825
+    },
+    {
+      "epoch": 0.25534500081781436,
+      "grad_norm": 0.44122767448425293,
+      "learning_rate": 1.695131350373612e-05,
+      "loss": 0.2884,
+      "step": 6830
+    },
+    {
+      "epoch": 0.25553192980816414,
+      "grad_norm": 0.4497906267642975,
+      "learning_rate": 1.6947090612560995e-05,
+      "loss": 0.2767,
+      "step": 6835
+    },
+    {
+      "epoch": 0.2557188587985139,
+      "grad_norm": 0.383256196975708,
+      "learning_rate": 1.6942865325533374e-05,
+      "loss": 0.3156,
+      "step": 6840
+    },
+    {
+      "epoch": 0.2559057877888637,
+      "grad_norm": 0.45405879616737366,
+      "learning_rate": 1.6938637644110442e-05,
+      "loss": 0.3549,
+      "step": 6845
+    },
+    {
+      "epoch": 0.25609271677921347,
+      "grad_norm": 0.26836884021759033,
+      "learning_rate": 1.6934407569750208e-05,
+      "loss": 0.265,
+      "step": 6850
+    },
+    {
+      "epoch": 0.2562796457695633,
+      "grad_norm": 0.42300936579704285,
+      "learning_rate": 1.6930175103911492e-05,
+      "loss": 0.3418,
+      "step": 6855
+    },
+    {
+      "epoch": 0.2564665747599131,
+      "grad_norm": 0.4314024746417999,
+      "learning_rate": 1.6925940248053964e-05,
+      "loss": 0.3864,
+      "step": 6860
+    },
+    {
+      "epoch": 0.25665350375026286,
+      "grad_norm": 0.4368232488632202,
+      "learning_rate": 1.6921703003638094e-05,
+      "loss": 0.3285,
+      "step": 6865
+    },
+    {
+      "epoch": 0.25684043274061263,
+      "grad_norm": 0.33699876070022583,
+      "learning_rate": 1.691746337212519e-05,
+      "loss": 0.3067,
+      "step": 6870
+    },
+    {
+      "epoch": 0.25702736173096247,
+      "grad_norm": 0.28076720237731934,
+      "learning_rate": 1.691322135497738e-05,
+      "loss": 0.2381,
+      "step": 6875
+    },
+    {
+      "epoch": 0.25721429072131224,
+      "grad_norm": 0.5777077674865723,
+      "learning_rate": 1.6908976953657612e-05,
+      "loss": 0.2946,
+      "step": 6880
+    },
+    {
+      "epoch": 0.257401219711662,
+      "grad_norm": 0.43163543939590454,
+      "learning_rate": 1.690473016962966e-05,
+      "loss": 0.307,
+      "step": 6885
+    },
+    {
+      "epoch": 0.2575881487020118,
+      "grad_norm": 0.4134599268436432,
+      "learning_rate": 1.6900481004358123e-05,
+      "loss": 0.2846,
+      "step": 6890
+    },
+    {
+      "epoch": 0.25777507769236163,
+      "grad_norm": 0.40419724583625793,
+      "learning_rate": 1.6896229459308405e-05,
+      "loss": 0.2843,
+      "step": 6895
+    },
+    {
+      "epoch": 0.2579620066827114,
+      "grad_norm": 0.4465784728527069,
+      "learning_rate": 1.6891975535946753e-05,
+      "loss": 0.277,
+      "step": 6900
+    },
+    {
+      "epoch": 0.2581489356730612,
+      "grad_norm": 0.39851880073547363,
+      "learning_rate": 1.6887719235740216e-05,
+      "loss": 0.2687,
+      "step": 6905
+    },
+    {
+      "epoch": 0.25833586466341096,
+      "grad_norm": 0.30198317766189575,
+      "learning_rate": 1.6883460560156673e-05,
+      "loss": 0.2999,
+      "step": 6910
+    },
+    {
+      "epoch": 0.2585227936537608,
+      "grad_norm": 0.43067798018455505,
+      "learning_rate": 1.687919951066482e-05,
+      "loss": 0.3473,
+      "step": 6915
+    },
+    {
+      "epoch": 0.2587097226441106,
+      "grad_norm": 0.2943825423717499,
+      "learning_rate": 1.687493608873417e-05,
+      "loss": 0.3303,
+      "step": 6920
+    },
+    {
+      "epoch": 0.25889665163446035,
+      "grad_norm": 0.3708643317222595,
+      "learning_rate": 1.6870670295835055e-05,
+      "loss": 0.3585,
+      "step": 6925
+    },
+    {
+      "epoch": 0.25908358062481013,
+      "grad_norm": 0.45866164565086365,
+      "learning_rate": 1.6866402133438623e-05,
+      "loss": 0.3606,
+      "step": 6930
+    },
+    {
+      "epoch": 0.25927050961515996,
+      "grad_norm": 0.4115564227104187,
+      "learning_rate": 1.6862131603016844e-05,
+      "loss": 0.2645,
+      "step": 6935
+    },
+    {
+      "epoch": 0.25945743860550974,
+      "grad_norm": 0.21493466198444366,
+      "learning_rate": 1.68578587060425e-05,
+      "loss": 0.312,
+      "step": 6940
+    },
+    {
+      "epoch": 0.2596443675958595,
+      "grad_norm": 0.8563496470451355,
+      "learning_rate": 1.6853583443989186e-05,
+      "loss": 0.362,
+      "step": 6945
+    },
+    {
+      "epoch": 0.2598312965862093,
+      "grad_norm": 0.36561504006385803,
+      "learning_rate": 1.684930581833133e-05,
+      "loss": 0.3779,
+      "step": 6950
+    },
+    {
+      "epoch": 0.2600182255765591,
+      "grad_norm": 0.37505069375038147,
+      "learning_rate": 1.6845025830544147e-05,
+      "loss": 0.301,
+      "step": 6955
+    },
+    {
+      "epoch": 0.2602051545669089,
+      "grad_norm": 0.45678603649139404,
+      "learning_rate": 1.68407434821037e-05,
+      "loss": 0.3282,
+      "step": 6960
+    },
+    {
+      "epoch": 0.2603920835572587,
+      "grad_norm": 0.5878984332084656,
+      "learning_rate": 1.6836458774486827e-05,
+      "loss": 0.3805,
+      "step": 6965
+    },
+    {
+      "epoch": 0.26057901254760846,
+      "grad_norm": 0.20040619373321533,
+      "learning_rate": 1.683217170917122e-05,
+      "loss": 0.2674,
+      "step": 6970
+    },
+    {
+      "epoch": 0.2607659415379583,
+      "grad_norm": 0.45022228360176086,
+      "learning_rate": 1.6827882287635353e-05,
+      "loss": 0.315,
+      "step": 6975
+    },
+    {
+      "epoch": 0.26095287052830807,
+      "grad_norm": 0.39582857489585876,
+      "learning_rate": 1.682359051135853e-05,
+      "loss": 0.3558,
+      "step": 6980
+    },
+    {
+      "epoch": 0.26113979951865784,
+      "grad_norm": 0.5623447895050049,
+      "learning_rate": 1.681929638182086e-05,
+      "loss": 0.2928,
+      "step": 6985
+    },
+    {
+      "epoch": 0.2613267285090076,
+      "grad_norm": 0.35395896434783936,
+      "learning_rate": 1.6814999900503265e-05,
+      "loss": 0.3276,
+      "step": 6990
+    },
+    {
+      "epoch": 0.26151365749935745,
+      "grad_norm": 0.2845790684223175,
+      "learning_rate": 1.681070106888748e-05,
+      "loss": 0.3025,
+      "step": 6995
+    },
+    {
+      "epoch": 0.26170058648970723,
+      "grad_norm": 0.3993929326534271,
+      "learning_rate": 1.6806399888456043e-05,
+      "loss": 0.327,
+      "step": 7000
+    },
+    {
+      "epoch": 0.261887515480057,
+      "grad_norm": 0.23860380053520203,
+      "learning_rate": 1.6802096360692316e-05,
+      "loss": 0.3101,
+      "step": 7005
+    },
+    {
+      "epoch": 0.2620744444704068,
+      "grad_norm": 0.20932121574878693,
+      "learning_rate": 1.679779048708046e-05,
+      "loss": 0.2384,
+      "step": 7010
+    },
+    {
+      "epoch": 0.2622613734607566,
+      "grad_norm": 0.4419725239276886,
+      "learning_rate": 1.6793482269105446e-05,
+      "loss": 0.2384,
+      "step": 7015
+    },
+    {
+      "epoch": 0.2624483024511064,
+      "grad_norm": 0.30131080746650696,
+      "learning_rate": 1.6789171708253052e-05,
+      "loss": 0.3222,
+      "step": 7020
+    },
+    {
+      "epoch": 0.2626352314414562,
+      "grad_norm": 0.2808610498905182,
+      "learning_rate": 1.6784858806009875e-05,
+      "loss": 0.249,
+      "step": 7025
+    },
+    {
+      "epoch": 0.26282216043180595,
+      "grad_norm": 0.4096542298793793,
+      "learning_rate": 1.67805435638633e-05,
+      "loss": 0.2912,
+      "step": 7030
+    },
+    {
+      "epoch": 0.2630090894221558,
+      "grad_norm": 0.32321956753730774,
+      "learning_rate": 1.6776225983301543e-05,
+      "loss": 0.2665,
+      "step": 7035
+    },
+    {
+      "epoch": 0.26319601841250556,
+      "grad_norm": 0.27172431349754333,
+      "learning_rate": 1.6771906065813607e-05,
+      "loss": 0.2614,
+      "step": 7040
+    },
+    {
+      "epoch": 0.26338294740285534,
+      "grad_norm": 0.2878829836845398,
+      "learning_rate": 1.676758381288931e-05,
+      "loss": 0.2974,
+      "step": 7045
+    },
+    {
+      "epoch": 0.2635698763932051,
+      "grad_norm": 0.3150484263896942,
+      "learning_rate": 1.6763259226019267e-05,
+      "loss": 0.2796,
+      "step": 7050
+    },
+    {
+      "epoch": 0.26375680538355495,
+      "grad_norm": 0.39884239435195923,
+      "learning_rate": 1.6758932306694913e-05,
+      "loss": 0.261,
+      "step": 7055
+    },
+    {
+      "epoch": 0.2639437343739047,
+      "grad_norm": 0.42523136734962463,
+      "learning_rate": 1.6754603056408473e-05,
+      "loss": 0.3706,
+      "step": 7060
+    },
+    {
+      "epoch": 0.2641306633642545,
+      "grad_norm": 0.5379177927970886,
+      "learning_rate": 1.675027147665298e-05,
+      "loss": 0.275,
+      "step": 7065
+    },
+    {
+      "epoch": 0.2643175923546043,
+      "grad_norm": 1.383110761642456,
+      "learning_rate": 1.674593756892228e-05,
+      "loss": 0.2515,
+      "step": 7070
+    },
+    {
+      "epoch": 0.2645045213449541,
+      "grad_norm": 0.3739633560180664,
+      "learning_rate": 1.6741601334711004e-05,
+      "loss": 0.3131,
+      "step": 7075
+    },
+    {
+      "epoch": 0.2646914503353039,
+      "grad_norm": 0.485005646944046,
+      "learning_rate": 1.67372627755146e-05,
+      "loss": 0.2868,
+      "step": 7080
+    },
+    {
+      "epoch": 0.26487837932565367,
+      "grad_norm": 0.4296754002571106,
+      "learning_rate": 1.6732921892829313e-05,
+      "loss": 0.3443,
+      "step": 7085
+    },
+    {
+      "epoch": 0.26506530831600345,
+      "grad_norm": 0.6618226170539856,
+      "learning_rate": 1.6728578688152186e-05,
+      "loss": 0.3642,
+      "step": 7090
+    },
+    {
+      "epoch": 0.2652522373063532,
+      "grad_norm": 0.3030281662940979,
+      "learning_rate": 1.6724233162981067e-05,
+      "loss": 0.2784,
+      "step": 7095
+    },
+    {
+      "epoch": 0.26543916629670306,
+      "grad_norm": 0.5461763143539429,
+      "learning_rate": 1.6719885318814604e-05,
+      "loss": 0.4215,
+      "step": 7100
+    },
+    {
+      "epoch": 0.26562609528705283,
+      "grad_norm": 0.38784629106521606,
+      "learning_rate": 1.6715535157152244e-05,
+      "loss": 0.3164,
+      "step": 7105
+    },
+    {
+      "epoch": 0.2658130242774026,
+      "grad_norm": 0.46060433983802795,
+      "learning_rate": 1.6711182679494232e-05,
+      "loss": 0.2347,
+      "step": 7110
+    },
+    {
+      "epoch": 0.2659999532677524,
+      "grad_norm": 0.7013404369354248,
+      "learning_rate": 1.6706827887341613e-05,
+      "loss": 0.2908,
+      "step": 7115
+    },
+    {
+      "epoch": 0.2661868822581022,
+      "grad_norm": 0.7355599999427795,
+      "learning_rate": 1.6702470782196237e-05,
+      "loss": 0.2816,
+      "step": 7120
+    },
+    {
+      "epoch": 0.266373811248452,
+      "grad_norm": 0.23429067432880402,
+      "learning_rate": 1.6698111365560733e-05,
+      "loss": 0.3136,
+      "step": 7125
+    },
+    {
+      "epoch": 0.2665607402388018,
+      "grad_norm": 0.7480411529541016,
+      "learning_rate": 1.6693749638938544e-05,
+      "loss": 0.2388,
+      "step": 7130
+    },
+    {
+      "epoch": 0.26674766922915155,
+      "grad_norm": 0.6238270998001099,
+      "learning_rate": 1.6689385603833907e-05,
+      "loss": 0.4418,
+      "step": 7135
+    },
+    {
+      "epoch": 0.2669345982195014,
+      "grad_norm": 0.15828560292720795,
+      "learning_rate": 1.6685019261751848e-05,
+      "loss": 0.251,
+      "step": 7140
+    },
+    {
+      "epoch": 0.26712152720985116,
+      "grad_norm": 0.44209611415863037,
+      "learning_rate": 1.6680650614198194e-05,
+      "loss": 0.2724,
+      "step": 7145
+    },
+    {
+      "epoch": 0.26730845620020094,
+      "grad_norm": 0.49470680952072144,
+      "learning_rate": 1.667627966267957e-05,
+      "loss": 0.2514,
+      "step": 7150
+    },
+    {
+      "epoch": 0.2674953851905507,
+      "grad_norm": 0.3461754620075226,
+      "learning_rate": 1.6671906408703394e-05,
+      "loss": 0.2894,
+      "step": 7155
+    },
+    {
+      "epoch": 0.26768231418090055,
+      "grad_norm": 0.31336525082588196,
+      "learning_rate": 1.6667530853777865e-05,
+      "loss": 0.2696,
+      "step": 7160
+    },
+    {
+      "epoch": 0.2678692431712503,
+      "grad_norm": 0.5694870352745056,
+      "learning_rate": 1.6663152999411998e-05,
+      "loss": 0.2751,
+      "step": 7165
+    },
+    {
+      "epoch": 0.2680561721616001,
+      "grad_norm": 0.8204647898674011,
+      "learning_rate": 1.6658772847115584e-05,
+      "loss": 0.4243,
+      "step": 7170
+    },
+    {
+      "epoch": 0.2682431011519499,
+      "grad_norm": 0.44578155875205994,
+      "learning_rate": 1.6654390398399213e-05,
+      "loss": 0.2863,
+      "step": 7175
+    },
+    {
+      "epoch": 0.2684300301422997,
+      "grad_norm": 0.42481744289398193,
+      "learning_rate": 1.6650005654774265e-05,
+      "loss": 0.2577,
+      "step": 7180
+    },
+    {
+      "epoch": 0.2686169591326495,
+      "grad_norm": 0.5261092782020569,
+      "learning_rate": 1.6645618617752914e-05,
+      "loss": 0.2702,
+      "step": 7185
+    },
+    {
+      "epoch": 0.26880388812299927,
+      "grad_norm": 0.468940794467926,
+      "learning_rate": 1.6641229288848123e-05,
+      "loss": 0.3976,
+      "step": 7190
+    },
+    {
+      "epoch": 0.26899081711334905,
+      "grad_norm": 0.5336934328079224,
+      "learning_rate": 1.6636837669573647e-05,
+      "loss": 0.3726,
+      "step": 7195
+    },
+    {
+      "epoch": 0.2691777461036989,
+      "grad_norm": 0.2591412365436554,
+      "learning_rate": 1.6632443761444027e-05,
+      "loss": 0.3307,
+      "step": 7200
+    },
+    {
+      "epoch": 0.26936467509404866,
+      "grad_norm": 0.5266516804695129,
+      "learning_rate": 1.6628047565974594e-05,
+      "loss": 0.307,
+      "step": 7205
+    },
+    {
+      "epoch": 0.26955160408439843,
+      "grad_norm": 0.38717594742774963,
+      "learning_rate": 1.6623649084681477e-05,
+      "loss": 0.3238,
+      "step": 7210
+    },
+    {
+      "epoch": 0.2697385330747482,
+      "grad_norm": 0.17658643424510956,
+      "learning_rate": 1.6619248319081583e-05,
+      "loss": 0.2985,
+      "step": 7215
+    },
+    {
+      "epoch": 0.26992546206509804,
+      "grad_norm": 0.30146273970603943,
+      "learning_rate": 1.6614845270692606e-05,
+      "loss": 0.294,
+      "step": 7220
+    },
+    {
+      "epoch": 0.2701123910554478,
+      "grad_norm": 0.6402953863143921,
+      "learning_rate": 1.6610439941033038e-05,
+      "loss": 0.2622,
+      "step": 7225
+    },
+    {
+      "epoch": 0.2702993200457976,
+      "grad_norm": 0.5695781111717224,
+      "learning_rate": 1.6606032331622148e-05,
+      "loss": 0.2279,
+      "step": 7230
+    },
+    {
+      "epoch": 0.2704862490361474,
+      "grad_norm": 0.23263224959373474,
+      "learning_rate": 1.6601622443979987e-05,
+      "loss": 0.2919,
+      "step": 7235
+    },
+    {
+      "epoch": 0.2706731780264972,
+      "grad_norm": 0.4788752496242523,
+      "learning_rate": 1.659721027962741e-05,
+      "loss": 0.2335,
+      "step": 7240
+    },
+    {
+      "epoch": 0.270860107016847,
+      "grad_norm": 0.4149765372276306,
+      "learning_rate": 1.659279584008604e-05,
+      "loss": 0.3214,
+      "step": 7245
+    },
+    {
+      "epoch": 0.27104703600719676,
+      "grad_norm": 0.5265377759933472,
+      "learning_rate": 1.6588379126878293e-05,
+      "loss": 0.3483,
+      "step": 7250
+    },
+    {
+      "epoch": 0.27123396499754654,
+      "grad_norm": 0.6144738793373108,
+      "learning_rate": 1.6583960141527367e-05,
+      "loss": 0.424,
+      "step": 7255
+    },
+    {
+      "epoch": 0.27142089398789637,
+      "grad_norm": 0.3015303909778595,
+      "learning_rate": 1.6579538885557242e-05,
+      "loss": 0.343,
+      "step": 7260
+    },
+    {
+      "epoch": 0.27160782297824615,
+      "grad_norm": 0.4408928155899048,
+      "learning_rate": 1.6575115360492683e-05,
+      "loss": 0.2704,
+      "step": 7265
+    },
+    {
+      "epoch": 0.2717947519685959,
+      "grad_norm": 0.3285323679447174,
+      "learning_rate": 1.6570689567859237e-05,
+      "loss": 0.2789,
+      "step": 7270
+    },
+    {
+      "epoch": 0.2719816809589457,
+      "grad_norm": 0.09705019742250443,
+      "learning_rate": 1.6566261509183232e-05,
+      "loss": 0.3737,
+      "step": 7275
+    },
+    {
+      "epoch": 0.27216860994929554,
+      "grad_norm": 0.24546971917152405,
+      "learning_rate": 1.6561831185991782e-05,
+      "loss": 0.2472,
+      "step": 7280
+    },
+    {
+      "epoch": 0.2723555389396453,
+      "grad_norm": 0.1780613362789154,
+      "learning_rate": 1.6557398599812774e-05,
+      "loss": 0.2864,
+      "step": 7285
+    },
+    {
+      "epoch": 0.2725424679299951,
+      "grad_norm": 0.5108126997947693,
+      "learning_rate": 1.655296375217488e-05,
+      "loss": 0.3593,
+      "step": 7290
+    },
+    {
+      "epoch": 0.27272939692034487,
+      "grad_norm": 0.14681376516819,
+      "learning_rate": 1.654852664460756e-05,
+      "loss": 0.318,
+      "step": 7295
+    },
+    {
+      "epoch": 0.2729163259106947,
+      "grad_norm": 0.42053887248039246,
+      "learning_rate": 1.6544087278641037e-05,
+      "loss": 0.2802,
+      "step": 7300
+    },
+    {
+      "epoch": 0.2731032549010445,
+      "grad_norm": 0.5258027911186218,
+      "learning_rate": 1.653964565580632e-05,
+      "loss": 0.3422,
+      "step": 7305
+    },
+    {
+      "epoch": 0.27329018389139426,
+      "grad_norm": 0.2713332772254944,
+      "learning_rate": 1.6535201777635206e-05,
+      "loss": 0.3497,
+      "step": 7310
+    },
+    {
+      "epoch": 0.27347711288174403,
+      "grad_norm": 0.5129302740097046,
+      "learning_rate": 1.6530755645660254e-05,
+      "loss": 0.2799,
+      "step": 7315
+    },
+    {
+      "epoch": 0.2736640418720938,
+      "grad_norm": 0.33113789558410645,
+      "learning_rate": 1.6526307261414812e-05,
+      "loss": 0.351,
+      "step": 7320
+    },
+    {
+      "epoch": 0.27385097086244364,
+      "grad_norm": 0.3590730130672455,
+      "learning_rate": 1.6521856626432992e-05,
+      "loss": 0.3425,
+      "step": 7325
+    },
+    {
+      "epoch": 0.2740378998527934,
+      "grad_norm": 0.2738291323184967,
+      "learning_rate": 1.65174037422497e-05,
+      "loss": 0.3775,
+      "step": 7330
+    },
+    {
+      "epoch": 0.2742248288431432,
+      "grad_norm": 0.3311813771724701,
+      "learning_rate": 1.6512948610400606e-05,
+      "loss": 0.3176,
+      "step": 7335
+    },
+    {
+      "epoch": 0.274411757833493,
+      "grad_norm": 0.2696897089481354,
+      "learning_rate": 1.6508491232422153e-05,
+      "loss": 0.314,
+      "step": 7340
+    },
+    {
+      "epoch": 0.2745986868238428,
+      "grad_norm": 0.2976570725440979,
+      "learning_rate": 1.6504031609851567e-05,
+      "loss": 0.4029,
+      "step": 7345
+    },
+    {
+      "epoch": 0.2747856158141926,
+      "grad_norm": 0.25655868649482727,
+      "learning_rate": 1.6499569744226843e-05,
+      "loss": 0.3326,
+      "step": 7350
+    },
+    {
+      "epoch": 0.27497254480454236,
+      "grad_norm": 0.21505354344844818,
+      "learning_rate": 1.649510563708675e-05,
+      "loss": 0.2943,
+      "step": 7355
+    },
+    {
+      "epoch": 0.27515947379489214,
+      "grad_norm": 0.4883136749267578,
+      "learning_rate": 1.6490639289970834e-05,
+      "loss": 0.2989,
+      "step": 7360
+    },
+    {
+      "epoch": 0.275346402785242,
+      "grad_norm": 0.26657021045684814,
+      "learning_rate": 1.6486170704419402e-05,
+      "loss": 0.2817,
+      "step": 7365
+    },
+    {
+      "epoch": 0.27553333177559175,
+      "grad_norm": 0.40542852878570557,
+      "learning_rate": 1.648169988197355e-05,
+      "loss": 0.2866,
+      "step": 7370
+    },
+    {
+      "epoch": 0.2757202607659415,
+      "grad_norm": 0.901739776134491,
+      "learning_rate": 1.647722682417513e-05,
+      "loss": 0.2885,
+      "step": 7375
+    },
+    {
+      "epoch": 0.2759071897562913,
+      "grad_norm": 0.42218196392059326,
+      "learning_rate": 1.6472751532566777e-05,
+      "loss": 0.225,
+      "step": 7380
+    },
+    {
+      "epoch": 0.27609411874664114,
+      "grad_norm": 0.5886632800102234,
+      "learning_rate": 1.6468274008691888e-05,
+      "loss": 0.3544,
+      "step": 7385
+    },
+    {
+      "epoch": 0.2762810477369909,
+      "grad_norm": 0.4941720962524414,
+      "learning_rate": 1.646379425409463e-05,
+      "loss": 0.3206,
+      "step": 7390
+    },
+    {
+      "epoch": 0.2764679767273407,
+      "grad_norm": 0.3011486530303955,
+      "learning_rate": 1.6459312270319946e-05,
+      "loss": 0.2848,
+      "step": 7395
+    },
+    {
+      "epoch": 0.27665490571769047,
+      "grad_norm": 0.5940350890159607,
+      "learning_rate": 1.6454828058913544e-05,
+      "loss": 0.3851,
+      "step": 7400
+    },
+    {
+      "epoch": 0.2768418347080403,
+      "grad_norm": 0.41424158215522766,
+      "learning_rate": 1.64503416214219e-05,
+      "loss": 0.3082,
+      "step": 7405
+    },
+    {
+      "epoch": 0.2770287636983901,
+      "grad_norm": 0.43419891595840454,
+      "learning_rate": 1.6445852959392257e-05,
+      "loss": 0.2973,
+      "step": 7410
+    },
+    {
+      "epoch": 0.27721569268873986,
+      "grad_norm": 0.4673122465610504,
+      "learning_rate": 1.644136207437262e-05,
+      "loss": 0.2796,
+      "step": 7415
+    },
+    {
+      "epoch": 0.27740262167908963,
+      "grad_norm": 0.5394338965415955,
+      "learning_rate": 1.6436868967911777e-05,
+      "loss": 0.4372,
+      "step": 7420
+    },
+    {
+      "epoch": 0.27758955066943947,
+      "grad_norm": 0.39527252316474915,
+      "learning_rate": 1.6432373641559266e-05,
+      "loss": 0.2922,
+      "step": 7425
+    },
+    {
+      "epoch": 0.27777647965978924,
+      "grad_norm": 0.35765504837036133,
+      "learning_rate": 1.6427876096865394e-05,
+      "loss": 0.3001,
+      "step": 7430
+    },
+    {
+      "epoch": 0.277963408650139,
+      "grad_norm": 0.6142287254333496,
+      "learning_rate": 1.642337633538124e-05,
+      "loss": 0.3109,
+      "step": 7435
+    },
+    {
+      "epoch": 0.2781503376404888,
+      "grad_norm": 0.45784667134284973,
+      "learning_rate": 1.641887435865864e-05,
+      "loss": 0.3899,
+      "step": 7440
+    },
+    {
+      "epoch": 0.27833726663083863,
+      "grad_norm": 0.3831481337547302,
+      "learning_rate": 1.64143701682502e-05,
+      "loss": 0.2393,
+      "step": 7445
+    },
+    {
+      "epoch": 0.2785241956211884,
+      "grad_norm": 0.628398060798645,
+      "learning_rate": 1.6409863765709282e-05,
+      "loss": 0.3072,
+      "step": 7450
+    },
+    {
+      "epoch": 0.2787111246115382,
+      "grad_norm": 0.3460518717765808,
+      "learning_rate": 1.6405355152590018e-05,
+      "loss": 0.3902,
+      "step": 7455
+    },
+    {
+      "epoch": 0.27889805360188796,
+      "grad_norm": 0.48453855514526367,
+      "learning_rate": 1.64008443304473e-05,
+      "loss": 0.339,
+      "step": 7460
+    },
+    {
+      "epoch": 0.2790849825922378,
+      "grad_norm": 0.4171224534511566,
+      "learning_rate": 1.6396331300836778e-05,
+      "loss": 0.358,
+      "step": 7465
+    },
+    {
+      "epoch": 0.2792719115825876,
+      "grad_norm": 0.3539869487285614,
+      "learning_rate": 1.6391816065314865e-05,
+      "loss": 0.3195,
+      "step": 7470
+    },
+    {
+      "epoch": 0.27945884057293735,
+      "grad_norm": 0.39307695627212524,
+      "learning_rate": 1.6387298625438743e-05,
+      "loss": 0.378,
+      "step": 7475
+    },
+    {
+      "epoch": 0.2796457695632871,
+      "grad_norm": 0.28309181332588196,
+      "learning_rate": 1.6382778982766347e-05,
+      "loss": 0.2541,
+      "step": 7480
+    },
+    {
+      "epoch": 0.27983269855363696,
+      "grad_norm": 0.4474238455295563,
+      "learning_rate": 1.6378257138856365e-05,
+      "loss": 0.2949,
+      "step": 7485
+    },
+    {
+      "epoch": 0.28001962754398674,
+      "grad_norm": 0.5505223274230957,
+      "learning_rate": 1.6373733095268258e-05,
+      "loss": 0.3328,
+      "step": 7490
+    },
+    {
+      "epoch": 0.2802065565343365,
+      "grad_norm": 0.3556166887283325,
+      "learning_rate": 1.636920685356224e-05,
+      "loss": 0.3287,
+      "step": 7495
+    },
+    {
+      "epoch": 0.2803934855246863,
+      "grad_norm": 0.3573717772960663,
+      "learning_rate": 1.6364678415299274e-05,
+      "loss": 0.363,
+      "step": 7500
+    },
+    {
+      "epoch": 0.2805804145150361,
+      "grad_norm": 0.3663536608219147,
+      "learning_rate": 1.6360147782041097e-05,
+      "loss": 0.3575,
+      "step": 7505
+    },
+    {
+      "epoch": 0.2807673435053859,
+      "grad_norm": 0.26305726170539856,
+      "learning_rate": 1.6355614955350187e-05,
+      "loss": 0.3127,
+      "step": 7510
+    },
+    {
+      "epoch": 0.2809542724957357,
+      "grad_norm": 0.5720646977424622,
+      "learning_rate": 1.6351079936789792e-05,
+      "loss": 0.377,
+      "step": 7515
+    },
+    {
+      "epoch": 0.28114120148608546,
+      "grad_norm": 0.3031104505062103,
+      "learning_rate": 1.634654272792391e-05,
+      "loss": 0.3199,
+      "step": 7520
+    },
+    {
+      "epoch": 0.2813281304764353,
+      "grad_norm": 0.4231772720813751,
+      "learning_rate": 1.6342003330317295e-05,
+      "loss": 0.3691,
+      "step": 7525
+    },
+    {
+      "epoch": 0.28151505946678507,
+      "grad_norm": 0.4373528063297272,
+      "learning_rate": 1.6337461745535446e-05,
+      "loss": 0.3065,
+      "step": 7530
+    },
+    {
+      "epoch": 0.28170198845713484,
+      "grad_norm": 0.3790377676486969,
+      "learning_rate": 1.6332917975144638e-05,
+      "loss": 0.3129,
+      "step": 7535
+    },
+    {
+      "epoch": 0.2818889174474846,
+      "grad_norm": 0.5475689172744751,
+      "learning_rate": 1.632837202071188e-05,
+      "loss": 0.3161,
+      "step": 7540
+    },
+    {
+      "epoch": 0.28207584643783445,
+      "grad_norm": 0.3444899320602417,
+      "learning_rate": 1.6323823883804942e-05,
+      "loss": 0.3612,
+      "step": 7545
+    },
+    {
+      "epoch": 0.28226277542818423,
+      "grad_norm": 0.38195663690567017,
+      "learning_rate": 1.631927356599235e-05,
+      "loss": 0.3729,
+      "step": 7550
+    },
+    {
+      "epoch": 0.282449704418534,
+      "grad_norm": 0.3962342143058777,
+      "learning_rate": 1.6314721068843367e-05,
+      "loss": 0.3536,
+      "step": 7555
+    },
+    {
+      "epoch": 0.2826366334088838,
+      "grad_norm": 0.41847673058509827,
+      "learning_rate": 1.6310166393928036e-05,
+      "loss": 0.3601,
+      "step": 7560
+    },
+    {
+      "epoch": 0.28282356239923356,
+      "grad_norm": 0.4625031650066376,
+      "learning_rate": 1.630560954281712e-05,
+      "loss": 0.3712,
+      "step": 7565
+    },
+    {
+      "epoch": 0.2830104913895834,
+      "grad_norm": 0.24907182157039642,
+      "learning_rate": 1.6301050517082154e-05,
+      "loss": 0.2722,
+      "step": 7570
+    },
+    {
+      "epoch": 0.2831974203799332,
+      "grad_norm": 0.4163358211517334,
+      "learning_rate": 1.629648931829541e-05,
+      "loss": 0.3019,
+      "step": 7575
+    },
+    {
+      "epoch": 0.28338434937028295,
+      "grad_norm": 0.24772894382476807,
+      "learning_rate": 1.6291925948029918e-05,
+      "loss": 0.2406,
+      "step": 7580
+    },
+    {
+      "epoch": 0.28357127836063273,
+      "grad_norm": 0.42197006940841675,
+      "learning_rate": 1.6287360407859452e-05,
+      "loss": 0.2812,
+      "step": 7585
+    },
+    {
+      "epoch": 0.28375820735098256,
+      "grad_norm": 0.38658830523490906,
+      "learning_rate": 1.628279269935854e-05,
+      "loss": 0.3223,
+      "step": 7590
+    },
+    {
+      "epoch": 0.28394513634133234,
+      "grad_norm": 0.6128913164138794,
+      "learning_rate": 1.627822282410245e-05,
+      "loss": 0.3035,
+      "step": 7595
+    },
+    {
+      "epoch": 0.2841320653316821,
+      "grad_norm": 0.44375917315483093,
+      "learning_rate": 1.62736507836672e-05,
+      "loss": 0.3087,
+      "step": 7600
+    },
+    {
+      "epoch": 0.2843189943220319,
+      "grad_norm": 0.35817980766296387,
+      "learning_rate": 1.626907657962956e-05,
+      "loss": 0.2697,
+      "step": 7605
+    },
+    {
+      "epoch": 0.2845059233123817,
+      "grad_norm": 0.3279922902584076,
+      "learning_rate": 1.6264500213567038e-05,
+      "loss": 0.2629,
+      "step": 7610
+    },
+    {
+      "epoch": 0.2846928523027315,
+      "grad_norm": 0.4176391363143921,
+      "learning_rate": 1.625992168705789e-05,
+      "loss": 0.2812,
+      "step": 7615
+    },
+    {
+      "epoch": 0.2848797812930813,
+      "grad_norm": 0.5126670002937317,
+      "learning_rate": 1.6255341001681125e-05,
+      "loss": 0.2976,
+      "step": 7620
+    },
+    {
+      "epoch": 0.28506671028343106,
+      "grad_norm": 0.40457814931869507,
+      "learning_rate": 1.625075815901649e-05,
+      "loss": 0.258,
+      "step": 7625
+    },
+    {
+      "epoch": 0.2852536392737809,
+      "grad_norm": 0.6561230421066284,
+      "learning_rate": 1.624617316064447e-05,
+      "loss": 0.3499,
+      "step": 7630
+    },
+    {
+      "epoch": 0.28544056826413067,
+      "grad_norm": 0.17372195422649384,
+      "learning_rate": 1.62415860081463e-05,
+      "loss": 0.2878,
+      "step": 7635
+    },
+    {
+      "epoch": 0.28562749725448044,
+      "grad_norm": 0.3161996603012085,
+      "learning_rate": 1.6236996703103963e-05,
+      "loss": 0.2953,
+      "step": 7640
+    },
+    {
+      "epoch": 0.2858144262448302,
+      "grad_norm": 0.41537266969680786,
+      "learning_rate": 1.6232405247100173e-05,
+      "loss": 0.4672,
+      "step": 7645
+    },
+    {
+      "epoch": 0.28600135523518005,
+      "grad_norm": 0.5545313358306885,
+      "learning_rate": 1.6227811641718392e-05,
+      "loss": 0.251,
+      "step": 7650
+    },
+    {
+      "epoch": 0.28618828422552983,
+      "grad_norm": 0.4810824692249298,
+      "learning_rate": 1.6223215888542832e-05,
+      "loss": 0.2838,
+      "step": 7655
+    },
+    {
+      "epoch": 0.2863752132158796,
+      "grad_norm": 0.3870065212249756,
+      "learning_rate": 1.6218617989158426e-05,
+      "loss": 0.2886,
+      "step": 7660
+    },
+    {
+      "epoch": 0.2865621422062294,
+      "grad_norm": 0.5083006024360657,
+      "learning_rate": 1.6214017945150863e-05,
+      "loss": 0.2963,
+      "step": 7665
+    },
+    {
+      "epoch": 0.2867490711965792,
+      "grad_norm": 0.4420605003833771,
+      "learning_rate": 1.6209415758106565e-05,
+      "loss": 0.3387,
+      "step": 7670
+    },
+    {
+      "epoch": 0.286936000186929,
+      "grad_norm": 0.29529669880867004,
+      "learning_rate": 1.620481142961269e-05,
+      "loss": 0.2956,
+      "step": 7675
+    },
+    {
+      "epoch": 0.2871229291772788,
+      "grad_norm": 0.25295814871788025,
+      "learning_rate": 1.6200204961257148e-05,
+      "loss": 0.2808,
+      "step": 7680
+    },
+    {
+      "epoch": 0.28730985816762855,
+      "grad_norm": 0.3369937241077423,
+      "learning_rate": 1.619559635462857e-05,
+      "loss": 0.3474,
+      "step": 7685
+    },
+    {
+      "epoch": 0.2874967871579784,
+      "grad_norm": 0.5575014352798462,
+      "learning_rate": 1.6190985611316336e-05,
+      "loss": 0.3019,
+      "step": 7690
+    },
+    {
+      "epoch": 0.28768371614832816,
+      "grad_norm": 0.5619780421257019,
+      "learning_rate": 1.618637273291056e-05,
+      "loss": 0.2889,
+      "step": 7695
+    },
+    {
+      "epoch": 0.28787064513867794,
+      "grad_norm": 0.6334633231163025,
+      "learning_rate": 1.6181757721002092e-05,
+      "loss": 0.406,
+      "step": 7700
+    },
+    {
+      "epoch": 0.2880575741290277,
+      "grad_norm": 0.27999556064605713,
+      "learning_rate": 1.617714057718251e-05,
+      "loss": 0.3111,
+      "step": 7705
+    },
+    {
+      "epoch": 0.28824450311937755,
+      "grad_norm": 0.2326452136039734,
+      "learning_rate": 1.6172521303044145e-05,
+      "loss": 0.3168,
+      "step": 7710
+    },
+    {
+      "epoch": 0.2884314321097273,
+      "grad_norm": 0.3384558856487274,
+      "learning_rate": 1.616789990018005e-05,
+      "loss": 0.3016,
+      "step": 7715
+    },
+    {
+      "epoch": 0.2886183611000771,
+      "grad_norm": 0.29935914278030396,
+      "learning_rate": 1.616327637018401e-05,
+      "loss": 0.2635,
+      "step": 7720
+    },
+    {
+      "epoch": 0.2888052900904269,
+      "grad_norm": 0.3248922526836395,
+      "learning_rate": 1.615865071465055e-05,
+      "loss": 0.3969,
+      "step": 7725
+    },
+    {
+      "epoch": 0.2889922190807767,
+      "grad_norm": 0.2685738205909729,
+      "learning_rate": 1.6154022935174923e-05,
+      "loss": 0.3624,
+      "step": 7730
+    },
+    {
+      "epoch": 0.2891791480711265,
+      "grad_norm": 0.34825339913368225,
+      "learning_rate": 1.614939303335312e-05,
+      "loss": 0.2855,
+      "step": 7735
+    },
+    {
+      "epoch": 0.28936607706147627,
+      "grad_norm": 0.41288071870803833,
+      "learning_rate": 1.6144761010781867e-05,
+      "loss": 0.2983,
+      "step": 7740
+    },
+    {
+      "epoch": 0.28955300605182605,
+      "grad_norm": 0.34375739097595215,
+      "learning_rate": 1.614012686905861e-05,
+      "loss": 0.2286,
+      "step": 7745
+    },
+    {
+      "epoch": 0.2897399350421759,
+      "grad_norm": 0.16923807561397552,
+      "learning_rate": 1.6135490609781534e-05,
+      "loss": 0.2648,
+      "step": 7750
+    },
+    {
+      "epoch": 0.28992686403252566,
+      "grad_norm": 0.391319215297699,
+      "learning_rate": 1.613085223454955e-05,
+      "loss": 0.3083,
+      "step": 7755
+    },
+    {
+      "epoch": 0.29011379302287543,
+      "grad_norm": 0.35019999742507935,
+      "learning_rate": 1.61262117449623e-05,
+      "loss": 0.2544,
+      "step": 7760
+    },
+    {
+      "epoch": 0.2903007220132252,
+      "grad_norm": 0.35723742842674255,
+      "learning_rate": 1.612156914262016e-05,
+      "loss": 0.2807,
+      "step": 7765
+    },
+    {
+      "epoch": 0.29048765100357504,
+      "grad_norm": 0.5157048106193542,
+      "learning_rate": 1.6116924429124222e-05,
+      "loss": 0.3461,
+      "step": 7770
+    },
+    {
+      "epoch": 0.2906745799939248,
+      "grad_norm": 0.504540741443634,
+      "learning_rate": 1.6112277606076325e-05,
+      "loss": 0.3168,
+      "step": 7775
+    },
+    {
+      "epoch": 0.2908615089842746,
+      "grad_norm": 0.4043572247028351,
+      "learning_rate": 1.6107628675079023e-05,
+      "loss": 0.2517,
+      "step": 7780
+    },
+    {
+      "epoch": 0.2910484379746244,
+      "grad_norm": 0.29459133744239807,
+      "learning_rate": 1.6102977637735587e-05,
+      "loss": 0.2881,
+      "step": 7785
+    },
+    {
+      "epoch": 0.2912353669649742,
+      "grad_norm": 0.20902110636234283,
+      "learning_rate": 1.6098324495650044e-05,
+      "loss": 0.3696,
+      "step": 7790
+    },
+    {
+      "epoch": 0.291422295955324,
+      "grad_norm": 0.3627723157405853,
+      "learning_rate": 1.609366925042712e-05,
+      "loss": 0.1937,
+      "step": 7795
+    },
+    {
+      "epoch": 0.29160922494567376,
+      "grad_norm": 0.4867866337299347,
+      "learning_rate": 1.6089011903672277e-05,
+      "loss": 0.362,
+      "step": 7800
+    },
+    {
+      "epoch": 0.29179615393602354,
+      "grad_norm": 0.2678326964378357,
+      "learning_rate": 1.6084352456991704e-05,
+      "loss": 0.36,
+      "step": 7805
+    },
+    {
+      "epoch": 0.2919830829263733,
+      "grad_norm": 0.7733646631240845,
+      "learning_rate": 1.6079690911992304e-05,
+      "loss": 0.2725,
+      "step": 7810
+    },
+    {
+      "epoch": 0.29217001191672315,
+      "grad_norm": 1.191166639328003,
+      "learning_rate": 1.6075027270281713e-05,
+      "loss": 0.2314,
+      "step": 7815
+    },
+    {
+      "epoch": 0.2923569409070729,
+      "grad_norm": 0.3552202880382538,
+      "learning_rate": 1.607036153346829e-05,
+      "loss": 0.3046,
+      "step": 7820
+    },
+    {
+      "epoch": 0.2925438698974227,
+      "grad_norm": 0.42656761407852173,
+      "learning_rate": 1.6065693703161113e-05,
+      "loss": 0.315,
+      "step": 7825
+    },
+    {
+      "epoch": 0.2927307988877725,
+      "grad_norm": 0.3393990099430084,
+      "learning_rate": 1.606102378096998e-05,
+      "loss": 0.2877,
+      "step": 7830
+    },
+    {
+      "epoch": 0.2929177278781223,
+      "grad_norm": 0.3915458619594574,
+      "learning_rate": 1.605635176850541e-05,
+      "loss": 0.3389,
+      "step": 7835
+    },
+    {
+      "epoch": 0.2931046568684721,
+      "grad_norm": 0.4257470369338989,
+      "learning_rate": 1.605167766737866e-05,
+      "loss": 0.317,
+      "step": 7840
+    },
+    {
+      "epoch": 0.29329158585882187,
+      "grad_norm": 0.36982858180999756,
+      "learning_rate": 1.604700147920168e-05,
+      "loss": 0.3218,
+      "step": 7845
+    },
+    {
+      "epoch": 0.29347851484917165,
+      "grad_norm": 0.4049963057041168,
+      "learning_rate": 1.604232320558716e-05,
+      "loss": 0.2894,
+      "step": 7850
+    },
+    {
+      "epoch": 0.2936654438395215,
+      "grad_norm": 0.7059416174888611,
+      "learning_rate": 1.6037642848148502e-05,
+      "loss": 0.2444,
+      "step": 7855
+    },
+    {
+      "epoch": 0.29385237282987126,
+      "grad_norm": 0.635959267616272,
+      "learning_rate": 1.6032960408499824e-05,
+      "loss": 0.2997,
+      "step": 7860
+    },
+    {
+      "epoch": 0.29403930182022103,
+      "grad_norm": 0.24313776195049286,
+      "learning_rate": 1.602827588825597e-05,
+      "loss": 0.2254,
+      "step": 7865
+    },
+    {
+      "epoch": 0.2942262308105708,
+      "grad_norm": 0.3153620660305023,
+      "learning_rate": 1.6023589289032494e-05,
+      "loss": 0.2558,
+      "step": 7870
+    },
+    {
+      "epoch": 0.29441315980092064,
+      "grad_norm": 0.45469948649406433,
+      "learning_rate": 1.6018900612445665e-05,
+      "loss": 0.2513,
+      "step": 7875
+    },
+    {
+      "epoch": 0.2946000887912704,
+      "grad_norm": 0.47144100069999695,
+      "learning_rate": 1.601420986011248e-05,
+      "loss": 0.2372,
+      "step": 7880
+    },
+    {
+      "epoch": 0.2947870177816202,
+      "grad_norm": 0.22326113283634186,
+      "learning_rate": 1.6009517033650643e-05,
+      "loss": 0.2896,
+      "step": 7885
+    },
+    {
+      "epoch": 0.29497394677197,
+      "grad_norm": 0.5031372308731079,
+      "learning_rate": 1.6004822134678577e-05,
+      "loss": 0.2962,
+      "step": 7890
+    },
+    {
+      "epoch": 0.2951608757623198,
+      "grad_norm": 0.6720777750015259,
+      "learning_rate": 1.6000125164815418e-05,
+      "loss": 0.3161,
+      "step": 7895
+    },
+    {
+      "epoch": 0.2953478047526696,
+      "grad_norm": 0.4197002351284027,
+      "learning_rate": 1.5995426125681014e-05,
+      "loss": 0.3376,
+      "step": 7900
+    },
+    {
+      "epoch": 0.29553473374301936,
+      "grad_norm": 0.13400337100028992,
+      "learning_rate": 1.599072501889593e-05,
+      "loss": 0.3168,
+      "step": 7905
+    },
+    {
+      "epoch": 0.29572166273336914,
+      "grad_norm": 0.3738100230693817,
+      "learning_rate": 1.598602184608144e-05,
+      "loss": 0.2846,
+      "step": 7910
+    },
+    {
+      "epoch": 0.29590859172371897,
+      "grad_norm": 0.5275940895080566,
+      "learning_rate": 1.598131660885954e-05,
+      "loss": 0.2793,
+      "step": 7915
+    },
+    {
+      "epoch": 0.29609552071406875,
+      "grad_norm": 0.41306841373443604,
+      "learning_rate": 1.597660930885293e-05,
+      "loss": 0.3141,
+      "step": 7920
+    },
+    {
+      "epoch": 0.2962824497044185,
+      "grad_norm": 0.2588225305080414,
+      "learning_rate": 1.5971899947685018e-05,
+      "loss": 0.2731,
+      "step": 7925
+    },
+    {
+      "epoch": 0.2964693786947683,
+      "grad_norm": 0.1754133701324463,
+      "learning_rate": 1.5967188526979928e-05,
+      "loss": 0.2432,
+      "step": 7930
+    },
+    {
+      "epoch": 0.29665630768511814,
+      "grad_norm": 0.18629756569862366,
+      "learning_rate": 1.5962475048362498e-05,
+      "loss": 0.3458,
+      "step": 7935
+    },
+    {
+      "epoch": 0.2968432366754679,
+      "grad_norm": 0.6092668175697327,
+      "learning_rate": 1.5957759513458274e-05,
+      "loss": 0.3054,
+      "step": 7940
+    },
+    {
+      "epoch": 0.2970301656658177,
+      "grad_norm": 0.3897908926010132,
+      "learning_rate": 1.59530419238935e-05,
+      "loss": 0.2734,
+      "step": 7945
+    },
+    {
+      "epoch": 0.29721709465616747,
+      "grad_norm": 0.4030907154083252,
+      "learning_rate": 1.5948322281295147e-05,
+      "loss": 0.3481,
+      "step": 7950
+    },
+    {
+      "epoch": 0.2974040236465173,
+      "grad_norm": 0.3852006793022156,
+      "learning_rate": 1.594360058729088e-05,
+      "loss": 0.2923,
+      "step": 7955
+    },
+    {
+      "epoch": 0.2975909526368671,
+      "grad_norm": 0.3748472034931183,
+      "learning_rate": 1.5938876843509072e-05,
+      "loss": 0.3313,
+      "step": 7960
+    },
+    {
+      "epoch": 0.29777788162721686,
+      "grad_norm": 0.5024619698524475,
+      "learning_rate": 1.5934151051578814e-05,
+      "loss": 0.2947,
+      "step": 7965
+    },
+    {
+      "epoch": 0.29796481061756663,
+      "grad_norm": 0.4308888018131256,
+      "learning_rate": 1.592942321312989e-05,
+      "loss": 0.2932,
+      "step": 7970
+    },
+    {
+      "epoch": 0.29815173960791647,
+      "grad_norm": 0.4150945842266083,
+      "learning_rate": 1.5924693329792808e-05,
+      "loss": 0.351,
+      "step": 7975
+    },
+    {
+      "epoch": 0.29833866859826624,
+      "grad_norm": 0.2940865755081177,
+      "learning_rate": 1.5919961403198752e-05,
+      "loss": 0.2743,
+      "step": 7980
+    },
+    {
+      "epoch": 0.298525597588616,
+      "grad_norm": 0.2742552161216736,
+      "learning_rate": 1.591522743497964e-05,
+      "loss": 0.266,
+      "step": 7985
+    },
+    {
+      "epoch": 0.2987125265789658,
+      "grad_norm": 0.5203600525856018,
+      "learning_rate": 1.591049142676808e-05,
+      "loss": 0.3469,
+      "step": 7990
+    },
+    {
+      "epoch": 0.29889945556931563,
+      "grad_norm": 0.29020723700523376,
+      "learning_rate": 1.590575338019738e-05,
+      "loss": 0.2792,
+      "step": 7995
+    },
+    {
+      "epoch": 0.2990863845596654,
+      "grad_norm": 0.5410657525062561,
+      "learning_rate": 1.590101329690156e-05,
+      "loss": 0.2447,
+      "step": 8000
+    },
+    {
+      "epoch": 0.2992733135500152,
+      "grad_norm": 0.36561673879623413,
+      "learning_rate": 1.589627117851534e-05,
+      "loss": 0.2502,
+      "step": 8005
+    },
+    {
+      "epoch": 0.29946024254036496,
+      "grad_norm": 0.2769913673400879,
+      "learning_rate": 1.589152702667414e-05,
+      "loss": 0.3428,
+      "step": 8010
+    },
+    {
+      "epoch": 0.2996471715307148,
+      "grad_norm": 0.34356051683425903,
+      "learning_rate": 1.5886780843014085e-05,
+      "loss": 0.3429,
+      "step": 8015
+    },
+    {
+      "epoch": 0.2998341005210646,
+      "grad_norm": 0.2786838710308075,
+      "learning_rate": 1.5882032629171993e-05,
+      "loss": 0.3215,
+      "step": 8020
+    },
+    {
+      "epoch": 0.30002102951141435,
+      "grad_norm": 0.46318674087524414,
+      "learning_rate": 1.587728238678539e-05,
+      "loss": 0.3538,
+      "step": 8025
+    },
+    {
+      "epoch": 0.3002079585017641,
+      "grad_norm": 0.3690614402294159,
+      "learning_rate": 1.5872530117492495e-05,
+      "loss": 0.3055,
+      "step": 8030
+    },
+    {
+      "epoch": 0.30039488749211396,
+      "grad_norm": 0.8592503070831299,
+      "learning_rate": 1.5867775822932233e-05,
+      "loss": 0.3292,
+      "step": 8035
+    },
+    {
+      "epoch": 0.30058181648246374,
+      "grad_norm": 0.4920271635055542,
+      "learning_rate": 1.5863019504744222e-05,
+      "loss": 0.2921,
+      "step": 8040
+    },
+    {
+      "epoch": 0.3007687454728135,
+      "grad_norm": 0.3858177363872528,
+      "learning_rate": 1.585826116456878e-05,
+      "loss": 0.3832,
+      "step": 8045
+    },
+    {
+      "epoch": 0.3009556744631633,
+      "grad_norm": 0.37820449471473694,
+      "learning_rate": 1.5853500804046926e-05,
+      "loss": 0.3231,
+      "step": 8050
+    },
+    {
+      "epoch": 0.30114260345351307,
+      "grad_norm": 0.24952290952205658,
+      "learning_rate": 1.5848738424820366e-05,
+      "loss": 0.3362,
+      "step": 8055
+    },
+    {
+      "epoch": 0.3013295324438629,
+      "grad_norm": 0.39352136850357056,
+      "learning_rate": 1.584397402853151e-05,
+      "loss": 0.2533,
+      "step": 8060
+    },
+    {
+      "epoch": 0.3015164614342127,
+      "grad_norm": 0.34481966495513916,
+      "learning_rate": 1.5839207616823468e-05,
+      "loss": 0.3445,
+      "step": 8065
+    },
+    {
+      "epoch": 0.30170339042456246,
+      "grad_norm": 0.6301359534263611,
+      "learning_rate": 1.583443919134003e-05,
+      "loss": 0.332,
+      "step": 8070
+    },
+    {
+      "epoch": 0.30189031941491223,
+      "grad_norm": 0.6552008986473083,
+      "learning_rate": 1.582966875372569e-05,
+      "loss": 0.3042,
+      "step": 8075
+    },
+    {
+      "epoch": 0.30207724840526207,
+      "grad_norm": 0.3802003860473633,
+      "learning_rate": 1.582489630562564e-05,
+      "loss": 0.2635,
+      "step": 8080
+    },
+    {
+      "epoch": 0.30226417739561184,
+      "grad_norm": 0.4597533941268921,
+      "learning_rate": 1.5820121848685758e-05,
+      "loss": 0.2614,
+      "step": 8085
+    },
+    {
+      "epoch": 0.3024511063859616,
+      "grad_norm": 0.4375554025173187,
+      "learning_rate": 1.581534538455262e-05,
+      "loss": 0.342,
+      "step": 8090
+    },
+    {
+      "epoch": 0.3026380353763114,
+      "grad_norm": 0.36787232756614685,
+      "learning_rate": 1.5810566914873487e-05,
+      "loss": 0.3305,
+      "step": 8095
+    },
+    {
+      "epoch": 0.30282496436666123,
+      "grad_norm": 0.342817485332489,
+      "learning_rate": 1.580578644129632e-05,
+      "loss": 0.3672,
+      "step": 8100
+    },
+    {
+      "epoch": 0.303011893357011,
+      "grad_norm": 0.8170397281646729,
+      "learning_rate": 1.5801003965469764e-05,
+      "loss": 0.2564,
+      "step": 8105
+    },
+    {
+      "epoch": 0.3031988223473608,
+      "grad_norm": 0.1437387764453888,
+      "learning_rate": 1.5796219489043164e-05,
+      "loss": 0.3617,
+      "step": 8110
+    },
+    {
+      "epoch": 0.30338575133771056,
+      "grad_norm": 0.4565275013446808,
+      "learning_rate": 1.5791433013666544e-05,
+      "loss": 0.2808,
+      "step": 8115
+    },
+    {
+      "epoch": 0.3035726803280604,
+      "grad_norm": 0.5784963369369507,
+      "learning_rate": 1.5786644540990622e-05,
+      "loss": 0.234,
+      "step": 8120
+    },
+    {
+      "epoch": 0.3037596093184102,
+      "grad_norm": 0.44726547598838806,
+      "learning_rate": 1.578185407266681e-05,
+      "loss": 0.3222,
+      "step": 8125
+    },
+    {
+      "epoch": 0.30394653830875995,
+      "grad_norm": 0.260372132062912,
+      "learning_rate": 1.5777061610347197e-05,
+      "loss": 0.2867,
+      "step": 8130
+    },
+    {
+      "epoch": 0.3041334672991097,
+      "grad_norm": 0.5115578174591064,
+      "learning_rate": 1.5772267155684565e-05,
+      "loss": 0.3603,
+      "step": 8135
+    },
+    {
+      "epoch": 0.30432039628945956,
+      "grad_norm": 0.4961742162704468,
+      "learning_rate": 1.576747071033239e-05,
+      "loss": 0.2761,
+      "step": 8140
+    },
+    {
+      "epoch": 0.30450732527980934,
+      "grad_norm": 0.07623039931058884,
+      "learning_rate": 1.5762672275944826e-05,
+      "loss": 0.3036,
+      "step": 8145
+    },
+    {
+      "epoch": 0.3046942542701591,
+      "grad_norm": 0.7987903356552124,
+      "learning_rate": 1.5757871854176716e-05,
+      "loss": 0.2701,
+      "step": 8150
+    },
+    {
+      "epoch": 0.3048811832605089,
+      "grad_norm": 0.34833961725234985,
+      "learning_rate": 1.5753069446683586e-05,
+      "loss": 0.2672,
+      "step": 8155
+    },
+    {
+      "epoch": 0.3050681122508587,
+      "grad_norm": 0.5639849305152893,
+      "learning_rate": 1.574826505512165e-05,
+      "loss": 0.3335,
+      "step": 8160
+    },
+    {
+      "epoch": 0.3052550412412085,
+      "grad_norm": 0.2684768736362457,
+      "learning_rate": 1.5743458681147807e-05,
+      "loss": 0.2879,
+      "step": 8165
+    },
+    {
+      "epoch": 0.3054419702315583,
+      "grad_norm": 0.32152795791625977,
+      "learning_rate": 1.5738650326419636e-05,
+      "loss": 0.3091,
+      "step": 8170
+    },
+    {
+      "epoch": 0.30562889922190806,
+      "grad_norm": 0.4024050533771515,
+      "learning_rate": 1.57338399925954e-05,
+      "loss": 0.3315,
+      "step": 8175
+    },
+    {
+      "epoch": 0.3058158282122579,
+      "grad_norm": 0.32922789454460144,
+      "learning_rate": 1.5729027681334043e-05,
+      "loss": 0.3058,
+      "step": 8180
+    },
+    {
+      "epoch": 0.30600275720260767,
+      "grad_norm": 0.37226831912994385,
+      "learning_rate": 1.57242133942952e-05,
+      "loss": 0.2949,
+      "step": 8185
+    },
+    {
+      "epoch": 0.30618968619295744,
+      "grad_norm": 0.5778080821037292,
+      "learning_rate": 1.5719397133139172e-05,
+      "loss": 0.3008,
+      "step": 8190
+    },
+    {
+      "epoch": 0.3063766151833072,
+      "grad_norm": 0.30780428647994995,
+      "learning_rate": 1.5714578899526957e-05,
+      "loss": 0.2412,
+      "step": 8195
+    },
+    {
+      "epoch": 0.30656354417365705,
+      "grad_norm": 0.13755351305007935,
+      "learning_rate": 1.5709758695120222e-05,
+      "loss": 0.2439,
+      "step": 8200
+    },
+    {
+      "epoch": 0.30675047316400683,
+      "grad_norm": 0.5074127316474915,
+      "learning_rate": 1.570493652158132e-05,
+      "loss": 0.3904,
+      "step": 8205
+    },
+    {
+      "epoch": 0.3069374021543566,
+      "grad_norm": 0.27860695123672485,
+      "learning_rate": 1.5700112380573277e-05,
+      "loss": 0.313,
+      "step": 8210
+    },
+    {
+      "epoch": 0.3071243311447064,
+      "grad_norm": 0.4412946403026581,
+      "learning_rate": 1.5695286273759805e-05,
+      "loss": 0.3682,
+      "step": 8215
+    },
+    {
+      "epoch": 0.3073112601350562,
+      "grad_norm": 0.402165025472641,
+      "learning_rate": 1.569045820280529e-05,
+      "loss": 0.2835,
+      "step": 8220
+    },
+    {
+      "epoch": 0.307498189125406,
+      "grad_norm": 0.49247145652770996,
+      "learning_rate": 1.5685628169374793e-05,
+      "loss": 0.2833,
+      "step": 8225
+    },
+    {
+      "epoch": 0.3076851181157558,
+      "grad_norm": 0.41755223274230957,
+      "learning_rate": 1.568079617513405e-05,
+      "loss": 0.2886,
+      "step": 8230
+    },
+    {
+      "epoch": 0.30787204710610555,
+      "grad_norm": 0.34908807277679443,
+      "learning_rate": 1.5675962221749478e-05,
+      "loss": 0.275,
+      "step": 8235
+    },
+    {
+      "epoch": 0.3080589760964554,
+      "grad_norm": 0.3702501356601715,
+      "learning_rate": 1.567112631088818e-05,
+      "loss": 0.3203,
+      "step": 8240
+    },
+    {
+      "epoch": 0.30824590508680516,
+      "grad_norm": 0.42112669348716736,
+      "learning_rate": 1.5666288444217915e-05,
+      "loss": 0.2694,
+      "step": 8245
+    },
+    {
+      "epoch": 0.30843283407715494,
+      "grad_norm": 0.29839271306991577,
+      "learning_rate": 1.5661448623407122e-05,
+      "loss": 0.288,
+      "step": 8250
+    },
+    {
+      "epoch": 0.3086197630675047,
+      "grad_norm": 0.4897206425666809,
+      "learning_rate": 1.565660685012492e-05,
+      "loss": 0.3968,
+      "step": 8255
+    },
+    {
+      "epoch": 0.30880669205785455,
+      "grad_norm": 0.6160528659820557,
+      "learning_rate": 1.5651763126041098e-05,
+      "loss": 0.2819,
+      "step": 8260
+    },
+    {
+      "epoch": 0.3089936210482043,
+      "grad_norm": 0.2819858193397522,
+      "learning_rate": 1.564691745282612e-05,
+      "loss": 0.2829,
+      "step": 8265
+    },
+    {
+      "epoch": 0.3091805500385541,
+      "grad_norm": 0.33518272638320923,
+      "learning_rate": 1.5642069832151116e-05,
+      "loss": 0.2391,
+      "step": 8270
+    },
+    {
+      "epoch": 0.3093674790289039,
+      "grad_norm": 0.3020849823951721,
+      "learning_rate": 1.5637220265687894e-05,
+      "loss": 0.2829,
+      "step": 8275
+    },
+    {
+      "epoch": 0.3095544080192537,
+      "grad_norm": 0.42169663310050964,
+      "learning_rate": 1.5632368755108926e-05,
+      "loss": 0.312,
+      "step": 8280
+    },
+    {
+      "epoch": 0.3097413370096035,
+      "grad_norm": 0.34373271465301514,
+      "learning_rate": 1.5627515302087362e-05,
+      "loss": 0.2879,
+      "step": 8285
+    },
+    {
+      "epoch": 0.30992826599995327,
+      "grad_norm": 0.4371539056301117,
+      "learning_rate": 1.562265990829702e-05,
+      "loss": 0.3221,
+      "step": 8290
+    },
+    {
+      "epoch": 0.31011519499030304,
+      "grad_norm": 0.3401534855365753,
+      "learning_rate": 1.5617802575412385e-05,
+      "loss": 0.2821,
+      "step": 8295
+    },
+    {
+      "epoch": 0.3103021239806528,
+      "grad_norm": 0.605153501033783,
+      "learning_rate": 1.5612943305108615e-05,
+      "loss": 0.3055,
+      "step": 8300
+    },
+    {
+      "epoch": 0.31048905297100265,
+      "grad_norm": 0.4191633462905884,
+      "learning_rate": 1.5608082099061528e-05,
+      "loss": 0.2807,
+      "step": 8305
+    },
+    {
+      "epoch": 0.31067598196135243,
+      "grad_norm": 0.2965903580188751,
+      "learning_rate": 1.560321895894762e-05,
+      "loss": 0.2722,
+      "step": 8310
+    },
+    {
+      "epoch": 0.3108629109517022,
+      "grad_norm": 0.39185696840286255,
+      "learning_rate": 1.559835388644404e-05,
+      "loss": 0.3032,
+      "step": 8315
+    },
+    {
+      "epoch": 0.311049839942052,
+      "grad_norm": 0.2846728265285492,
+      "learning_rate": 1.5593486883228617e-05,
+      "loss": 0.2977,
+      "step": 8320
+    },
+    {
+      "epoch": 0.3112367689324018,
+      "grad_norm": 0.35575684905052185,
+      "learning_rate": 1.5588617950979846e-05,
+      "loss": 0.3072,
+      "step": 8325
+    },
+    {
+      "epoch": 0.3114236979227516,
+      "grad_norm": 0.12662628293037415,
+      "learning_rate": 1.5583747091376877e-05,
+      "loss": 0.2764,
+      "step": 8330
+    },
+    {
+      "epoch": 0.3116106269131014,
+      "grad_norm": 0.439426451921463,
+      "learning_rate": 1.5578874306099533e-05,
+      "loss": 0.2643,
+      "step": 8335
+    },
+    {
+      "epoch": 0.31179755590345115,
+      "grad_norm": 0.13574817776679993,
+      "learning_rate": 1.5573999596828292e-05,
+      "loss": 0.2879,
+      "step": 8340
+    },
+    {
+      "epoch": 0.311984484893801,
+      "grad_norm": 0.6407355666160583,
+      "learning_rate": 1.5569122965244306e-05,
+      "loss": 0.2412,
+      "step": 8345
+    },
+    {
+      "epoch": 0.31217141388415076,
+      "grad_norm": 0.5395145416259766,
+      "learning_rate": 1.5564244413029385e-05,
+      "loss": 0.3068,
+      "step": 8350
+    },
+    {
+      "epoch": 0.31235834287450054,
+      "grad_norm": 0.17821340262889862,
+      "learning_rate": 1.5559363941866005e-05,
+      "loss": 0.4122,
+      "step": 8355
+    },
+    {
+      "epoch": 0.3125452718648503,
+      "grad_norm": 0.3062966465950012,
+      "learning_rate": 1.5554481553437294e-05,
+      "loss": 0.2104,
+      "step": 8360
+    },
+    {
+      "epoch": 0.31273220085520015,
+      "grad_norm": 0.5049282908439636,
+      "learning_rate": 1.5549597249427052e-05,
+      "loss": 0.253,
+      "step": 8365
+    },
+    {
+      "epoch": 0.3129191298455499,
+      "grad_norm": 0.49796777963638306,
+      "learning_rate": 1.5544711031519736e-05,
+      "loss": 0.2669,
+      "step": 8370
+    },
+    {
+      "epoch": 0.3131060588358997,
+      "grad_norm": 0.34029170870780945,
+      "learning_rate": 1.553982290140046e-05,
+      "loss": 0.3638,
+      "step": 8375
+    },
+    {
+      "epoch": 0.3132929878262495,
+      "grad_norm": 0.5282084941864014,
+      "learning_rate": 1.5534932860755e-05,
+      "loss": 0.2841,
+      "step": 8380
+    },
+    {
+      "epoch": 0.3134799168165993,
+      "grad_norm": 0.4704183340072632,
+      "learning_rate": 1.553004091126979e-05,
+      "loss": 0.3326,
+      "step": 8385
+    },
+    {
+      "epoch": 0.3136668458069491,
+      "grad_norm": 0.442396879196167,
+      "learning_rate": 1.552514705463193e-05,
+      "loss": 0.2584,
+      "step": 8390
+    },
+    {
+      "epoch": 0.31385377479729887,
+      "grad_norm": 0.40408533811569214,
+      "learning_rate": 1.5520251292529165e-05,
+      "loss": 0.3079,
+      "step": 8395
+    },
+    {
+      "epoch": 0.31404070378764865,
+      "grad_norm": 0.34462475776672363,
+      "learning_rate": 1.5515353626649905e-05,
+      "loss": 0.2727,
+      "step": 8400
+    },
+    {
+      "epoch": 0.3142276327779985,
+      "grad_norm": 0.043053608387708664,
+      "learning_rate": 1.5510454058683216e-05,
+      "loss": 0.2879,
+      "step": 8405
+    },
+    {
+      "epoch": 0.31441456176834826,
+      "grad_norm": 0.3124004006385803,
+      "learning_rate": 1.5505552590318814e-05,
+      "loss": 0.3576,
+      "step": 8410
+    },
+    {
+      "epoch": 0.31460149075869803,
+      "grad_norm": 0.4973233938217163,
+      "learning_rate": 1.5500649223247076e-05,
+      "loss": 0.2801,
+      "step": 8415
+    },
+    {
+      "epoch": 0.3147884197490478,
+      "grad_norm": 0.4319716989994049,
+      "learning_rate": 1.5495743959159034e-05,
+      "loss": 0.2454,
+      "step": 8420
+    },
+    {
+      "epoch": 0.31497534873939764,
+      "grad_norm": 0.4788093864917755,
+      "learning_rate": 1.549083679974638e-05,
+      "loss": 0.3537,
+      "step": 8425
+    },
+    {
+      "epoch": 0.3151622777297474,
+      "grad_norm": 0.6235319972038269,
+      "learning_rate": 1.548592774670144e-05,
+      "loss": 0.2974,
+      "step": 8430
+    },
+    {
+      "epoch": 0.3153492067200972,
+      "grad_norm": 0.3328862190246582,
+      "learning_rate": 1.5481016801717213e-05,
+      "loss": 0.37,
+      "step": 8435
+    },
+    {
+      "epoch": 0.315536135710447,
+      "grad_norm": 0.6881264448165894,
+      "learning_rate": 1.5476103966487345e-05,
+      "loss": 0.3738,
+      "step": 8440
+    },
+    {
+      "epoch": 0.3157230647007968,
+      "grad_norm": 0.4371614456176758,
+      "learning_rate": 1.547118924270613e-05,
+      "loss": 0.2477,
+      "step": 8445
+    },
+    {
+      "epoch": 0.3159099936911466,
+      "grad_norm": 0.6535786986351013,
+      "learning_rate": 1.546627263206851e-05,
+      "loss": 0.3318,
+      "step": 8450
+    },
+    {
+      "epoch": 0.31609692268149636,
+      "grad_norm": 0.5133426189422607,
+      "learning_rate": 1.546135413627009e-05,
+      "loss": 0.2697,
+      "step": 8455
+    },
+    {
+      "epoch": 0.31628385167184614,
+      "grad_norm": 0.5738974809646606,
+      "learning_rate": 1.5456433757007115e-05,
+      "loss": 0.247,
+      "step": 8460
+    },
+    {
+      "epoch": 0.31647078066219597,
+      "grad_norm": 0.12999732792377472,
+      "learning_rate": 1.5451511495976483e-05,
+      "loss": 0.2726,
+      "step": 8465
+    },
+    {
+      "epoch": 0.31665770965254575,
+      "grad_norm": 0.30150121450424194,
+      "learning_rate": 1.5446587354875742e-05,
+      "loss": 0.2792,
+      "step": 8470
+    },
+    {
+      "epoch": 0.3168446386428955,
+      "grad_norm": 0.36538806557655334,
+      "learning_rate": 1.544166133540309e-05,
+      "loss": 0.3656,
+      "step": 8475
+    },
+    {
+      "epoch": 0.3170315676332453,
+      "grad_norm": 0.5512821078300476,
+      "learning_rate": 1.5436733439257362e-05,
+      "loss": 0.2357,
+      "step": 8480
+    },
+    {
+      "epoch": 0.31721849662359514,
+      "grad_norm": 0.4331134259700775,
+      "learning_rate": 1.5431803668138056e-05,
+      "loss": 0.2833,
+      "step": 8485
+    },
+    {
+      "epoch": 0.3174054256139449,
+      "grad_norm": 0.2866262197494507,
+      "learning_rate": 1.5426872023745305e-05,
+      "loss": 0.1892,
+      "step": 8490
+    },
+    {
+      "epoch": 0.3175923546042947,
+      "grad_norm": 0.36052101850509644,
+      "learning_rate": 1.5421938507779893e-05,
+      "loss": 0.2636,
+      "step": 8495
+    },
+    {
+      "epoch": 0.31777928359464447,
+      "grad_norm": 0.4259335398674011,
+      "learning_rate": 1.5417003121943247e-05,
+      "loss": 0.2615,
+      "step": 8500
+    },
+    {
+      "epoch": 0.3179662125849943,
+      "grad_norm": 0.270948201417923,
+      "learning_rate": 1.541206586793744e-05,
+      "loss": 0.3241,
+      "step": 8505
+    },
+    {
+      "epoch": 0.3181531415753441,
+      "grad_norm": 0.2681352496147156,
+      "learning_rate": 1.5407126747465195e-05,
+      "loss": 0.2701,
+      "step": 8510
+    },
+    {
+      "epoch": 0.31834007056569386,
+      "grad_norm": 0.4917761981487274,
+      "learning_rate": 1.5402185762229864e-05,
+      "loss": 0.2454,
+      "step": 8515
+    },
+    {
+      "epoch": 0.31852699955604363,
+      "grad_norm": 0.4351048767566681,
+      "learning_rate": 1.539724291393546e-05,
+      "loss": 0.2836,
+      "step": 8520
+    },
+    {
+      "epoch": 0.31871392854639347,
+      "grad_norm": 0.7462584376335144,
+      "learning_rate": 1.5392298204286623e-05,
+      "loss": 0.2905,
+      "step": 8525
+    },
+    {
+      "epoch": 0.31890085753674324,
+      "grad_norm": 0.36334559321403503,
+      "learning_rate": 1.5387351634988644e-05,
+      "loss": 0.2733,
+      "step": 8530
+    },
+    {
+      "epoch": 0.319087786527093,
+      "grad_norm": 0.22820745408535004,
+      "learning_rate": 1.5382403207747453e-05,
+      "loss": 0.253,
+      "step": 8535
+    },
+    {
+      "epoch": 0.3192747155174428,
+      "grad_norm": 0.45713189244270325,
+      "learning_rate": 1.537745292426962e-05,
+      "loss": 0.2474,
+      "step": 8540
+    },
+    {
+      "epoch": 0.3194616445077926,
+      "grad_norm": 0.1870334893465042,
+      "learning_rate": 1.5372500786262357e-05,
+      "loss": 0.3835,
+      "step": 8545
+    },
+    {
+      "epoch": 0.3196485734981424,
+      "grad_norm": 0.6428795456886292,
+      "learning_rate": 1.5367546795433517e-05,
+      "loss": 0.3194,
+      "step": 8550
+    },
+    {
+      "epoch": 0.3198355024884922,
+      "grad_norm": 0.47562816739082336,
+      "learning_rate": 1.5362590953491586e-05,
+      "loss": 0.2485,
+      "step": 8555
+    },
+    {
+      "epoch": 0.32002243147884196,
+      "grad_norm": 0.418270468711853,
+      "learning_rate": 1.535763326214569e-05,
+      "loss": 0.318,
+      "step": 8560
+    },
+    {
+      "epoch": 0.32020936046919174,
+      "grad_norm": 0.7207987904548645,
+      "learning_rate": 1.53526737231056e-05,
+      "loss": 0.347,
+      "step": 8565
+    },
+    {
+      "epoch": 0.32039628945954157,
+      "grad_norm": 0.4449211061000824,
+      "learning_rate": 1.5347712338081717e-05,
+      "loss": 0.2959,
+      "step": 8570
+    },
+    {
+      "epoch": 0.32058321844989135,
+      "grad_norm": 0.29263216257095337,
+      "learning_rate": 1.5342749108785084e-05,
+      "loss": 0.2846,
+      "step": 8575
+    },
+    {
+      "epoch": 0.3207701474402411,
+      "grad_norm": 0.5871028900146484,
+      "learning_rate": 1.5337784036927367e-05,
+      "loss": 0.3511,
+      "step": 8580
+    },
+    {
+      "epoch": 0.3209570764305909,
+      "grad_norm": 0.3245595395565033,
+      "learning_rate": 1.533281712422088e-05,
+      "loss": 0.3302,
+      "step": 8585
+    },
+    {
+      "epoch": 0.32114400542094074,
+      "grad_norm": 0.6525892019271851,
+      "learning_rate": 1.5327848372378574e-05,
+      "loss": 0.3053,
+      "step": 8590
+    },
+    {
+      "epoch": 0.3213309344112905,
+      "grad_norm": 0.3770609498023987,
+      "learning_rate": 1.5322877783114027e-05,
+      "loss": 0.3547,
+      "step": 8595
+    },
+    {
+      "epoch": 0.3215178634016403,
+      "grad_norm": 0.4422697126865387,
+      "learning_rate": 1.5317905358141456e-05,
+      "loss": 0.2285,
+      "step": 8600
+    },
+    {
+      "epoch": 0.32170479239199007,
+      "grad_norm": 0.3939485549926758,
+      "learning_rate": 1.53129310991757e-05,
+      "loss": 0.3421,
+      "step": 8605
+    },
+    {
+      "epoch": 0.3218917213823399,
+      "grad_norm": 0.3674333095550537,
+      "learning_rate": 1.5307955007932243e-05,
+      "loss": 0.2389,
+      "step": 8610
+    },
+    {
+      "epoch": 0.3220786503726897,
+      "grad_norm": 0.27263256907463074,
+      "learning_rate": 1.5302977086127194e-05,
+      "loss": 0.2633,
+      "step": 8615
+    },
+    {
+      "epoch": 0.32226557936303946,
+      "grad_norm": 0.3871941864490509,
+      "learning_rate": 1.5297997335477302e-05,
+      "loss": 0.314,
+      "step": 8620
+    },
+    {
+      "epoch": 0.32245250835338923,
+      "grad_norm": 0.4470764398574829,
+      "learning_rate": 1.5293015757699935e-05,
+      "loss": 0.2881,
+      "step": 8625
+    },
+    {
+      "epoch": 0.32263943734373907,
+      "grad_norm": 0.6860572695732117,
+      "learning_rate": 1.5288032354513095e-05,
+      "loss": 0.3377,
+      "step": 8630
+    },
+    {
+      "epoch": 0.32282636633408884,
+      "grad_norm": 0.4774250090122223,
+      "learning_rate": 1.5283047127635418e-05,
+      "loss": 0.2989,
+      "step": 8635
+    },
+    {
+      "epoch": 0.3230132953244386,
+      "grad_norm": 0.2891339957714081,
+      "learning_rate": 1.5278060078786166e-05,
+      "loss": 0.2577,
+      "step": 8640
+    },
+    {
+      "epoch": 0.3232002243147884,
+      "grad_norm": 0.7874906659126282,
+      "learning_rate": 1.5273071209685227e-05,
+      "loss": 0.3161,
+      "step": 8645
+    },
+    {
+      "epoch": 0.32338715330513823,
+      "grad_norm": 0.3914336562156677,
+      "learning_rate": 1.526808052205312e-05,
+      "loss": 0.2692,
+      "step": 8650
+    },
+    {
+      "epoch": 0.323574082295488,
+      "grad_norm": 0.9589570164680481,
+      "learning_rate": 1.526308801761099e-05,
+      "loss": 0.2913,
+      "step": 8655
+    },
+    {
+      "epoch": 0.3237610112858378,
+      "grad_norm": 0.19636771082878113,
+      "learning_rate": 1.5258093698080614e-05,
+      "loss": 0.3008,
+      "step": 8660
+    },
+    {
+      "epoch": 0.32394794027618756,
+      "grad_norm": 0.27566975355148315,
+      "learning_rate": 1.5253097565184382e-05,
+      "loss": 0.2644,
+      "step": 8665
+    },
+    {
+      "epoch": 0.3241348692665374,
+      "grad_norm": 0.3337952494621277,
+      "learning_rate": 1.5248099620645321e-05,
+      "loss": 0.3182,
+      "step": 8670
+    },
+    {
+      "epoch": 0.3243217982568872,
+      "grad_norm": 0.30488675832748413,
+      "learning_rate": 1.5243099866187076e-05,
+      "loss": 0.3166,
+      "step": 8675
+    },
+    {
+      "epoch": 0.32450872724723695,
+      "grad_norm": 0.2742447853088379,
+      "learning_rate": 1.5238098303533923e-05,
+      "loss": 0.2807,
+      "step": 8680
+    },
+    {
+      "epoch": 0.3246956562375867,
+      "grad_norm": 0.35254549980163574,
+      "learning_rate": 1.5233094934410755e-05,
+      "loss": 0.2999,
+      "step": 8685
+    },
+    {
+      "epoch": 0.32488258522793656,
+      "grad_norm": 0.1497417390346527,
+      "learning_rate": 1.522808976054309e-05,
+      "loss": 0.3525,
+      "step": 8690
+    },
+    {
+      "epoch": 0.32506951421828634,
+      "grad_norm": 0.22152622044086456,
+      "learning_rate": 1.522308278365707e-05,
+      "loss": 0.3084,
+      "step": 8695
+    },
+    {
+      "epoch": 0.3252564432086361,
+      "grad_norm": 0.4548346996307373,
+      "learning_rate": 1.521807400547946e-05,
+      "loss": 0.3306,
+      "step": 8700
+    },
+    {
+      "epoch": 0.3254433721989859,
+      "grad_norm": 0.7274771928787231,
+      "learning_rate": 1.5213063427737639e-05,
+      "loss": 0.2431,
+      "step": 8705
+    },
+    {
+      "epoch": 0.3256303011893357,
+      "grad_norm": 0.4136773347854614,
+      "learning_rate": 1.5208051052159618e-05,
+      "loss": 0.2497,
+      "step": 8710
+    },
+    {
+      "epoch": 0.3258172301796855,
+      "grad_norm": 0.4158473014831543,
+      "learning_rate": 1.520303688047402e-05,
+      "loss": 0.3343,
+      "step": 8715
+    },
+    {
+      "epoch": 0.3260041591700353,
+      "grad_norm": 0.26390954852104187,
+      "learning_rate": 1.5198020914410085e-05,
+      "loss": 0.2489,
+      "step": 8720
+    },
+    {
+      "epoch": 0.32619108816038506,
+      "grad_norm": 0.3240726888179779,
+      "learning_rate": 1.5193003155697681e-05,
+      "loss": 0.3173,
+      "step": 8725
+    },
+    {
+      "epoch": 0.3263780171507349,
+      "grad_norm": 0.4203800857067108,
+      "learning_rate": 1.5187983606067284e-05,
+      "loss": 0.2545,
+      "step": 8730
+    },
+    {
+      "epoch": 0.32656494614108467,
+      "grad_norm": 0.40211477875709534,
+      "learning_rate": 1.5182962267249997e-05,
+      "loss": 0.2805,
+      "step": 8735
+    },
+    {
+      "epoch": 0.32675187513143444,
+      "grad_norm": 0.35995474457740784,
+      "learning_rate": 1.5177939140977535e-05,
+      "loss": 0.3068,
+      "step": 8740
+    },
+    {
+      "epoch": 0.3269388041217842,
+      "grad_norm": 0.36302804946899414,
+      "learning_rate": 1.517291422898223e-05,
+      "loss": 0.2778,
+      "step": 8745
+    },
+    {
+      "epoch": 0.32712573311213405,
+      "grad_norm": 0.4798825681209564,
+      "learning_rate": 1.5167887532997032e-05,
+      "loss": 0.2644,
+      "step": 8750
+    },
+    {
+      "epoch": 0.32731266210248383,
+      "grad_norm": 0.3247123658657074,
+      "learning_rate": 1.51628590547555e-05,
+      "loss": 0.259,
+      "step": 8755
+    },
+    {
+      "epoch": 0.3274995910928336,
+      "grad_norm": 0.23680318892002106,
+      "learning_rate": 1.5157828795991813e-05,
+      "loss": 0.3716,
+      "step": 8760
+    },
+    {
+      "epoch": 0.3276865200831834,
+      "grad_norm": 0.4908321797847748,
+      "learning_rate": 1.5152796758440769e-05,
+      "loss": 0.2849,
+      "step": 8765
+    },
+    {
+      "epoch": 0.3278734490735332,
+      "grad_norm": 0.8543740510940552,
+      "learning_rate": 1.514776294383777e-05,
+      "loss": 0.3428,
+      "step": 8770
+    },
+    {
+      "epoch": 0.328060378063883,
+      "grad_norm": 0.6384404301643372,
+      "learning_rate": 1.514272735391883e-05,
+      "loss": 0.291,
+      "step": 8775
+    },
+    {
+      "epoch": 0.3282473070542328,
+      "grad_norm": 0.33483198285102844,
+      "learning_rate": 1.5137689990420583e-05,
+      "loss": 0.3363,
+      "step": 8780
+    },
+    {
+      "epoch": 0.32843423604458255,
+      "grad_norm": 0.26816031336784363,
+      "learning_rate": 1.5132650855080275e-05,
+      "loss": 0.298,
+      "step": 8785
+    },
+    {
+      "epoch": 0.3286211650349323,
+      "grad_norm": 0.3883032500743866,
+      "learning_rate": 1.5127609949635753e-05,
+      "loss": 0.2433,
+      "step": 8790
+    },
+    {
+      "epoch": 0.32880809402528216,
+      "grad_norm": 0.31391748785972595,
+      "learning_rate": 1.5122567275825486e-05,
+      "loss": 0.3271,
+      "step": 8795
+    },
+    {
+      "epoch": 0.32899502301563194,
+      "grad_norm": 0.5481862425804138,
+      "learning_rate": 1.5117522835388545e-05,
+      "loss": 0.2805,
+      "step": 8800
+    },
+    {
+      "epoch": 0.3291819520059817,
+      "grad_norm": 0.38565173745155334,
+      "learning_rate": 1.5112476630064615e-05,
+      "loss": 0.2974,
+      "step": 8805
+    },
+    {
+      "epoch": 0.3293688809963315,
+      "grad_norm": 0.3033200800418854,
+      "learning_rate": 1.5107428661593983e-05,
+      "loss": 0.2851,
+      "step": 8810
+    },
+    {
+      "epoch": 0.3295558099866813,
+      "grad_norm": 0.3411104381084442,
+      "learning_rate": 1.5102378931717556e-05,
+      "loss": 0.3146,
+      "step": 8815
+    },
+    {
+      "epoch": 0.3297427389770311,
+      "grad_norm": 0.32924318313598633,
+      "learning_rate": 1.5097327442176837e-05,
+      "loss": 0.2499,
+      "step": 8820
+    },
+    {
+      "epoch": 0.3299296679673809,
+      "grad_norm": 0.6186853647232056,
+      "learning_rate": 1.5092274194713933e-05,
+      "loss": 0.3388,
+      "step": 8825
+    },
+    {
+      "epoch": 0.33011659695773066,
+      "grad_norm": 0.2822064757347107,
+      "learning_rate": 1.5087219191071579e-05,
+      "loss": 0.307,
+      "step": 8830
+    },
+    {
+      "epoch": 0.3303035259480805,
+      "grad_norm": 0.4423992931842804,
+      "learning_rate": 1.5082162432993092e-05,
+      "loss": 0.2772,
+      "step": 8835
+    },
+    {
+      "epoch": 0.33049045493843027,
+      "grad_norm": 0.6045133471488953,
+      "learning_rate": 1.5077103922222402e-05,
+      "loss": 0.3682,
+      "step": 8840
+    },
+    {
+      "epoch": 0.33067738392878004,
+      "grad_norm": 0.26116663217544556,
+      "learning_rate": 1.507204366050405e-05,
+      "loss": 0.4336,
+      "step": 8845
+    },
+    {
+      "epoch": 0.3308643129191298,
+      "grad_norm": 0.621884286403656,
+      "learning_rate": 1.5066981649583168e-05,
+      "loss": 0.2781,
+      "step": 8850
+    },
+    {
+      "epoch": 0.33105124190947965,
+      "grad_norm": 0.43372565507888794,
+      "learning_rate": 1.5061917891205504e-05,
+      "loss": 0.3472,
+      "step": 8855
+    },
+    {
+      "epoch": 0.33123817089982943,
+      "grad_norm": 1.028779149055481,
+      "learning_rate": 1.5056852387117405e-05,
+      "loss": 0.4705,
+      "step": 8860
+    },
+    {
+      "epoch": 0.3314250998901792,
+      "grad_norm": 0.4312193691730499,
+      "learning_rate": 1.505178513906581e-05,
+      "loss": 0.3649,
+      "step": 8865
+    },
+    {
+      "epoch": 0.331612028880529,
+      "grad_norm": 0.3952298164367676,
+      "learning_rate": 1.5046716148798273e-05,
+      "loss": 0.3109,
+      "step": 8870
+    },
+    {
+      "epoch": 0.3317989578708788,
+      "grad_norm": 0.19110475480556488,
+      "learning_rate": 1.5041645418062942e-05,
+      "loss": 0.3217,
+      "step": 8875
+    },
+    {
+      "epoch": 0.3319858868612286,
+      "grad_norm": 0.7014629244804382,
+      "learning_rate": 1.5036572948608572e-05,
+      "loss": 0.3183,
+      "step": 8880
+    },
+    {
+      "epoch": 0.3321728158515784,
+      "grad_norm": 0.27941253781318665,
+      "learning_rate": 1.5031498742184507e-05,
+      "loss": 0.2359,
+      "step": 8885
+    },
+    {
+      "epoch": 0.33235974484192815,
+      "grad_norm": 0.29573166370391846,
+      "learning_rate": 1.5026422800540694e-05,
+      "loss": 0.2383,
+      "step": 8890
+    },
+    {
+      "epoch": 0.332546673832278,
+      "grad_norm": 0.36960336565971375,
+      "learning_rate": 1.5021345125427684e-05,
+      "loss": 0.3709,
+      "step": 8895
+    },
+    {
+      "epoch": 0.33273360282262776,
+      "grad_norm": 0.3210650086402893,
+      "learning_rate": 1.501626571859662e-05,
+      "loss": 0.3651,
+      "step": 8900
+    },
+    {
+      "epoch": 0.33292053181297754,
+      "grad_norm": 0.6550555229187012,
+      "learning_rate": 1.5011184581799243e-05,
+      "loss": 0.3616,
+      "step": 8905
+    },
+    {
+      "epoch": 0.3331074608033273,
+      "grad_norm": 0.3219583034515381,
+      "learning_rate": 1.5006101716787896e-05,
+      "loss": 0.3333,
+      "step": 8910
+    },
+    {
+      "epoch": 0.33329438979367715,
+      "grad_norm": 0.40844401717185974,
+      "learning_rate": 1.5001017125315503e-05,
+      "loss": 0.3207,
+      "step": 8915
+    },
+    {
+      "epoch": 0.3334813187840269,
+      "grad_norm": 0.5592089295387268,
+      "learning_rate": 1.4995930809135604e-05,
+      "loss": 0.2584,
+      "step": 8920
+    },
+    {
+      "epoch": 0.3336682477743767,
+      "grad_norm": 0.3412552773952484,
+      "learning_rate": 1.4990842770002321e-05,
+      "loss": 0.3286,
+      "step": 8925
+    },
+    {
+      "epoch": 0.3338551767647265,
+      "grad_norm": 0.2517843246459961,
+      "learning_rate": 1.4985753009670375e-05,
+      "loss": 0.3027,
+      "step": 8930
+    },
+    {
+      "epoch": 0.3340421057550763,
+      "grad_norm": 0.274789422750473,
+      "learning_rate": 1.4980661529895073e-05,
+      "loss": 0.3058,
+      "step": 8935
+    },
+    {
+      "epoch": 0.3342290347454261,
+      "grad_norm": 0.3245002031326294,
+      "learning_rate": 1.4975568332432322e-05,
+      "loss": 0.356,
+      "step": 8940
+    },
+    {
+      "epoch": 0.33441596373577587,
+      "grad_norm": 0.2871999442577362,
+      "learning_rate": 1.4970473419038623e-05,
+      "loss": 0.2805,
+      "step": 8945
+    },
+    {
+      "epoch": 0.33460289272612564,
+      "grad_norm": 0.3988129496574402,
+      "learning_rate": 1.4965376791471062e-05,
+      "loss": 0.2596,
+      "step": 8950
+    },
+    {
+      "epoch": 0.3347898217164755,
+      "grad_norm": 0.3798826038837433,
+      "learning_rate": 1.4960278451487327e-05,
+      "loss": 0.3168,
+      "step": 8955
+    },
+    {
+      "epoch": 0.33497675070682525,
+      "grad_norm": 0.336479127407074,
+      "learning_rate": 1.4955178400845678e-05,
+      "loss": 0.294,
+      "step": 8960
+    },
+    {
+      "epoch": 0.33516367969717503,
+      "grad_norm": 0.31239262223243713,
+      "learning_rate": 1.4950076641304984e-05,
+      "loss": 0.2486,
+      "step": 8965
+    },
+    {
+      "epoch": 0.3353506086875248,
+      "grad_norm": 0.4820721745491028,
+      "learning_rate": 1.4944973174624695e-05,
+      "loss": 0.3584,
+      "step": 8970
+    },
+    {
+      "epoch": 0.33553753767787464,
+      "grad_norm": 0.7084938287734985,
+      "learning_rate": 1.493986800256485e-05,
+      "loss": 0.2835,
+      "step": 8975
+    },
+    {
+      "epoch": 0.3357244666682244,
+      "grad_norm": 0.13603666424751282,
+      "learning_rate": 1.4934761126886077e-05,
+      "loss": 0.2406,
+      "step": 8980
+    },
+    {
+      "epoch": 0.3359113956585742,
+      "grad_norm": 0.28574085235595703,
+      "learning_rate": 1.4929652549349587e-05,
+      "loss": 0.331,
+      "step": 8985
+    },
+    {
+      "epoch": 0.336098324648924,
+      "grad_norm": 0.3157019317150116,
+      "learning_rate": 1.4924542271717186e-05,
+      "loss": 0.2754,
+      "step": 8990
+    },
+    {
+      "epoch": 0.3362852536392738,
+      "grad_norm": 0.4523063004016876,
+      "learning_rate": 1.4919430295751262e-05,
+      "loss": 0.2865,
+      "step": 8995
+    },
+    {
+      "epoch": 0.3364721826296236,
+      "grad_norm": 0.32611748576164246,
+      "learning_rate": 1.4914316623214788e-05,
+      "loss": 0.3052,
+      "step": 9000
+    },
+    {
+      "epoch": 0.33665911161997336,
+      "grad_norm": 0.34001392126083374,
+      "learning_rate": 1.4909201255871325e-05,
+      "loss": 0.253,
+      "step": 9005
+    },
+    {
+      "epoch": 0.33684604061032314,
+      "grad_norm": 0.3103577792644501,
+      "learning_rate": 1.4904084195485014e-05,
+      "loss": 0.2962,
+      "step": 9010
+    },
+    {
+      "epoch": 0.33703296960067297,
+      "grad_norm": 0.3384459614753723,
+      "learning_rate": 1.4898965443820584e-05,
+      "loss": 0.2931,
+      "step": 9015
+    },
+    {
+      "epoch": 0.33721989859102275,
+      "grad_norm": 0.47207170724868774,
+      "learning_rate": 1.4893845002643345e-05,
+      "loss": 0.3106,
+      "step": 9020
+    },
+    {
+      "epoch": 0.3374068275813725,
+      "grad_norm": 0.25210464000701904,
+      "learning_rate": 1.4888722873719195e-05,
+      "loss": 0.3618,
+      "step": 9025
+    },
+    {
+      "epoch": 0.3375937565717223,
+      "grad_norm": 0.4364457130432129,
+      "learning_rate": 1.4883599058814602e-05,
+      "loss": 0.3376,
+      "step": 9030
+    },
+    {
+      "epoch": 0.3377806855620721,
+      "grad_norm": 0.4352976381778717,
+      "learning_rate": 1.4878473559696625e-05,
+      "loss": 0.3076,
+      "step": 9035
+    },
+    {
+      "epoch": 0.3379676145524219,
+      "grad_norm": 0.4709399342536926,
+      "learning_rate": 1.4873346378132905e-05,
+      "loss": 0.301,
+      "step": 9040
+    },
+    {
+      "epoch": 0.3381545435427717,
+      "grad_norm": 0.4114588499069214,
+      "learning_rate": 1.4868217515891657e-05,
+      "loss": 0.2685,
+      "step": 9045
+    },
+    {
+      "epoch": 0.33834147253312147,
+      "grad_norm": 0.7446982860565186,
+      "learning_rate": 1.4863086974741684e-05,
+      "loss": 0.294,
+      "step": 9050
+    },
+    {
+      "epoch": 0.33852840152347125,
+      "grad_norm": 0.29010334610939026,
+      "learning_rate": 1.4857954756452353e-05,
+      "loss": 0.356,
+      "step": 9055
+    },
+    {
+      "epoch": 0.3387153305138211,
+      "grad_norm": 0.4005201756954193,
+      "learning_rate": 1.4852820862793626e-05,
+      "loss": 0.2966,
+      "step": 9060
+    },
+    {
+      "epoch": 0.33890225950417086,
+      "grad_norm": 0.7483821511268616,
+      "learning_rate": 1.4847685295536037e-05,
+      "loss": 0.3161,
+      "step": 9065
+    },
+    {
+      "epoch": 0.33908918849452063,
+      "grad_norm": 0.4686259925365448,
+      "learning_rate": 1.4842548056450692e-05,
+      "loss": 0.3532,
+      "step": 9070
+    },
+    {
+      "epoch": 0.3392761174848704,
+      "grad_norm": 0.44194498658180237,
+      "learning_rate": 1.4837409147309276e-05,
+      "loss": 0.2582,
+      "step": 9075
+    },
+    {
+      "epoch": 0.33946304647522024,
+      "grad_norm": 0.47664833068847656,
+      "learning_rate": 1.4832268569884058e-05,
+      "loss": 0.3903,
+      "step": 9080
+    },
+    {
+      "epoch": 0.33964997546557,
+      "grad_norm": 0.488346129655838,
+      "learning_rate": 1.4827126325947872e-05,
+      "loss": 0.2816,
+      "step": 9085
+    },
+    {
+      "epoch": 0.3398369044559198,
+      "grad_norm": 0.3450961709022522,
+      "learning_rate": 1.4821982417274128e-05,
+      "loss": 0.3073,
+      "step": 9090
+    },
+    {
+      "epoch": 0.3400238334462696,
+      "grad_norm": 0.32565736770629883,
+      "learning_rate": 1.4816836845636817e-05,
+      "loss": 0.3467,
+      "step": 9095
+    },
+    {
+      "epoch": 0.3402107624366194,
+      "grad_norm": 0.3779214322566986,
+      "learning_rate": 1.4811689612810498e-05,
+      "loss": 0.2372,
+      "step": 9100
+    },
+    {
+      "epoch": 0.3403976914269692,
+      "grad_norm": 11.968618392944336,
+      "learning_rate": 1.4806540720570306e-05,
+      "loss": 0.4737,
+      "step": 9105
+    },
+    {
+      "epoch": 0.34058462041731896,
+      "grad_norm": 0.4189259111881256,
+      "learning_rate": 1.4801390170691941e-05,
+      "loss": 0.2739,
+      "step": 9110
+    },
+    {
+      "epoch": 0.34077154940766874,
+      "grad_norm": 0.33259811997413635,
+      "learning_rate": 1.4796237964951686e-05,
+      "loss": 0.2829,
+      "step": 9115
+    },
+    {
+      "epoch": 0.34095847839801857,
+      "grad_norm": 0.2681080400943756,
+      "learning_rate": 1.4791084105126385e-05,
+      "loss": 0.2564,
+      "step": 9120
+    },
+    {
+      "epoch": 0.34114540738836835,
+      "grad_norm": 0.35501420497894287,
+      "learning_rate": 1.478592859299346e-05,
+      "loss": 0.2472,
+      "step": 9125
+    },
+    {
+      "epoch": 0.3413323363787181,
+      "grad_norm": 0.4222002625465393,
+      "learning_rate": 1.4780771430330894e-05,
+      "loss": 0.3695,
+      "step": 9130
+    },
+    {
+      "epoch": 0.3415192653690679,
+      "grad_norm": 0.6013131737709045,
+      "learning_rate": 1.477561261891725e-05,
+      "loss": 0.2431,
+      "step": 9135
+    },
+    {
+      "epoch": 0.34170619435941774,
+      "grad_norm": 0.4116293787956238,
+      "learning_rate": 1.4770452160531652e-05,
+      "loss": 0.2915,
+      "step": 9140
+    },
+    {
+      "epoch": 0.3418931233497675,
+      "grad_norm": 0.4117601811885834,
+      "learning_rate": 1.4765290056953796e-05,
+      "loss": 0.2381,
+      "step": 9145
+    },
+    {
+      "epoch": 0.3420800523401173,
+      "grad_norm": 0.5726933479309082,
+      "learning_rate": 1.476012630996394e-05,
+      "loss": 0.3691,
+      "step": 9150
+    },
+    {
+      "epoch": 0.34226698133046707,
+      "grad_norm": 0.43034738302230835,
+      "learning_rate": 1.4754960921342916e-05,
+      "loss": 0.2324,
+      "step": 9155
+    },
+    {
+      "epoch": 0.3424539103208169,
+      "grad_norm": 0.23042064905166626,
+      "learning_rate": 1.4749793892872115e-05,
+      "loss": 0.2359,
+      "step": 9160
+    },
+    {
+      "epoch": 0.3426408393111667,
+      "grad_norm": 0.4107089340686798,
+      "learning_rate": 1.4744625226333502e-05,
+      "loss": 0.2775,
+      "step": 9165
+    },
+    {
+      "epoch": 0.34282776830151646,
+      "grad_norm": 0.1352885216474533,
+      "learning_rate": 1.47394549235096e-05,
+      "loss": 0.2663,
+      "step": 9170
+    },
+    {
+      "epoch": 0.34301469729186623,
+      "grad_norm": 0.34713098406791687,
+      "learning_rate": 1.4734282986183494e-05,
+      "loss": 0.3152,
+      "step": 9175
+    },
+    {
+      "epoch": 0.34320162628221607,
+      "grad_norm": 0.3076459765434265,
+      "learning_rate": 1.4729109416138843e-05,
+      "loss": 0.3371,
+      "step": 9180
+    },
+    {
+      "epoch": 0.34338855527256584,
+      "grad_norm": 0.2737236022949219,
+      "learning_rate": 1.4723934215159858e-05,
+      "loss": 0.3022,
+      "step": 9185
+    },
+    {
+      "epoch": 0.3435754842629156,
+      "grad_norm": 0.3773277997970581,
+      "learning_rate": 1.4718757385031321e-05,
+      "loss": 0.2924,
+      "step": 9190
+    },
+    {
+      "epoch": 0.3437624132532654,
+      "grad_norm": 0.39832785725593567,
+      "learning_rate": 1.4713578927538573e-05,
+      "loss": 0.2438,
+      "step": 9195
+    },
+    {
+      "epoch": 0.34394934224361523,
+      "grad_norm": 0.19407474994659424,
+      "learning_rate": 1.4708398844467512e-05,
+      "loss": 0.3117,
+      "step": 9200
+    },
+    {
+      "epoch": 0.344136271233965,
+      "grad_norm": 0.3654159605503082,
+      "learning_rate": 1.4703217137604604e-05,
+      "loss": 0.3485,
+      "step": 9205
+    },
+    {
+      "epoch": 0.3443232002243148,
+      "grad_norm": 0.37599438428878784,
+      "learning_rate": 1.4698033808736867e-05,
+      "loss": 0.2486,
+      "step": 9210
+    },
+    {
+      "epoch": 0.34451012921466456,
+      "grad_norm": 0.42682337760925293,
+      "learning_rate": 1.4692848859651889e-05,
+      "loss": 0.2533,
+      "step": 9215
+    },
+    {
+      "epoch": 0.3446970582050144,
+      "grad_norm": 0.3610582947731018,
+      "learning_rate": 1.4687662292137804e-05,
+      "loss": 0.2781,
+      "step": 9220
+    },
+    {
+      "epoch": 0.3448839871953642,
+      "grad_norm": 0.4574204385280609,
+      "learning_rate": 1.4682474107983314e-05,
+      "loss": 0.2965,
+      "step": 9225
+    },
+    {
+      "epoch": 0.34507091618571395,
+      "grad_norm": 0.513286828994751,
+      "learning_rate": 1.467728430897767e-05,
+      "loss": 0.2859,
+      "step": 9230
+    },
+    {
+      "epoch": 0.3452578451760637,
+      "grad_norm": 0.23457178473472595,
+      "learning_rate": 1.4672092896910692e-05,
+      "loss": 0.2623,
+      "step": 9235
+    },
+    {
+      "epoch": 0.34544477416641356,
+      "grad_norm": 0.34765762090682983,
+      "learning_rate": 1.4666899873572747e-05,
+      "loss": 0.291,
+      "step": 9240
+    },
+    {
+      "epoch": 0.34563170315676334,
+      "grad_norm": 0.3738343417644501,
+      "learning_rate": 1.4661705240754757e-05,
+      "loss": 0.2701,
+      "step": 9245
+    },
+    {
+      "epoch": 0.3458186321471131,
+      "grad_norm": 0.3144400119781494,
+      "learning_rate": 1.4656509000248207e-05,
+      "loss": 0.2842,
+      "step": 9250
+    },
+    {
+      "epoch": 0.3460055611374629,
+      "grad_norm": 0.5743440985679626,
+      "learning_rate": 1.4651311153845127e-05,
+      "loss": 0.2436,
+      "step": 9255
+    },
+    {
+      "epoch": 0.3461924901278127,
+      "grad_norm": 0.2858171761035919,
+      "learning_rate": 1.4646111703338108e-05,
+      "loss": 0.318,
+      "step": 9260
+    },
+    {
+      "epoch": 0.3463794191181625,
+      "grad_norm": 0.4376906156539917,
+      "learning_rate": 1.464091065052029e-05,
+      "loss": 0.2822,
+      "step": 9265
+    },
+    {
+      "epoch": 0.3465663481085123,
+      "grad_norm": 0.2567811906337738,
+      "learning_rate": 1.4635707997185367e-05,
+      "loss": 0.2571,
+      "step": 9270
+    },
+    {
+      "epoch": 0.34675327709886206,
+      "grad_norm": 0.3876672089099884,
+      "learning_rate": 1.4630503745127587e-05,
+      "loss": 0.3234,
+      "step": 9275
+    },
+    {
+      "epoch": 0.34694020608921183,
+      "grad_norm": 0.48617517948150635,
+      "learning_rate": 1.4625297896141741e-05,
+      "loss": 0.2896,
+      "step": 9280
+    },
+    {
+      "epoch": 0.34712713507956167,
+      "grad_norm": 0.4633631110191345,
+      "learning_rate": 1.462009045202319e-05,
+      "loss": 0.3156,
+      "step": 9285
+    },
+    {
+      "epoch": 0.34731406406991144,
+      "grad_norm": 0.18796226382255554,
+      "learning_rate": 1.461488141456782e-05,
+      "loss": 0.2849,
+      "step": 9290
+    },
+    {
+      "epoch": 0.3475009930602612,
+      "grad_norm": 0.8571456074714661,
+      "learning_rate": 1.4609670785572084e-05,
+      "loss": 0.3728,
+      "step": 9295
+    },
+    {
+      "epoch": 0.347687922050611,
+      "grad_norm": 0.22898966073989868,
+      "learning_rate": 1.4604458566832977e-05,
+      "loss": 0.2989,
+      "step": 9300
+    },
+    {
+      "epoch": 0.34787485104096083,
+      "grad_norm": 0.32521340250968933,
+      "learning_rate": 1.4599244760148046e-05,
+      "loss": 0.2966,
+      "step": 9305
+    },
+    {
+      "epoch": 0.3480617800313106,
+      "grad_norm": 0.6073753833770752,
+      "learning_rate": 1.459402936731538e-05,
+      "loss": 0.3742,
+      "step": 9310
+    },
+    {
+      "epoch": 0.3482487090216604,
+      "grad_norm": 0.3876205384731293,
+      "learning_rate": 1.4588812390133624e-05,
+      "loss": 0.249,
+      "step": 9315
+    },
+    {
+      "epoch": 0.34843563801201016,
+      "grad_norm": 0.46591299772262573,
+      "learning_rate": 1.4583593830401958e-05,
+      "loss": 0.2723,
+      "step": 9320
+    },
+    {
+      "epoch": 0.34862256700236,
+      "grad_norm": 0.22240598499774933,
+      "learning_rate": 1.4578373689920116e-05,
+      "loss": 0.3037,
+      "step": 9325
+    },
+    {
+      "epoch": 0.3488094959927098,
+      "grad_norm": 0.5356066823005676,
+      "learning_rate": 1.457315197048838e-05,
+      "loss": 0.3212,
+      "step": 9330
+    },
+    {
+      "epoch": 0.34899642498305955,
+      "grad_norm": 0.5799317955970764,
+      "learning_rate": 1.4567928673907563e-05,
+      "loss": 0.3065,
+      "step": 9335
+    },
+    {
+      "epoch": 0.3491833539734093,
+      "grad_norm": 0.2084297090768814,
+      "learning_rate": 1.4562703801979035e-05,
+      "loss": 0.3081,
+      "step": 9340
+    },
+    {
+      "epoch": 0.34937028296375916,
+      "grad_norm": 0.520650327205658,
+      "learning_rate": 1.4557477356504702e-05,
+      "loss": 0.3058,
+      "step": 9345
+    },
+    {
+      "epoch": 0.34955721195410894,
+      "grad_norm": 0.34877529740333557,
+      "learning_rate": 1.455224933928702e-05,
+      "loss": 0.2876,
+      "step": 9350
+    },
+    {
+      "epoch": 0.3497441409444587,
+      "grad_norm": 0.30944302678108215,
+      "learning_rate": 1.4547019752128977e-05,
+      "loss": 0.3491,
+      "step": 9355
+    },
+    {
+      "epoch": 0.3499310699348085,
+      "grad_norm": 0.5403395295143127,
+      "learning_rate": 1.4541788596834111e-05,
+      "loss": 0.3018,
+      "step": 9360
+    },
+    {
+      "epoch": 0.3501179989251583,
+      "grad_norm": 0.43393194675445557,
+      "learning_rate": 1.4536555875206497e-05,
+      "loss": 0.3163,
+      "step": 9365
+    },
+    {
+      "epoch": 0.3503049279155081,
+      "grad_norm": 0.3072391450405121,
+      "learning_rate": 1.4531321589050749e-05,
+      "loss": 0.2298,
+      "step": 9370
+    },
+    {
+      "epoch": 0.3504918569058579,
+      "grad_norm": 0.18442721664905548,
+      "learning_rate": 1.4526085740172025e-05,
+      "loss": 0.4021,
+      "step": 9375
+    },
+    {
+      "epoch": 0.35067878589620766,
+      "grad_norm": 0.4270346462726593,
+      "learning_rate": 1.4520848330376019e-05,
+      "loss": 0.2749,
+      "step": 9380
+    },
+    {
+      "epoch": 0.3508657148865575,
+      "grad_norm": 0.4557771384716034,
+      "learning_rate": 1.4515609361468959e-05,
+      "loss": 0.2412,
+      "step": 9385
+    },
+    {
+      "epoch": 0.35105264387690727,
+      "grad_norm": 0.7857339978218079,
+      "learning_rate": 1.4510368835257622e-05,
+      "loss": 0.2772,
+      "step": 9390
+    },
+    {
+      "epoch": 0.35123957286725704,
+      "grad_norm": 0.30803146958351135,
+      "learning_rate": 1.450512675354931e-05,
+      "loss": 0.3163,
+      "step": 9395
+    },
+    {
+      "epoch": 0.3514265018576068,
+      "grad_norm": 0.4320001006126404,
+      "learning_rate": 1.449988311815187e-05,
+      "loss": 0.2661,
+      "step": 9400
+    },
+    {
+      "epoch": 0.35161343084795665,
+      "grad_norm": 0.13867487013339996,
+      "learning_rate": 1.4494637930873683e-05,
+      "loss": 0.2416,
+      "step": 9405
+    },
+    {
+      "epoch": 0.35180035983830643,
+      "grad_norm": 0.27354222536087036,
+      "learning_rate": 1.4489391193523658e-05,
+      "loss": 0.2661,
+      "step": 9410
+    },
+    {
+      "epoch": 0.3519872888286562,
+      "grad_norm": 0.3791649043560028,
+      "learning_rate": 1.4484142907911247e-05,
+      "loss": 0.2581,
+      "step": 9415
+    },
+    {
+      "epoch": 0.352174217819006,
+      "grad_norm": 0.38127240538597107,
+      "learning_rate": 1.4478893075846436e-05,
+      "loss": 0.2605,
+      "step": 9420
+    },
+    {
+      "epoch": 0.3523611468093558,
+      "grad_norm": 0.23698148131370544,
+      "learning_rate": 1.447364169913974e-05,
+      "loss": 0.3725,
+      "step": 9425
+    },
+    {
+      "epoch": 0.3525480757997056,
+      "grad_norm": 0.3475520610809326,
+      "learning_rate": 1.4468388779602207e-05,
+      "loss": 0.2851,
+      "step": 9430
+    },
+    {
+      "epoch": 0.3527350047900554,
+      "grad_norm": 0.45811226963996887,
+      "learning_rate": 1.446313431904542e-05,
+      "loss": 0.3581,
+      "step": 9435
+    },
+    {
+      "epoch": 0.35292193378040515,
+      "grad_norm": 0.6400048136711121,
+      "learning_rate": 1.4457878319281491e-05,
+      "loss": 0.2837,
+      "step": 9440
+    },
+    {
+      "epoch": 0.353108862770755,
+      "grad_norm": 0.4961070716381073,
+      "learning_rate": 1.4452620782123063e-05,
+      "loss": 0.3202,
+      "step": 9445
+    },
+    {
+      "epoch": 0.35329579176110476,
+      "grad_norm": 0.33612340688705444,
+      "learning_rate": 1.4447361709383312e-05,
+      "loss": 0.2923,
+      "step": 9450
+    },
+    {
+      "epoch": 0.35348272075145454,
+      "grad_norm": 0.4333072006702423,
+      "learning_rate": 1.4442101102875942e-05,
+      "loss": 0.2749,
+      "step": 9455
+    },
+    {
+      "epoch": 0.3536696497418043,
+      "grad_norm": 0.3720749616622925,
+      "learning_rate": 1.443683896441518e-05,
+      "loss": 0.333,
+      "step": 9460
+    },
+    {
+      "epoch": 0.35385657873215415,
+      "grad_norm": 0.638292670249939,
+      "learning_rate": 1.4431575295815793e-05,
+      "loss": 0.2825,
+      "step": 9465
+    },
+    {
+      "epoch": 0.3540435077225039,
+      "grad_norm": 0.27355220913887024,
+      "learning_rate": 1.4426310098893069e-05,
+      "loss": 0.334,
+      "step": 9470
+    },
+    {
+      "epoch": 0.3542304367128537,
+      "grad_norm": 0.3119034171104431,
+      "learning_rate": 1.4421043375462821e-05,
+      "loss": 0.3194,
+      "step": 9475
+    },
+    {
+      "epoch": 0.3544173657032035,
+      "grad_norm": 0.3145153820514679,
+      "learning_rate": 1.441577512734139e-05,
+      "loss": 0.2671,
+      "step": 9480
+    },
+    {
+      "epoch": 0.3546042946935533,
+      "grad_norm": 0.25988125801086426,
+      "learning_rate": 1.4410505356345645e-05,
+      "loss": 0.2174,
+      "step": 9485
+    },
+    {
+      "epoch": 0.3547912236839031,
+      "grad_norm": 0.4911235570907593,
+      "learning_rate": 1.440523406429298e-05,
+      "loss": 0.4387,
+      "step": 9490
+    },
+    {
+      "epoch": 0.35497815267425287,
+      "grad_norm": 0.21677200496196747,
+      "learning_rate": 1.4399961253001316e-05,
+      "loss": 0.2656,
+      "step": 9495
+    },
+    {
+      "epoch": 0.35516508166460264,
+      "grad_norm": 0.3045201897621155,
+      "learning_rate": 1.4394686924289087e-05,
+      "loss": 0.2927,
+      "step": 9500
+    },
+    {
+      "epoch": 0.3553520106549525,
+      "grad_norm": 0.38031861186027527,
+      "learning_rate": 1.4389411079975262e-05,
+      "loss": 0.333,
+      "step": 9505
+    },
+    {
+      "epoch": 0.35553893964530225,
+      "grad_norm": 0.31916332244873047,
+      "learning_rate": 1.4384133721879327e-05,
+      "loss": 0.2274,
+      "step": 9510
+    },
+    {
+      "epoch": 0.35572586863565203,
+      "grad_norm": 0.4681374430656433,
+      "learning_rate": 1.4378854851821294e-05,
+      "loss": 0.2061,
+      "step": 9515
+    },
+    {
+      "epoch": 0.3559127976260018,
+      "grad_norm": 0.43480637669563293,
+      "learning_rate": 1.4373574471621694e-05,
+      "loss": 0.3167,
+      "step": 9520
+    },
+    {
+      "epoch": 0.3560997266163516,
+      "grad_norm": 0.2742566764354706,
+      "learning_rate": 1.4368292583101578e-05,
+      "loss": 0.3607,
+      "step": 9525
+    },
+    {
+      "epoch": 0.3562866556067014,
+      "grad_norm": 0.2994193732738495,
+      "learning_rate": 1.4363009188082513e-05,
+      "loss": 0.34,
+      "step": 9530
+    },
+    {
+      "epoch": 0.3564735845970512,
+      "grad_norm": 0.38814255595207214,
+      "learning_rate": 1.43577242883866e-05,
+      "loss": 0.2461,
+      "step": 9535
+    },
+    {
+      "epoch": 0.356660513587401,
+      "grad_norm": 0.29775434732437134,
+      "learning_rate": 1.4352437885836441e-05,
+      "loss": 0.2581,
+      "step": 9540
+    },
+    {
+      "epoch": 0.35684744257775075,
+      "grad_norm": 0.36987629532814026,
+      "learning_rate": 1.4347149982255168e-05,
+      "loss": 0.3376,
+      "step": 9545
+    },
+    {
+      "epoch": 0.3570343715681006,
+      "grad_norm": 0.2949317693710327,
+      "learning_rate": 1.4341860579466428e-05,
+      "loss": 0.3628,
+      "step": 9550
+    },
+    {
+      "epoch": 0.35722130055845036,
+      "grad_norm": 0.3889232873916626,
+      "learning_rate": 1.4336569679294385e-05,
+      "loss": 0.2386,
+      "step": 9555
+    },
+    {
+      "epoch": 0.35740822954880014,
+      "grad_norm": 0.498597115278244,
+      "learning_rate": 1.433127728356372e-05,
+      "loss": 0.2282,
+      "step": 9560
+    },
+    {
+      "epoch": 0.3575951585391499,
+      "grad_norm": 0.332461416721344,
+      "learning_rate": 1.4325983394099626e-05,
+      "loss": 0.3273,
+      "step": 9565
+    },
+    {
+      "epoch": 0.35778208752949975,
+      "grad_norm": 0.5877690315246582,
+      "learning_rate": 1.4320688012727814e-05,
+      "loss": 0.2397,
+      "step": 9570
+    },
+    {
+      "epoch": 0.3579690165198495,
+      "grad_norm": 0.1286364197731018,
+      "learning_rate": 1.4315391141274508e-05,
+      "loss": 0.2346,
+      "step": 9575
+    },
+    {
+      "epoch": 0.3581559455101993,
+      "grad_norm": 0.3990229368209839,
+      "learning_rate": 1.4310092781566452e-05,
+      "loss": 0.3775,
+      "step": 9580
+    },
+    {
+      "epoch": 0.3583428745005491,
+      "grad_norm": 0.4049658179283142,
+      "learning_rate": 1.4304792935430893e-05,
+      "loss": 0.2323,
+      "step": 9585
+    },
+    {
+      "epoch": 0.3585298034908989,
+      "grad_norm": 0.27659088373184204,
+      "learning_rate": 1.42994916046956e-05,
+      "loss": 0.2626,
+      "step": 9590
+    },
+    {
+      "epoch": 0.3587167324812487,
+      "grad_norm": 0.38999220728874207,
+      "learning_rate": 1.4294188791188846e-05,
+      "loss": 0.3191,
+      "step": 9595
+    },
+    {
+      "epoch": 0.35890366147159847,
+      "grad_norm": 0.5159905552864075,
+      "learning_rate": 1.4288884496739424e-05,
+      "loss": 0.2849,
+      "step": 9600
+    },
+    {
+      "epoch": 0.35909059046194824,
+      "grad_norm": 0.3492240905761719,
+      "learning_rate": 1.4283578723176632e-05,
+      "loss": 0.3292,
+      "step": 9605
+    },
+    {
+      "epoch": 0.3592775194522981,
+      "grad_norm": 0.3294449746608734,
+      "learning_rate": 1.4278271472330277e-05,
+      "loss": 0.3421,
+      "step": 9610
+    },
+    {
+      "epoch": 0.35946444844264785,
+      "grad_norm": 0.543605625629425,
+      "learning_rate": 1.4272962746030678e-05,
+      "loss": 0.3525,
+      "step": 9615
+    },
+    {
+      "epoch": 0.35965137743299763,
+      "grad_norm": 0.5134566426277161,
+      "learning_rate": 1.4267652546108668e-05,
+      "loss": 0.2269,
+      "step": 9620
+    },
+    {
+      "epoch": 0.3598383064233474,
+      "grad_norm": 0.30253830552101135,
+      "learning_rate": 1.4262340874395574e-05,
+      "loss": 0.2389,
+      "step": 9625
+    },
+    {
+      "epoch": 0.36002523541369724,
+      "grad_norm": 0.30939239263534546,
+      "learning_rate": 1.4257027732723247e-05,
+      "loss": 0.3656,
+      "step": 9630
+    },
+    {
+      "epoch": 0.360212164404047,
+      "grad_norm": 0.3319133222103119,
+      "learning_rate": 1.4251713122924034e-05,
+      "loss": 0.2553,
+      "step": 9635
+    },
+    {
+      "epoch": 0.3603990933943968,
+      "grad_norm": 0.37184301018714905,
+      "learning_rate": 1.424639704683079e-05,
+      "loss": 0.305,
+      "step": 9640
+    },
+    {
+      "epoch": 0.3605860223847466,
+      "grad_norm": 0.4830484688282013,
+      "learning_rate": 1.424107950627688e-05,
+      "loss": 0.2698,
+      "step": 9645
+    },
+    {
+      "epoch": 0.3607729513750964,
+      "grad_norm": 0.5102275013923645,
+      "learning_rate": 1.423576050309617e-05,
+      "loss": 0.2744,
+      "step": 9650
+    },
+    {
+      "epoch": 0.3609598803654462,
+      "grad_norm": 0.2823433578014374,
+      "learning_rate": 1.4230440039123032e-05,
+      "loss": 0.2687,
+      "step": 9655
+    },
+    {
+      "epoch": 0.36114680935579596,
+      "grad_norm": 0.36009663343429565,
+      "learning_rate": 1.4225118116192344e-05,
+      "loss": 0.3377,
+      "step": 9660
+    },
+    {
+      "epoch": 0.36133373834614574,
+      "grad_norm": 0.5339310765266418,
+      "learning_rate": 1.4219794736139484e-05,
+      "loss": 0.2668,
+      "step": 9665
+    },
+    {
+      "epoch": 0.36152066733649557,
+      "grad_norm": 0.565954864025116,
+      "learning_rate": 1.421446990080033e-05,
+      "loss": 0.3104,
+      "step": 9670
+    },
+    {
+      "epoch": 0.36170759632684535,
+      "grad_norm": 0.5683127641677856,
+      "learning_rate": 1.4209143612011268e-05,
+      "loss": 0.2727,
+      "step": 9675
+    },
+    {
+      "epoch": 0.3618945253171951,
+      "grad_norm": 0.4476057291030884,
+      "learning_rate": 1.4203815871609177e-05,
+      "loss": 0.2485,
+      "step": 9680
+    },
+    {
+      "epoch": 0.3620814543075449,
+      "grad_norm": 0.3725626468658447,
+      "learning_rate": 1.419848668143145e-05,
+      "loss": 0.288,
+      "step": 9685
+    },
+    {
+      "epoch": 0.36226838329789474,
+      "grad_norm": 0.5492215752601624,
+      "learning_rate": 1.4193156043315967e-05,
+      "loss": 0.3209,
+      "step": 9690
+    },
+    {
+      "epoch": 0.3624553122882445,
+      "grad_norm": 0.3779137134552002,
+      "learning_rate": 1.4187823959101116e-05,
+      "loss": 0.2553,
+      "step": 9695
+    },
+    {
+      "epoch": 0.3626422412785943,
+      "grad_norm": 0.296917587518692,
+      "learning_rate": 1.4182490430625775e-05,
+      "loss": 0.3266,
+      "step": 9700
+    },
+    {
+      "epoch": 0.36282917026894407,
+      "grad_norm": 0.2786087691783905,
+      "learning_rate": 1.417715545972933e-05,
+      "loss": 0.2164,
+      "step": 9705
+    },
+    {
+      "epoch": 0.3630160992592939,
+      "grad_norm": 0.44238272309303284,
+      "learning_rate": 1.4171819048251658e-05,
+      "loss": 0.2699,
+      "step": 9710
+    },
+    {
+      "epoch": 0.3632030282496437,
+      "grad_norm": 0.5811543464660645,
+      "learning_rate": 1.4166481198033136e-05,
+      "loss": 0.2559,
+      "step": 9715
+    },
+    {
+      "epoch": 0.36338995723999346,
+      "grad_norm": 0.26523107290267944,
+      "learning_rate": 1.4161141910914632e-05,
+      "loss": 0.2773,
+      "step": 9720
+    },
+    {
+      "epoch": 0.36357688623034323,
+      "grad_norm": 0.5079830288887024,
+      "learning_rate": 1.415580118873751e-05,
+      "loss": 0.2821,
+      "step": 9725
+    },
+    {
+      "epoch": 0.36376381522069307,
+      "grad_norm": 0.44632482528686523,
+      "learning_rate": 1.4150459033343639e-05,
+      "loss": 0.3243,
+      "step": 9730
+    },
+    {
+      "epoch": 0.36395074421104284,
+      "grad_norm": 0.3747384250164032,
+      "learning_rate": 1.4145115446575377e-05,
+      "loss": 0.3253,
+      "step": 9735
+    },
+    {
+      "epoch": 0.3641376732013926,
+      "grad_norm": 0.423495888710022,
+      "learning_rate": 1.4139770430275567e-05,
+      "loss": 0.259,
+      "step": 9740
+    },
+    {
+      "epoch": 0.3643246021917424,
+      "grad_norm": 0.365408718585968,
+      "learning_rate": 1.4134423986287555e-05,
+      "loss": 0.3372,
+      "step": 9745
+    },
+    {
+      "epoch": 0.36451153118209223,
+      "grad_norm": 0.31306222081184387,
+      "learning_rate": 1.4129076116455176e-05,
+      "loss": 0.2578,
+      "step": 9750
+    },
+    {
+      "epoch": 0.364698460172442,
+      "grad_norm": 0.4542996287345886,
+      "learning_rate": 1.4123726822622758e-05,
+      "loss": 0.3076,
+      "step": 9755
+    },
+    {
+      "epoch": 0.3648853891627918,
+      "grad_norm": 0.2852497398853302,
+      "learning_rate": 1.411837610663512e-05,
+      "loss": 0.2407,
+      "step": 9760
+    },
+    {
+      "epoch": 0.36507231815314156,
+      "grad_norm": 0.4517068564891815,
+      "learning_rate": 1.411302397033757e-05,
+      "loss": 0.2958,
+      "step": 9765
+    },
+    {
+      "epoch": 0.36525924714349134,
+      "grad_norm": 0.24500373005867004,
+      "learning_rate": 1.4107670415575902e-05,
+      "loss": 0.3155,
+      "step": 9770
+    },
+    {
+      "epoch": 0.36544617613384117,
+      "grad_norm": 0.35521599650382996,
+      "learning_rate": 1.410231544419641e-05,
+      "loss": 0.2573,
+      "step": 9775
+    },
+    {
+      "epoch": 0.36563310512419095,
+      "grad_norm": 0.5142762660980225,
+      "learning_rate": 1.4096959058045867e-05,
+      "loss": 0.3081,
+      "step": 9780
+    },
+    {
+      "epoch": 0.3658200341145407,
+      "grad_norm": 0.46503591537475586,
+      "learning_rate": 1.4091601258971537e-05,
+      "loss": 0.2943,
+      "step": 9785
+    },
+    {
+      "epoch": 0.3660069631048905,
+      "grad_norm": 0.35828927159309387,
+      "learning_rate": 1.4086242048821172e-05,
+      "loss": 0.2945,
+      "step": 9790
+    },
+    {
+      "epoch": 0.36619389209524034,
+      "grad_norm": 0.43609440326690674,
+      "learning_rate": 1.408088142944301e-05,
+      "loss": 0.3362,
+      "step": 9795
+    },
+    {
+      "epoch": 0.3663808210855901,
+      "grad_norm": 0.7864788770675659,
+      "learning_rate": 1.4075519402685775e-05,
+      "loss": 0.2748,
+      "step": 9800
+    },
+    {
+      "epoch": 0.3665677500759399,
+      "grad_norm": 0.2505030930042267,
+      "learning_rate": 1.4070155970398676e-05,
+      "loss": 0.3193,
+      "step": 9805
+    },
+    {
+      "epoch": 0.36675467906628967,
+      "grad_norm": 0.2732783257961273,
+      "learning_rate": 1.406479113443141e-05,
+      "loss": 0.4463,
+      "step": 9810
+    },
+    {
+      "epoch": 0.3669416080566395,
+      "grad_norm": 0.4646552801132202,
+      "learning_rate": 1.405942489663415e-05,
+      "loss": 0.2432,
+      "step": 9815
+    },
+    {
+      "epoch": 0.3671285370469893,
+      "grad_norm": 0.23349085450172424,
+      "learning_rate": 1.405405725885756e-05,
+      "loss": 0.27,
+      "step": 9820
+    },
+    {
+      "epoch": 0.36731546603733906,
+      "grad_norm": 0.318560391664505,
+      "learning_rate": 1.4048688222952787e-05,
+      "loss": 0.2645,
+      "step": 9825
+    },
+    {
+      "epoch": 0.36750239502768883,
+      "grad_norm": 0.3994881510734558,
+      "learning_rate": 1.4043317790771455e-05,
+      "loss": 0.2549,
+      "step": 9830
+    },
+    {
+      "epoch": 0.36768932401803867,
+      "grad_norm": 0.4488486647605896,
+      "learning_rate": 1.4037945964165674e-05,
+      "loss": 0.2432,
+      "step": 9835
+    },
+    {
+      "epoch": 0.36787625300838844,
+      "grad_norm": 0.47077760100364685,
+      "learning_rate": 1.4032572744988029e-05,
+      "loss": 0.2743,
+      "step": 9840
+    },
+    {
+      "epoch": 0.3680631819987382,
+      "grad_norm": 0.3115878403186798,
+      "learning_rate": 1.4027198135091594e-05,
+      "loss": 0.3284,
+      "step": 9845
+    },
+    {
+      "epoch": 0.368250110989088,
+      "grad_norm": 0.802329421043396,
+      "learning_rate": 1.4021822136329914e-05,
+      "loss": 0.3841,
+      "step": 9850
+    },
+    {
+      "epoch": 0.36843703997943783,
+      "grad_norm": 0.22909800708293915,
+      "learning_rate": 1.4016444750557022e-05,
+      "loss": 0.3512,
+      "step": 9855
+    },
+    {
+      "epoch": 0.3686239689697876,
+      "grad_norm": 0.4044862687587738,
+      "learning_rate": 1.4011065979627418e-05,
+      "loss": 0.2226,
+      "step": 9860
+    },
+    {
+      "epoch": 0.3688108979601374,
+      "grad_norm": 0.7386347651481628,
+      "learning_rate": 1.4005685825396091e-05,
+      "loss": 0.2814,
+      "step": 9865
+    },
+    {
+      "epoch": 0.36899782695048716,
+      "grad_norm": 0.49540725350379944,
+      "learning_rate": 1.4000304289718498e-05,
+      "loss": 0.3408,
+      "step": 9870
+    },
+    {
+      "epoch": 0.369184755940837,
+      "grad_norm": 0.33938613533973694,
+      "learning_rate": 1.3994921374450584e-05,
+      "loss": 0.3386,
+      "step": 9875
+    },
+    {
+      "epoch": 0.3693716849311868,
+      "grad_norm": 0.6717550158500671,
+      "learning_rate": 1.3989537081448752e-05,
+      "loss": 0.351,
+      "step": 9880
+    },
+    {
+      "epoch": 0.36955861392153655,
+      "grad_norm": 0.3765832483768463,
+      "learning_rate": 1.3984151412569898e-05,
+      "loss": 0.2666,
+      "step": 9885
+    },
+    {
+      "epoch": 0.3697455429118863,
+      "grad_norm": 0.38686302304267883,
+      "learning_rate": 1.3978764369671378e-05,
+      "loss": 0.2667,
+      "step": 9890
+    },
+    {
+      "epoch": 0.36993247190223616,
+      "grad_norm": 0.4677604138851166,
+      "learning_rate": 1.3973375954611037e-05,
+      "loss": 0.263,
+      "step": 9895
+    },
+    {
+      "epoch": 0.37011940089258594,
+      "grad_norm": 0.19471901655197144,
+      "learning_rate": 1.396798616924718e-05,
+      "loss": 0.4968,
+      "step": 9900
+    },
+    {
+      "epoch": 0.3703063298829357,
+      "grad_norm": 0.37783077359199524,
+      "learning_rate": 1.3962595015438592e-05,
+      "loss": 0.3632,
+      "step": 9905
+    },
+    {
+      "epoch": 0.3704932588732855,
+      "grad_norm": 0.26753556728363037,
+      "learning_rate": 1.3957202495044525e-05,
+      "loss": 0.2798,
+      "step": 9910
+    },
+    {
+      "epoch": 0.3706801878636353,
+      "grad_norm": 0.3343743085861206,
+      "learning_rate": 1.3951808609924703e-05,
+      "loss": 0.2431,
+      "step": 9915
+    },
+    {
+      "epoch": 0.3708671168539851,
+      "grad_norm": 0.2634985148906708,
+      "learning_rate": 1.3946413361939324e-05,
+      "loss": 0.2888,
+      "step": 9920
+    },
+    {
+      "epoch": 0.3710540458443349,
+      "grad_norm": 0.8201472163200378,
+      "learning_rate": 1.3941016752949059e-05,
+      "loss": 0.3404,
+      "step": 9925
+    },
+    {
+      "epoch": 0.37124097483468466,
+      "grad_norm": 0.5062202215194702,
+      "learning_rate": 1.3935618784815042e-05,
+      "loss": 0.2689,
+      "step": 9930
+    },
+    {
+      "epoch": 0.3714279038250345,
+      "grad_norm": 0.5348239541053772,
+      "learning_rate": 1.3930219459398872e-05,
+      "loss": 0.2577,
+      "step": 9935
+    },
+    {
+      "epoch": 0.37161483281538427,
+      "grad_norm": 0.3352089822292328,
+      "learning_rate": 1.3924818778562627e-05,
+      "loss": 0.3134,
+      "step": 9940
+    },
+    {
+      "epoch": 0.37180176180573404,
+      "grad_norm": 0.4589434266090393,
+      "learning_rate": 1.3919416744168846e-05,
+      "loss": 0.3105,
+      "step": 9945
+    },
+    {
+      "epoch": 0.3719886907960838,
+      "grad_norm": 0.5832353234291077,
+      "learning_rate": 1.3914013358080536e-05,
+      "loss": 0.2674,
+      "step": 9950
+    },
+    {
+      "epoch": 0.37217561978643365,
+      "grad_norm": 0.46412187814712524,
+      "learning_rate": 1.3908608622161168e-05,
+      "loss": 0.2989,
+      "step": 9955
+    },
+    {
+      "epoch": 0.37236254877678343,
+      "grad_norm": 0.22348807752132416,
+      "learning_rate": 1.390320253827468e-05,
+      "loss": 0.306,
+      "step": 9960
+    },
+    {
+      "epoch": 0.3725494777671332,
+      "grad_norm": 0.3156212270259857,
+      "learning_rate": 1.389779510828548e-05,
+      "loss": 0.2838,
+      "step": 9965
+    },
+    {
+      "epoch": 0.372736406757483,
+      "grad_norm": 0.32380181550979614,
+      "learning_rate": 1.3892386334058433e-05,
+      "loss": 0.2335,
+      "step": 9970
+    },
+    {
+      "epoch": 0.3729233357478328,
+      "grad_norm": 0.39215198159217834,
+      "learning_rate": 1.3886976217458872e-05,
+      "loss": 0.2667,
+      "step": 9975
+    },
+    {
+      "epoch": 0.3731102647381826,
+      "grad_norm": 0.6366074085235596,
+      "learning_rate": 1.3881564760352589e-05,
+      "loss": 0.2668,
+      "step": 9980
+    },
+    {
+      "epoch": 0.3732971937285324,
+      "grad_norm": 0.4511488676071167,
+      "learning_rate": 1.387615196460584e-05,
+      "loss": 0.2636,
+      "step": 9985
+    },
+    {
+      "epoch": 0.37348412271888215,
+      "grad_norm": 0.2915279269218445,
+      "learning_rate": 1.3870737832085344e-05,
+      "loss": 0.2942,
+      "step": 9990
+    },
+    {
+      "epoch": 0.373671051709232,
+      "grad_norm": 0.2640349864959717,
+      "learning_rate": 1.386532236465828e-05,
+      "loss": 0.2608,
+      "step": 9995
+    },
+    {
+      "epoch": 0.37385798069958176,
+      "grad_norm": 0.5220035910606384,
+      "learning_rate": 1.3859905564192291e-05,
+      "loss": 0.3061,
+      "step": 10000
+    },
+    {
+      "epoch": 0.37404490968993154,
+      "grad_norm": 0.26642879843711853,
+      "learning_rate": 1.3854487432555467e-05,
+      "loss": 0.2558,
+      "step": 10005
+    },
+    {
+      "epoch": 0.3742318386802813,
+      "grad_norm": 0.5303171277046204,
+      "learning_rate": 1.3849067971616377e-05,
+      "loss": 0.264,
+      "step": 10010
+    },
+    {
+      "epoch": 0.3744187676706311,
+      "grad_norm": 0.412943035364151,
+      "learning_rate": 1.384364718324403e-05,
+      "loss": 0.363,
+      "step": 10015
+    },
+    {
+      "epoch": 0.3746056966609809,
+      "grad_norm": 0.7517222166061401,
+      "learning_rate": 1.3838225069307911e-05,
+      "loss": 0.3029,
+      "step": 10020
+    },
+    {
+      "epoch": 0.3747926256513307,
+      "grad_norm": 0.34779641032218933,
+      "learning_rate": 1.383280163167794e-05,
+      "loss": 0.2702,
+      "step": 10025
+    },
+    {
+      "epoch": 0.3749795546416805,
+      "grad_norm": 0.5778641104698181,
+      "learning_rate": 1.382737687222451e-05,
+      "loss": 0.2748,
+      "step": 10030
+    },
+    {
+      "epoch": 0.37516648363203026,
+      "grad_norm": 0.43767881393432617,
+      "learning_rate": 1.3821950792818464e-05,
+      "loss": 0.3038,
+      "step": 10035
+    },
+    {
+      "epoch": 0.3753534126223801,
+      "grad_norm": 0.43027085065841675,
+      "learning_rate": 1.3816523395331104e-05,
+      "loss": 0.2593,
+      "step": 10040
+    },
+    {
+      "epoch": 0.37554034161272987,
+      "grad_norm": 0.40595659613609314,
+      "learning_rate": 1.3811094681634183e-05,
+      "loss": 0.3294,
+      "step": 10045
+    },
+    {
+      "epoch": 0.37572727060307964,
+      "grad_norm": 0.21250957250595093,
+      "learning_rate": 1.3805664653599905e-05,
+      "loss": 0.3234,
+      "step": 10050
+    },
+    {
+      "epoch": 0.3759141995934294,
+      "grad_norm": 0.3901183605194092,
+      "learning_rate": 1.3800233313100935e-05,
+      "loss": 0.3291,
+      "step": 10055
+    },
+    {
+      "epoch": 0.37610112858377925,
+      "grad_norm": 0.40727564692497253,
+      "learning_rate": 1.3794800662010386e-05,
+      "loss": 0.2554,
+      "step": 10060
+    },
+    {
+      "epoch": 0.37628805757412903,
+      "grad_norm": 0.6711639165878296,
+      "learning_rate": 1.3789366702201824e-05,
+      "loss": 0.2909,
+      "step": 10065
+    },
+    {
+      "epoch": 0.3764749865644788,
+      "grad_norm": 0.3564465045928955,
+      "learning_rate": 1.3783931435549266e-05,
+      "loss": 0.2833,
+      "step": 10070
+    },
+    {
+      "epoch": 0.3766619155548286,
+      "grad_norm": 0.3789138197898865,
+      "learning_rate": 1.377849486392718e-05,
+      "loss": 0.2947,
+      "step": 10075
+    },
+    {
+      "epoch": 0.3768488445451784,
+      "grad_norm": 0.19771960377693176,
+      "learning_rate": 1.377305698921048e-05,
+      "loss": 0.2381,
+      "step": 10080
+    },
+    {
+      "epoch": 0.3770357735355282,
+      "grad_norm": 0.12256388366222382,
+      "learning_rate": 1.3767617813274537e-05,
+      "loss": 0.2924,
+      "step": 10085
+    },
+    {
+      "epoch": 0.377222702525878,
+      "grad_norm": 0.656947672367096,
+      "learning_rate": 1.376217733799517e-05,
+      "loss": 0.326,
+      "step": 10090
+    },
+    {
+      "epoch": 0.37740963151622775,
+      "grad_norm": 0.2615041136741638,
+      "learning_rate": 1.3756735565248634e-05,
+      "loss": 0.2551,
+      "step": 10095
+    },
+    {
+      "epoch": 0.3775965605065776,
+      "grad_norm": 0.4703410267829895,
+      "learning_rate": 1.3751292496911646e-05,
+      "loss": 0.273,
+      "step": 10100
+    },
+    {
+      "epoch": 0.37778348949692736,
+      "grad_norm": 0.3060661256313324,
+      "learning_rate": 1.3745848134861367e-05,
+      "loss": 0.381,
+      "step": 10105
+    },
+    {
+      "epoch": 0.37797041848727714,
+      "grad_norm": 0.30581894516944885,
+      "learning_rate": 1.3740402480975394e-05,
+      "loss": 0.2714,
+      "step": 10110
+    },
+    {
+      "epoch": 0.3781573474776269,
+      "grad_norm": 0.2676241397857666,
+      "learning_rate": 1.3734955537131786e-05,
+      "loss": 0.3834,
+      "step": 10115
+    },
+    {
+      "epoch": 0.37834427646797675,
+      "grad_norm": 0.28147178888320923,
+      "learning_rate": 1.3729507305209033e-05,
+      "loss": 0.2662,
+      "step": 10120
+    },
+    {
+      "epoch": 0.3785312054583265,
+      "grad_norm": 0.34182241559028625,
+      "learning_rate": 1.3724057787086073e-05,
+      "loss": 0.2477,
+      "step": 10125
+    },
+    {
+      "epoch": 0.3787181344486763,
+      "grad_norm": 0.30519115924835205,
+      "learning_rate": 1.3718606984642292e-05,
+      "loss": 0.2354,
+      "step": 10130
+    },
+    {
+      "epoch": 0.3789050634390261,
+      "grad_norm": 0.2952529788017273,
+      "learning_rate": 1.3713154899757508e-05,
+      "loss": 0.3169,
+      "step": 10135
+    },
+    {
+      "epoch": 0.3790919924293759,
+      "grad_norm": 0.4137369692325592,
+      "learning_rate": 1.3707701534311999e-05,
+      "loss": 0.3089,
+      "step": 10140
+    },
+    {
+      "epoch": 0.3792789214197257,
+      "grad_norm": 0.2354464828968048,
+      "learning_rate": 1.3702246890186469e-05,
+      "loss": 0.318,
+      "step": 10145
+    },
+    {
+      "epoch": 0.37946585041007547,
+      "grad_norm": 0.15742668509483337,
+      "learning_rate": 1.3696790969262068e-05,
+      "loss": 0.2672,
+      "step": 10150
+    },
+    {
+      "epoch": 0.37965277940042524,
+      "grad_norm": 0.21879169344902039,
+      "learning_rate": 1.369133377342039e-05,
+      "loss": 0.2698,
+      "step": 10155
+    },
+    {
+      "epoch": 0.3798397083907751,
+      "grad_norm": 0.8999713063240051,
+      "learning_rate": 1.3685875304543463e-05,
+      "loss": 0.4024,
+      "step": 10160
+    },
+    {
+      "epoch": 0.38002663738112485,
+      "grad_norm": 0.2633039355278015,
+      "learning_rate": 1.3680415564513759e-05,
+      "loss": 0.347,
+      "step": 10165
+    },
+    {
+      "epoch": 0.38021356637147463,
+      "grad_norm": 0.343563437461853,
+      "learning_rate": 1.3674954555214182e-05,
+      "loss": 0.3596,
+      "step": 10170
+    },
+    {
+      "epoch": 0.3804004953618244,
+      "grad_norm": 0.5179651975631714,
+      "learning_rate": 1.3669492278528079e-05,
+      "loss": 0.2785,
+      "step": 10175
+    },
+    {
+      "epoch": 0.38058742435217424,
+      "grad_norm": 0.24854378402233124,
+      "learning_rate": 1.3664028736339234e-05,
+      "loss": 0.2958,
+      "step": 10180
+    },
+    {
+      "epoch": 0.380774353342524,
+      "grad_norm": 0.39850738644599915,
+      "learning_rate": 1.3658563930531865e-05,
+      "loss": 0.313,
+      "step": 10185
+    },
+    {
+      "epoch": 0.3809612823328738,
+      "grad_norm": 0.3405385911464691,
+      "learning_rate": 1.365309786299063e-05,
+      "loss": 0.2671,
+      "step": 10190
+    },
+    {
+      "epoch": 0.3811482113232236,
+      "grad_norm": 0.3381601572036743,
+      "learning_rate": 1.3647630535600613e-05,
+      "loss": 0.2986,
+      "step": 10195
+    },
+    {
+      "epoch": 0.3813351403135734,
+      "grad_norm": 0.3975812792778015,
+      "learning_rate": 1.3642161950247345e-05,
+      "loss": 0.3056,
+      "step": 10200
+    },
+    {
+      "epoch": 0.3815220693039232,
+      "grad_norm": 0.5125154852867126,
+      "learning_rate": 1.363669210881678e-05,
+      "loss": 0.2916,
+      "step": 10205
+    },
+    {
+      "epoch": 0.38170899829427296,
+      "grad_norm": 0.29686829447746277,
+      "learning_rate": 1.3631221013195313e-05,
+      "loss": 0.3361,
+      "step": 10210
+    },
+    {
+      "epoch": 0.38189592728462274,
+      "grad_norm": 0.596336841583252,
+      "learning_rate": 1.3625748665269765e-05,
+      "loss": 0.2738,
+      "step": 10215
+    },
+    {
+      "epoch": 0.38208285627497257,
+      "grad_norm": 0.24768860638141632,
+      "learning_rate": 1.3620275066927392e-05,
+      "loss": 0.3018,
+      "step": 10220
+    },
+    {
+      "epoch": 0.38226978526532235,
+      "grad_norm": 0.30476251244544983,
+      "learning_rate": 1.3614800220055884e-05,
+      "loss": 0.2602,
+      "step": 10225
+    },
+    {
+      "epoch": 0.3824567142556721,
+      "grad_norm": 0.46634262800216675,
+      "learning_rate": 1.3609324126543357e-05,
+      "loss": 0.2068,
+      "step": 10230
+    },
+    {
+      "epoch": 0.3826436432460219,
+      "grad_norm": 0.264059454202652,
+      "learning_rate": 1.3603846788278357e-05,
+      "loss": 0.3386,
+      "step": 10235
+    },
+    {
+      "epoch": 0.38283057223637174,
+      "grad_norm": 0.37096327543258667,
+      "learning_rate": 1.3598368207149865e-05,
+      "loss": 0.3036,
+      "step": 10240
+    },
+    {
+      "epoch": 0.3830175012267215,
+      "grad_norm": 0.35662344098091125,
+      "learning_rate": 1.3592888385047285e-05,
+      "loss": 0.3593,
+      "step": 10245
+    },
+    {
+      "epoch": 0.3832044302170713,
+      "grad_norm": 0.4874442517757416,
+      "learning_rate": 1.358740732386045e-05,
+      "loss": 0.2687,
+      "step": 10250
+    },
+    {
+      "epoch": 0.38339135920742107,
+      "grad_norm": 0.3303247392177582,
+      "learning_rate": 1.3581925025479618e-05,
+      "loss": 0.3349,
+      "step": 10255
+    },
+    {
+      "epoch": 0.38357828819777084,
+      "grad_norm": 0.31140390038490295,
+      "learning_rate": 1.3576441491795484e-05,
+      "loss": 0.2713,
+      "step": 10260
+    },
+    {
+      "epoch": 0.3837652171881207,
+      "grad_norm": 0.2794208824634552,
+      "learning_rate": 1.3570956724699156e-05,
+      "loss": 0.2632,
+      "step": 10265
+    },
+    {
+      "epoch": 0.38395214617847045,
+      "grad_norm": 0.5220376253128052,
+      "learning_rate": 1.3565470726082176e-05,
+      "loss": 0.3356,
+      "step": 10270
+    },
+    {
+      "epoch": 0.38413907516882023,
+      "grad_norm": 0.31608107686042786,
+      "learning_rate": 1.3559983497836504e-05,
+      "loss": 0.3079,
+      "step": 10275
+    },
+    {
+      "epoch": 0.38432600415917,
+      "grad_norm": 0.44200223684310913,
+      "learning_rate": 1.3554495041854532e-05,
+      "loss": 0.2978,
+      "step": 10280
+    },
+    {
+      "epoch": 0.38451293314951984,
+      "grad_norm": 0.2726639211177826,
+      "learning_rate": 1.3549005360029066e-05,
+      "loss": 0.2671,
+      "step": 10285
+    },
+    {
+      "epoch": 0.3846998621398696,
+      "grad_norm": 0.410034716129303,
+      "learning_rate": 1.3543514454253346e-05,
+      "loss": 0.2938,
+      "step": 10290
+    },
+    {
+      "epoch": 0.3848867911302194,
+      "grad_norm": 0.33769235014915466,
+      "learning_rate": 1.3538022326421023e-05,
+      "loss": 0.3121,
+      "step": 10295
+    },
+    {
+      "epoch": 0.3850737201205692,
+      "grad_norm": 0.4343665540218353,
+      "learning_rate": 1.3532528978426178e-05,
+      "loss": 0.2907,
+      "step": 10300
+    },
+    {
+      "epoch": 0.385260649110919,
+      "grad_norm": 0.45513853430747986,
+      "learning_rate": 1.352703441216331e-05,
+      "loss": 0.2766,
+      "step": 10305
+    },
+    {
+      "epoch": 0.3854475781012688,
+      "grad_norm": 0.49546775221824646,
+      "learning_rate": 1.3521538629527336e-05,
+      "loss": 0.3089,
+      "step": 10310
+    },
+    {
+      "epoch": 0.38563450709161856,
+      "grad_norm": 0.6482167840003967,
+      "learning_rate": 1.351604163241359e-05,
+      "loss": 0.2859,
+      "step": 10315
+    },
+    {
+      "epoch": 0.38582143608196834,
+      "grad_norm": 0.6202002763748169,
+      "learning_rate": 1.3510543422717835e-05,
+      "loss": 0.3149,
+      "step": 10320
+    },
+    {
+      "epoch": 0.38600836507231817,
+      "grad_norm": 0.7666960954666138,
+      "learning_rate": 1.3505044002336241e-05,
+      "loss": 0.2812,
+      "step": 10325
+    },
+    {
+      "epoch": 0.38619529406266795,
+      "grad_norm": 0.42803874611854553,
+      "learning_rate": 1.3499543373165402e-05,
+      "loss": 0.3083,
+      "step": 10330
+    },
+    {
+      "epoch": 0.3863822230530177,
+      "grad_norm": 0.5874767899513245,
+      "learning_rate": 1.3494041537102332e-05,
+      "loss": 0.3362,
+      "step": 10335
+    },
+    {
+      "epoch": 0.3865691520433675,
+      "grad_norm": 0.5874844789505005,
+      "learning_rate": 1.348853849604445e-05,
+      "loss": 0.3628,
+      "step": 10340
+    },
+    {
+      "epoch": 0.38675608103371734,
+      "grad_norm": 0.454783171415329,
+      "learning_rate": 1.34830342518896e-05,
+      "loss": 0.2621,
+      "step": 10345
+    },
+    {
+      "epoch": 0.3869430100240671,
+      "grad_norm": 0.3771190345287323,
+      "learning_rate": 1.347752880653604e-05,
+      "loss": 0.2833,
+      "step": 10350
+    },
+    {
+      "epoch": 0.3871299390144169,
+      "grad_norm": 0.30891087651252747,
+      "learning_rate": 1.3472022161882439e-05,
+      "loss": 0.2938,
+      "step": 10355
+    },
+    {
+      "epoch": 0.38731686800476667,
+      "grad_norm": 0.5641288757324219,
+      "learning_rate": 1.346651431982788e-05,
+      "loss": 0.3008,
+      "step": 10360
+    },
+    {
+      "epoch": 0.3875037969951165,
+      "grad_norm": 0.19588641822338104,
+      "learning_rate": 1.3461005282271857e-05,
+      "loss": 0.2773,
+      "step": 10365
+    },
+    {
+      "epoch": 0.3876907259854663,
+      "grad_norm": 0.6388927102088928,
+      "learning_rate": 1.3455495051114283e-05,
+      "loss": 0.2966,
+      "step": 10370
+    },
+    {
+      "epoch": 0.38787765497581606,
+      "grad_norm": 0.4509037733078003,
+      "learning_rate": 1.344998362825548e-05,
+      "loss": 0.3202,
+      "step": 10375
+    },
+    {
+      "epoch": 0.38806458396616583,
+      "grad_norm": 0.3036980628967285,
+      "learning_rate": 1.344447101559618e-05,
+      "loss": 0.2795,
+      "step": 10380
+    },
+    {
+      "epoch": 0.38825151295651567,
+      "grad_norm": 0.34260863065719604,
+      "learning_rate": 1.343895721503752e-05,
+      "loss": 0.2886,
+      "step": 10385
+    },
+    {
+      "epoch": 0.38843844194686544,
+      "grad_norm": 0.2766103744506836,
+      "learning_rate": 1.3433442228481055e-05,
+      "loss": 0.3462,
+      "step": 10390
+    },
+    {
+      "epoch": 0.3886253709372152,
+      "grad_norm": 0.39682698249816895,
+      "learning_rate": 1.3427926057828749e-05,
+      "loss": 0.294,
+      "step": 10395
+    },
+    {
+      "epoch": 0.388812299927565,
+      "grad_norm": 0.3901638388633728,
+      "learning_rate": 1.3422408704982968e-05,
+      "loss": 0.2803,
+      "step": 10400
+    },
+    {
+      "epoch": 0.38899922891791483,
+      "grad_norm": 0.39301908016204834,
+      "learning_rate": 1.3416890171846486e-05,
+      "loss": 0.3174,
+      "step": 10405
+    },
+    {
+      "epoch": 0.3891861579082646,
+      "grad_norm": 0.39801543951034546,
+      "learning_rate": 1.3411370460322493e-05,
+      "loss": 0.3569,
+      "step": 10410
+    },
+    {
+      "epoch": 0.3893730868986144,
+      "grad_norm": 0.3659820854663849,
+      "learning_rate": 1.3405849572314574e-05,
+      "loss": 0.3338,
+      "step": 10415
+    },
+    {
+      "epoch": 0.38956001588896416,
+      "grad_norm": 0.28667154908180237,
+      "learning_rate": 1.3400327509726726e-05,
+      "loss": 0.2569,
+      "step": 10420
+    },
+    {
+      "epoch": 0.389746944879314,
+      "grad_norm": 0.4689178466796875,
+      "learning_rate": 1.3394804274463358e-05,
+      "loss": 0.279,
+      "step": 10425
+    },
+    {
+      "epoch": 0.38993387386966377,
+      "grad_norm": 0.4040241539478302,
+      "learning_rate": 1.3389279868429264e-05,
+      "loss": 0.2931,
+      "step": 10430
+    },
+    {
+      "epoch": 0.39012080286001355,
+      "grad_norm": 0.33305883407592773,
+      "learning_rate": 1.3383754293529659e-05,
+      "loss": 0.3057,
+      "step": 10435
+    },
+    {
+      "epoch": 0.3903077318503633,
+      "grad_norm": 0.37228256464004517,
+      "learning_rate": 1.3378227551670155e-05,
+      "loss": 0.315,
+      "step": 10440
+    },
+    {
+      "epoch": 0.39049466084071316,
+      "grad_norm": 0.4526333510875702,
+      "learning_rate": 1.3372699644756769e-05,
+      "loss": 0.294,
+      "step": 10445
+    },
+    {
+      "epoch": 0.39068158983106294,
+      "grad_norm": 0.4754495620727539,
+      "learning_rate": 1.3367170574695916e-05,
+      "loss": 0.2909,
+      "step": 10450
+    },
+    {
+      "epoch": 0.3908685188214127,
+      "grad_norm": 0.4171309769153595,
+      "learning_rate": 1.3361640343394411e-05,
+      "loss": 0.2908,
+      "step": 10455
+    },
+    {
+      "epoch": 0.3910554478117625,
+      "grad_norm": 0.3807438313961029,
+      "learning_rate": 1.3356108952759472e-05,
+      "loss": 0.3104,
+      "step": 10460
+    },
+    {
+      "epoch": 0.3912423768021123,
+      "grad_norm": 0.20601804554462433,
+      "learning_rate": 1.3350576404698725e-05,
+      "loss": 0.299,
+      "step": 10465
+    },
+    {
+      "epoch": 0.3914293057924621,
+      "grad_norm": 0.4232652187347412,
+      "learning_rate": 1.334504270112018e-05,
+      "loss": 0.3072,
+      "step": 10470
+    },
+    {
+      "epoch": 0.3916162347828119,
+      "grad_norm": 0.40767017006874084,
+      "learning_rate": 1.3339507843932259e-05,
+      "loss": 0.273,
+      "step": 10475
+    },
+    {
+      "epoch": 0.39180316377316166,
+      "grad_norm": 0.49084317684173584,
+      "learning_rate": 1.3333971835043767e-05,
+      "loss": 0.2653,
+      "step": 10480
+    },
+    {
+      "epoch": 0.3919900927635115,
+      "grad_norm": 0.5394593477249146,
+      "learning_rate": 1.332843467636392e-05,
+      "loss": 0.3149,
+      "step": 10485
+    },
+    {
+      "epoch": 0.39217702175386127,
+      "grad_norm": 0.2563055157661438,
+      "learning_rate": 1.3322896369802325e-05,
+      "loss": 0.3443,
+      "step": 10490
+    },
+    {
+      "epoch": 0.39236395074421104,
+      "grad_norm": 0.2230866700410843,
+      "learning_rate": 1.3317356917268987e-05,
+      "loss": 0.2512,
+      "step": 10495
+    },
+    {
+      "epoch": 0.3925508797345608,
+      "grad_norm": 0.5398097038269043,
+      "learning_rate": 1.33118163206743e-05,
+      "loss": 0.2493,
+      "step": 10500
+    },
+    {
+      "epoch": 0.3927378087249106,
+      "grad_norm": 0.36351341009140015,
+      "learning_rate": 1.3306274581929059e-05,
+      "loss": 0.3472,
+      "step": 10505
+    },
+    {
+      "epoch": 0.39292473771526043,
+      "grad_norm": 0.34674593806266785,
+      "learning_rate": 1.3300731702944453e-05,
+      "loss": 0.3117,
+      "step": 10510
+    },
+    {
+      "epoch": 0.3931116667056102,
+      "grad_norm": 0.2768007516860962,
+      "learning_rate": 1.329518768563206e-05,
+      "loss": 0.2991,
+      "step": 10515
+    },
+    {
+      "epoch": 0.39329859569596,
+      "grad_norm": 0.7572014331817627,
+      "learning_rate": 1.3289642531903857e-05,
+      "loss": 0.3921,
+      "step": 10520
+    },
+    {
+      "epoch": 0.39348552468630976,
+      "grad_norm": 0.34211087226867676,
+      "learning_rate": 1.32840962436722e-05,
+      "loss": 0.3197,
+      "step": 10525
+    },
+    {
+      "epoch": 0.3936724536766596,
+      "grad_norm": 0.5510493516921997,
+      "learning_rate": 1.3278548822849853e-05,
+      "loss": 0.2646,
+      "step": 10530
+    },
+    {
+      "epoch": 0.3938593826670094,
+      "grad_norm": 0.39801833033561707,
+      "learning_rate": 1.3273000271349959e-05,
+      "loss": 0.2889,
+      "step": 10535
+    },
+    {
+      "epoch": 0.39404631165735915,
+      "grad_norm": 0.3157884180545807,
+      "learning_rate": 1.3267450591086051e-05,
+      "loss": 0.3597,
+      "step": 10540
+    },
+    {
+      "epoch": 0.3942332406477089,
+      "grad_norm": 0.5300944447517395,
+      "learning_rate": 1.3261899783972062e-05,
+      "loss": 0.3158,
+      "step": 10545
+    },
+    {
+      "epoch": 0.39442016963805876,
+      "grad_norm": 0.48455068469047546,
+      "learning_rate": 1.32563478519223e-05,
+      "loss": 0.2964,
+      "step": 10550
+    },
+    {
+      "epoch": 0.39460709862840854,
+      "grad_norm": 0.33385178446769714,
+      "learning_rate": 1.3250794796851474e-05,
+      "loss": 0.3481,
+      "step": 10555
+    },
+    {
+      "epoch": 0.3947940276187583,
+      "grad_norm": 0.35784122347831726,
+      "learning_rate": 1.3245240620674667e-05,
+      "loss": 0.3654,
+      "step": 10560
+    },
+    {
+      "epoch": 0.3949809566091081,
+      "grad_norm": 0.19756436347961426,
+      "learning_rate": 1.3239685325307359e-05,
+      "loss": 0.3123,
+      "step": 10565
+    },
+    {
+      "epoch": 0.3951678855994579,
+      "grad_norm": 0.33110281825065613,
+      "learning_rate": 1.323412891266541e-05,
+      "loss": 0.291,
+      "step": 10570
+    },
+    {
+      "epoch": 0.3953548145898077,
+      "grad_norm": 0.2022867351770401,
+      "learning_rate": 1.322857138466507e-05,
+      "loss": 0.3221,
+      "step": 10575
+    },
+    {
+      "epoch": 0.3955417435801575,
+      "grad_norm": 0.49643459916114807,
+      "learning_rate": 1.3223012743222967e-05,
+      "loss": 0.2867,
+      "step": 10580
+    },
+    {
+      "epoch": 0.39572867257050726,
+      "grad_norm": 0.43748632073402405,
+      "learning_rate": 1.3217452990256122e-05,
+      "loss": 0.334,
+      "step": 10585
+    },
+    {
+      "epoch": 0.3959156015608571,
+      "grad_norm": 0.6567883491516113,
+      "learning_rate": 1.3211892127681934e-05,
+      "loss": 0.3096,
+      "step": 10590
+    },
+    {
+      "epoch": 0.39610253055120687,
+      "grad_norm": 0.5701555609703064,
+      "learning_rate": 1.320633015741818e-05,
+      "loss": 0.249,
+      "step": 10595
+    },
+    {
+      "epoch": 0.39628945954155664,
+      "grad_norm": 0.33283892273902893,
+      "learning_rate": 1.3200767081383028e-05,
+      "loss": 0.3294,
+      "step": 10600
+    },
+    {
+      "epoch": 0.3964763885319064,
+      "grad_norm": 0.3843041956424713,
+      "learning_rate": 1.3195202901495024e-05,
+      "loss": 0.2428,
+      "step": 10605
+    },
+    {
+      "epoch": 0.39666331752225625,
+      "grad_norm": 0.2713102698326111,
+      "learning_rate": 1.3189637619673095e-05,
+      "loss": 0.2753,
+      "step": 10610
+    },
+    {
+      "epoch": 0.39685024651260603,
+      "grad_norm": 0.28656595945358276,
+      "learning_rate": 1.3184071237836544e-05,
+      "loss": 0.2965,
+      "step": 10615
+    },
+    {
+      "epoch": 0.3970371755029558,
+      "grad_norm": 0.30529990792274475,
+      "learning_rate": 1.3178503757905058e-05,
+      "loss": 0.3222,
+      "step": 10620
+    },
+    {
+      "epoch": 0.3972241044933056,
+      "grad_norm": 0.3871190845966339,
+      "learning_rate": 1.3172935181798703e-05,
+      "loss": 0.362,
+      "step": 10625
+    },
+    {
+      "epoch": 0.3974110334836554,
+      "grad_norm": 0.5080790519714355,
+      "learning_rate": 1.3167365511437919e-05,
+      "loss": 0.238,
+      "step": 10630
+    },
+    {
+      "epoch": 0.3975979624740052,
+      "grad_norm": 0.3145684003829956,
+      "learning_rate": 1.3161794748743525e-05,
+      "loss": 0.2148,
+      "step": 10635
+    },
+    {
+      "epoch": 0.397784891464355,
+      "grad_norm": 0.7068213820457458,
+      "learning_rate": 1.315622289563672e-05,
+      "loss": 0.2557,
+      "step": 10640
+    },
+    {
+      "epoch": 0.39797182045470475,
+      "grad_norm": 0.3006476163864136,
+      "learning_rate": 1.3150649954039078e-05,
+      "loss": 0.3824,
+      "step": 10645
+    },
+    {
+      "epoch": 0.3981587494450546,
+      "grad_norm": 0.269481360912323,
+      "learning_rate": 1.3145075925872543e-05,
+      "loss": 0.3222,
+      "step": 10650
+    },
+    {
+      "epoch": 0.39834567843540436,
+      "grad_norm": 0.42372992634773254,
+      "learning_rate": 1.3139500813059438e-05,
+      "loss": 0.2562,
+      "step": 10655
+    },
+    {
+      "epoch": 0.39853260742575414,
+      "grad_norm": 0.4050086736679077,
+      "learning_rate": 1.3133924617522462e-05,
+      "loss": 0.2818,
+      "step": 10660
+    },
+    {
+      "epoch": 0.3987195364161039,
+      "grad_norm": 0.34156128764152527,
+      "learning_rate": 1.3128347341184684e-05,
+      "loss": 0.3041,
+      "step": 10665
+    },
+    {
+      "epoch": 0.39890646540645375,
+      "grad_norm": 0.25688794255256653,
+      "learning_rate": 1.3122768985969546e-05,
+      "loss": 0.3239,
+      "step": 10670
+    },
+    {
+      "epoch": 0.3990933943968035,
+      "grad_norm": 0.4633738398551941,
+      "learning_rate": 1.3117189553800861e-05,
+      "loss": 0.267,
+      "step": 10675
+    },
+    {
+      "epoch": 0.3992803233871533,
+      "grad_norm": 0.41393613815307617,
+      "learning_rate": 1.311160904660282e-05,
+      "loss": 0.2762,
+      "step": 10680
+    },
+    {
+      "epoch": 0.3994672523775031,
+      "grad_norm": 0.41371089220046997,
+      "learning_rate": 1.3106027466299977e-05,
+      "loss": 0.2862,
+      "step": 10685
+    },
+    {
+      "epoch": 0.3996541813678529,
+      "grad_norm": 0.4804210662841797,
+      "learning_rate": 1.310044481481726e-05,
+      "loss": 0.2986,
+      "step": 10690
+    },
+    {
+      "epoch": 0.3998411103582027,
+      "grad_norm": 0.5706376433372498,
+      "learning_rate": 1.3094861094079965e-05,
+      "loss": 0.3143,
+      "step": 10695
+    },
+    {
+      "epoch": 0.40002803934855247,
+      "grad_norm": 0.6555120944976807,
+      "learning_rate": 1.3089276306013759e-05,
+      "loss": 0.2877,
+      "step": 10700
+    },
+    {
+      "epoch": 0.40021496833890224,
+      "grad_norm": 0.40146076679229736,
+      "learning_rate": 1.3083690452544673e-05,
+      "loss": 0.2458,
+      "step": 10705
+    },
+    {
+      "epoch": 0.4004018973292521,
+      "grad_norm": 0.24645639955997467,
+      "learning_rate": 1.307810353559911e-05,
+      "loss": 0.3734,
+      "step": 10710
+    },
+    {
+      "epoch": 0.40058882631960185,
+      "grad_norm": 0.34604454040527344,
+      "learning_rate": 1.3072515557103835e-05,
+      "loss": 0.3489,
+      "step": 10715
+    },
+    {
+      "epoch": 0.40077575530995163,
+      "grad_norm": 0.3194914758205414,
+      "learning_rate": 1.3066926518985984e-05,
+      "loss": 0.2924,
+      "step": 10720
+    },
+    {
+      "epoch": 0.4009626843003014,
+      "grad_norm": 0.38813045620918274,
+      "learning_rate": 1.3061336423173053e-05,
+      "loss": 0.3895,
+      "step": 10725
+    },
+    {
+      "epoch": 0.40114961329065124,
+      "grad_norm": 0.4124408960342407,
+      "learning_rate": 1.305574527159291e-05,
+      "loss": 0.2377,
+      "step": 10730
+    },
+    {
+      "epoch": 0.401336542281001,
+      "grad_norm": 0.37710294127464294,
+      "learning_rate": 1.305015306617378e-05,
+      "loss": 0.2393,
+      "step": 10735
+    },
+    {
+      "epoch": 0.4015234712713508,
+      "grad_norm": 0.34080588817596436,
+      "learning_rate": 1.3044559808844257e-05,
+      "loss": 0.3022,
+      "step": 10740
+    },
+    {
+      "epoch": 0.4017104002617006,
+      "grad_norm": 0.4511391520500183,
+      "learning_rate": 1.3038965501533291e-05,
+      "loss": 0.2905,
+      "step": 10745
+    },
+    {
+      "epoch": 0.40189732925205035,
+      "grad_norm": 0.5546760559082031,
+      "learning_rate": 1.3033370146170201e-05,
+      "loss": 0.3291,
+      "step": 10750
+    },
+    {
+      "epoch": 0.4020842582424002,
+      "grad_norm": 0.3765566647052765,
+      "learning_rate": 1.3027773744684669e-05,
+      "loss": 0.2396,
+      "step": 10755
+    },
+    {
+      "epoch": 0.40227118723274996,
+      "grad_norm": 0.32864052057266235,
+      "learning_rate": 1.3022176299006726e-05,
+      "loss": 0.3394,
+      "step": 10760
+    },
+    {
+      "epoch": 0.40245811622309974,
+      "grad_norm": 0.3548523187637329,
+      "learning_rate": 1.3016577811066775e-05,
+      "loss": 0.3865,
+      "step": 10765
+    },
+    {
+      "epoch": 0.4026450452134495,
+      "grad_norm": 0.8239120841026306,
+      "learning_rate": 1.3010978282795569e-05,
+      "loss": 0.2682,
+      "step": 10770
+    },
+    {
+      "epoch": 0.40283197420379935,
+      "grad_norm": 0.3894842565059662,
+      "learning_rate": 1.3005377716124232e-05,
+      "loss": 0.335,
+      "step": 10775
+    },
+    {
+      "epoch": 0.4030189031941491,
+      "grad_norm": 0.5358176231384277,
+      "learning_rate": 1.2999776112984237e-05,
+      "loss": 0.2949,
+      "step": 10780
+    },
+    {
+      "epoch": 0.4032058321844989,
+      "grad_norm": 0.5368191003799438,
+      "learning_rate": 1.2994173475307412e-05,
+      "loss": 0.2752,
+      "step": 10785
+    },
+    {
+      "epoch": 0.4033927611748487,
+      "grad_norm": 0.30042341351509094,
+      "learning_rate": 1.298856980502595e-05,
+      "loss": 0.2578,
+      "step": 10790
+    },
+    {
+      "epoch": 0.4035796901651985,
+      "grad_norm": 0.3132897913455963,
+      "learning_rate": 1.2982965104072397e-05,
+      "loss": 0.3735,
+      "step": 10795
+    },
+    {
+      "epoch": 0.4037666191555483,
+      "grad_norm": 2.302870988845825,
+      "learning_rate": 1.2977359374379652e-05,
+      "loss": 0.3051,
+      "step": 10800
+    },
+    {
+      "epoch": 0.40395354814589807,
+      "grad_norm": 0.31478455662727356,
+      "learning_rate": 1.2971752617880972e-05,
+      "loss": 0.2558,
+      "step": 10805
+    },
+    {
+      "epoch": 0.40414047713624784,
+      "grad_norm": 0.5957106947898865,
+      "learning_rate": 1.2966144836509964e-05,
+      "loss": 0.3231,
+      "step": 10810
+    },
+    {
+      "epoch": 0.4043274061265977,
+      "grad_norm": 0.49217236042022705,
+      "learning_rate": 1.2960536032200592e-05,
+      "loss": 0.2385,
+      "step": 10815
+    },
+    {
+      "epoch": 0.40451433511694745,
+      "grad_norm": 0.39011693000793457,
+      "learning_rate": 1.2954926206887174e-05,
+      "loss": 0.3561,
+      "step": 10820
+    },
+    {
+      "epoch": 0.40470126410729723,
+      "grad_norm": 0.4122646152973175,
+      "learning_rate": 1.2949315362504376e-05,
+      "loss": 0.3262,
+      "step": 10825
+    },
+    {
+      "epoch": 0.404888193097647,
+      "grad_norm": 0.3655274510383606,
+      "learning_rate": 1.2943703500987218e-05,
+      "loss": 0.2638,
+      "step": 10830
+    },
+    {
+      "epoch": 0.40507512208799684,
+      "grad_norm": 0.39280015230178833,
+      "learning_rate": 1.293809062427107e-05,
+      "loss": 0.3355,
+      "step": 10835
+    },
+    {
+      "epoch": 0.4052620510783466,
+      "grad_norm": 0.5871797800064087,
+      "learning_rate": 1.2932476734291652e-05,
+      "loss": 0.3217,
+      "step": 10840
+    },
+    {
+      "epoch": 0.4054489800686964,
+      "grad_norm": 0.22152677178382874,
+      "learning_rate": 1.2926861832985036e-05,
+      "loss": 0.2709,
+      "step": 10845
+    },
+    {
+      "epoch": 0.4056359090590462,
+      "grad_norm": 0.4744000732898712,
+      "learning_rate": 1.292124592228764e-05,
+      "loss": 0.3501,
+      "step": 10850
+    },
+    {
+      "epoch": 0.405822838049396,
+      "grad_norm": 0.34133681654930115,
+      "learning_rate": 1.2915629004136228e-05,
+      "loss": 0.2422,
+      "step": 10855
+    },
+    {
+      "epoch": 0.4060097670397458,
+      "grad_norm": 0.5644049048423767,
+      "learning_rate": 1.2910011080467917e-05,
+      "loss": 0.2745,
+      "step": 10860
+    },
+    {
+      "epoch": 0.40619669603009556,
+      "grad_norm": 0.48984065651893616,
+      "learning_rate": 1.2904392153220164e-05,
+      "loss": 0.3229,
+      "step": 10865
+    },
+    {
+      "epoch": 0.40638362502044534,
+      "grad_norm": 0.30956414341926575,
+      "learning_rate": 1.2898772224330778e-05,
+      "loss": 0.3053,
+      "step": 10870
+    },
+    {
+      "epoch": 0.40657055401079517,
+      "grad_norm": 0.24445974826812744,
+      "learning_rate": 1.2893151295737916e-05,
+      "loss": 0.2011,
+      "step": 10875
+    },
+    {
+      "epoch": 0.40675748300114495,
+      "grad_norm": 1.567341923713684,
+      "learning_rate": 1.288752936938007e-05,
+      "loss": 0.3163,
+      "step": 10880
+    },
+    {
+      "epoch": 0.4069444119914947,
+      "grad_norm": 0.28451815247535706,
+      "learning_rate": 1.2881906447196082e-05,
+      "loss": 0.2405,
+      "step": 10885
+    },
+    {
+      "epoch": 0.4071313409818445,
+      "grad_norm": 0.252210408449173,
+      "learning_rate": 1.287628253112514e-05,
+      "loss": 0.245,
+      "step": 10890
+    },
+    {
+      "epoch": 0.40731826997219434,
+      "grad_norm": 0.4983895421028137,
+      "learning_rate": 1.2870657623106766e-05,
+      "loss": 0.3025,
+      "step": 10895
+    },
+    {
+      "epoch": 0.4075051989625441,
+      "grad_norm": 0.6142060160636902,
+      "learning_rate": 1.2865031725080834e-05,
+      "loss": 0.3184,
+      "step": 10900
+    },
+    {
+      "epoch": 0.4076921279528939,
+      "grad_norm": 0.34046661853790283,
+      "learning_rate": 1.2859404838987552e-05,
+      "loss": 0.3181,
+      "step": 10905
+    },
+    {
+      "epoch": 0.40787905694324367,
+      "grad_norm": 0.7890579700469971,
+      "learning_rate": 1.2853776966767475e-05,
+      "loss": 0.3095,
+      "step": 10910
+    },
+    {
+      "epoch": 0.4080659859335935,
+      "grad_norm": 0.4729807376861572,
+      "learning_rate": 1.284814811036149e-05,
+      "loss": 0.3228,
+      "step": 10915
+    },
+    {
+      "epoch": 0.4082529149239433,
+      "grad_norm": 0.5021012425422668,
+      "learning_rate": 1.2842518271710836e-05,
+      "loss": 0.2234,
+      "step": 10920
+    },
+    {
+      "epoch": 0.40843984391429305,
+      "grad_norm": 0.36319997906684875,
+      "learning_rate": 1.2836887452757076e-05,
+      "loss": 0.2692,
+      "step": 10925
+    },
+    {
+      "epoch": 0.40862677290464283,
+      "grad_norm": 0.44691482186317444,
+      "learning_rate": 1.2831255655442122e-05,
+      "loss": 0.3243,
+      "step": 10930
+    },
+    {
+      "epoch": 0.40881370189499266,
+      "grad_norm": 0.2757837772369385,
+      "learning_rate": 1.2825622881708218e-05,
+      "loss": 0.3383,
+      "step": 10935
+    },
+    {
+      "epoch": 0.40900063088534244,
+      "grad_norm": 0.3655282258987427,
+      "learning_rate": 1.2819989133497945e-05,
+      "loss": 0.216,
+      "step": 10940
+    },
+    {
+      "epoch": 0.4091875598756922,
+      "grad_norm": 0.2834169864654541,
+      "learning_rate": 1.2814354412754227e-05,
+      "loss": 0.2807,
+      "step": 10945
+    },
+    {
+      "epoch": 0.409374488866042,
+      "grad_norm": 0.49131330847740173,
+      "learning_rate": 1.2808718721420308e-05,
+      "loss": 0.2887,
+      "step": 10950
+    },
+    {
+      "epoch": 0.40956141785639183,
+      "grad_norm": 0.2866436541080475,
+      "learning_rate": 1.2803082061439784e-05,
+      "loss": 0.3064,
+      "step": 10955
+    },
+    {
+      "epoch": 0.4097483468467416,
+      "grad_norm": 0.5651163458824158,
+      "learning_rate": 1.2797444434756571e-05,
+      "loss": 0.2649,
+      "step": 10960
+    },
+    {
+      "epoch": 0.4099352758370914,
+      "grad_norm": 0.3305072486400604,
+      "learning_rate": 1.2791805843314937e-05,
+      "loss": 0.3414,
+      "step": 10965
+    },
+    {
+      "epoch": 0.41012220482744116,
+      "grad_norm": 0.4713725745677948,
+      "learning_rate": 1.2786166289059461e-05,
+      "loss": 0.2735,
+      "step": 10970
+    },
+    {
+      "epoch": 0.410309133817791,
+      "grad_norm": 0.45371806621551514,
+      "learning_rate": 1.2780525773935063e-05,
+      "loss": 0.231,
+      "step": 10975
+    },
+    {
+      "epoch": 0.41049606280814077,
+      "grad_norm": 0.3463742136955261,
+      "learning_rate": 1.2774884299887e-05,
+      "loss": 0.338,
+      "step": 10980
+    },
+    {
+      "epoch": 0.41068299179849055,
+      "grad_norm": 0.30682888627052307,
+      "learning_rate": 1.2769241868860851e-05,
+      "loss": 0.2203,
+      "step": 10985
+    },
+    {
+      "epoch": 0.4108699207888403,
+      "grad_norm": 0.5777801275253296,
+      "learning_rate": 1.2763598482802531e-05,
+      "loss": 0.2488,
+      "step": 10990
+    },
+    {
+      "epoch": 0.4110568497791901,
+      "grad_norm": 0.3914337754249573,
+      "learning_rate": 1.2757954143658285e-05,
+      "loss": 0.3193,
+      "step": 10995
+    },
+    {
+      "epoch": 0.41124377876953994,
+      "grad_norm": 0.37003767490386963,
+      "learning_rate": 1.2752308853374675e-05,
+      "loss": 0.3423,
+      "step": 11000
+    },
+    {
+      "epoch": 0.4114307077598897,
+      "grad_norm": 0.6970301866531372,
+      "learning_rate": 1.2746662613898605e-05,
+      "loss": 0.3753,
+      "step": 11005
+    },
+    {
+      "epoch": 0.4116176367502395,
+      "grad_norm": 0.5969115495681763,
+      "learning_rate": 1.2741015427177304e-05,
+      "loss": 0.3479,
+      "step": 11010
+    },
+    {
+      "epoch": 0.41180456574058927,
+      "grad_norm": 0.4280428886413574,
+      "learning_rate": 1.2735367295158324e-05,
+      "loss": 0.2633,
+      "step": 11015
+    },
+    {
+      "epoch": 0.4119914947309391,
+      "grad_norm": 0.47678300738334656,
+      "learning_rate": 1.2729718219789538e-05,
+      "loss": 0.2922,
+      "step": 11020
+    },
+    {
+      "epoch": 0.4121784237212889,
+      "grad_norm": 0.4477633833885193,
+      "learning_rate": 1.2724068203019155e-05,
+      "loss": 0.3189,
+      "step": 11025
+    },
+    {
+      "epoch": 0.41236535271163866,
+      "grad_norm": 0.330090194940567,
+      "learning_rate": 1.2718417246795702e-05,
+      "loss": 0.2159,
+      "step": 11030
+    },
+    {
+      "epoch": 0.41255228170198843,
+      "grad_norm": 0.33988553285598755,
+      "learning_rate": 1.2712765353068036e-05,
+      "loss": 0.2479,
+      "step": 11035
+    },
+    {
+      "epoch": 0.41273921069233827,
+      "grad_norm": 0.22661490738391876,
+      "learning_rate": 1.2707112523785327e-05,
+      "loss": 0.3321,
+      "step": 11040
+    },
+    {
+      "epoch": 0.41292613968268804,
+      "grad_norm": 0.29933539032936096,
+      "learning_rate": 1.2701458760897083e-05,
+      "loss": 0.2927,
+      "step": 11045
+    },
+    {
+      "epoch": 0.4131130686730378,
+      "grad_norm": 0.39779335260391235,
+      "learning_rate": 1.2695804066353113e-05,
+      "loss": 0.3024,
+      "step": 11050
+    },
+    {
+      "epoch": 0.4132999976633876,
+      "grad_norm": 0.3741395175457001,
+      "learning_rate": 1.2690148442103567e-05,
+      "loss": 0.3496,
+      "step": 11055
+    },
+    {
+      "epoch": 0.41348692665373743,
+      "grad_norm": 0.33514922857284546,
+      "learning_rate": 1.2684491890098907e-05,
+      "loss": 0.2939,
+      "step": 11060
+    },
+    {
+      "epoch": 0.4136738556440872,
+      "grad_norm": 0.2723214030265808,
+      "learning_rate": 1.2678834412289915e-05,
+      "loss": 0.2748,
+      "step": 11065
+    },
+    {
+      "epoch": 0.413860784634437,
+      "grad_norm": 0.2613365054130554,
+      "learning_rate": 1.267317601062769e-05,
+      "loss": 0.268,
+      "step": 11070
+    },
+    {
+      "epoch": 0.41404771362478676,
+      "grad_norm": 0.3753436207771301,
+      "learning_rate": 1.2667516687063657e-05,
+      "loss": 0.2862,
+      "step": 11075
+    },
+    {
+      "epoch": 0.4142346426151366,
+      "grad_norm": 0.38865718245506287,
+      "learning_rate": 1.2661856443549551e-05,
+      "loss": 0.3599,
+      "step": 11080
+    },
+    {
+      "epoch": 0.41442157160548637,
+      "grad_norm": 0.3524521291255951,
+      "learning_rate": 1.2656195282037432e-05,
+      "loss": 0.2869,
+      "step": 11085
+    },
+    {
+      "epoch": 0.41460850059583615,
+      "grad_norm": 0.28500306606292725,
+      "learning_rate": 1.265053320447967e-05,
+      "loss": 0.2769,
+      "step": 11090
+    },
+    {
+      "epoch": 0.4147954295861859,
+      "grad_norm": 0.514741837978363,
+      "learning_rate": 1.2644870212828949e-05,
+      "loss": 0.234,
+      "step": 11095
+    },
+    {
+      "epoch": 0.41498235857653576,
+      "grad_norm": 0.33767464756965637,
+      "learning_rate": 1.2639206309038279e-05,
+      "loss": 0.3528,
+      "step": 11100
+    },
+    {
+      "epoch": 0.41516928756688554,
+      "grad_norm": 0.4422234296798706,
+      "learning_rate": 1.2633541495060975e-05,
+      "loss": 0.2436,
+      "step": 11105
+    },
+    {
+      "epoch": 0.4153562165572353,
+      "grad_norm": 0.41787955164909363,
+      "learning_rate": 1.2627875772850672e-05,
+      "loss": 0.3395,
+      "step": 11110
+    },
+    {
+      "epoch": 0.4155431455475851,
+      "grad_norm": 1.9292072057724,
+      "learning_rate": 1.2622209144361313e-05,
+      "loss": 0.2723,
+      "step": 11115
+    },
+    {
+      "epoch": 0.4157300745379349,
+      "grad_norm": 0.3203314244747162,
+      "learning_rate": 1.2616541611547155e-05,
+      "loss": 0.2928,
+      "step": 11120
+    },
+    {
+      "epoch": 0.4159170035282847,
+      "grad_norm": 0.49962037801742554,
+      "learning_rate": 1.2610873176362767e-05,
+      "loss": 0.3809,
+      "step": 11125
+    },
+    {
+      "epoch": 0.4161039325186345,
+      "grad_norm": 0.4445996880531311,
+      "learning_rate": 1.2605203840763034e-05,
+      "loss": 0.3369,
+      "step": 11130
+    },
+    {
+      "epoch": 0.41629086150898426,
+      "grad_norm": 0.43643853068351746,
+      "learning_rate": 1.2599533606703144e-05,
+      "loss": 0.4172,
+      "step": 11135
+    },
+    {
+      "epoch": 0.4164777904993341,
+      "grad_norm": 0.7465499639511108,
+      "learning_rate": 1.2593862476138598e-05,
+      "loss": 0.4366,
+      "step": 11140
+    },
+    {
+      "epoch": 0.41666471948968387,
+      "grad_norm": 0.34575432538986206,
+      "learning_rate": 1.2588190451025209e-05,
+      "loss": 0.2783,
+      "step": 11145
+    },
+    {
+      "epoch": 0.41685164848003364,
+      "grad_norm": 0.33487460017204285,
+      "learning_rate": 1.2582517533319094e-05,
+      "loss": 0.3308,
+      "step": 11150
+    },
+    {
+      "epoch": 0.4170385774703834,
+      "grad_norm": 0.33010798692703247,
+      "learning_rate": 1.257684372497668e-05,
+      "loss": 0.3056,
+      "step": 11155
+    },
+    {
+      "epoch": 0.41722550646073325,
+      "grad_norm": 0.49482452869415283,
+      "learning_rate": 1.2571169027954702e-05,
+      "loss": 0.2821,
+      "step": 11160
+    },
+    {
+      "epoch": 0.41741243545108303,
+      "grad_norm": 0.23514960706233978,
+      "learning_rate": 1.25654934442102e-05,
+      "loss": 0.3404,
+      "step": 11165
+    },
+    {
+      "epoch": 0.4175993644414328,
+      "grad_norm": 0.7681819200515747,
+      "learning_rate": 1.2559816975700518e-05,
+      "loss": 0.3222,
+      "step": 11170
+    },
+    {
+      "epoch": 0.4177862934317826,
+      "grad_norm": 0.32423636317253113,
+      "learning_rate": 1.2554139624383307e-05,
+      "loss": 0.263,
+      "step": 11175
+    },
+    {
+      "epoch": 0.4179732224221324,
+      "grad_norm": 0.26687395572662354,
+      "learning_rate": 1.2548461392216531e-05,
+      "loss": 0.2467,
+      "step": 11180
+    },
+    {
+      "epoch": 0.4181601514124822,
+      "grad_norm": 0.3909884989261627,
+      "learning_rate": 1.2542782281158438e-05,
+      "loss": 0.2949,
+      "step": 11185
+    },
+    {
+      "epoch": 0.418347080402832,
+      "grad_norm": 0.27226680517196655,
+      "learning_rate": 1.2537102293167598e-05,
+      "loss": 0.2839,
+      "step": 11190
+    },
+    {
+      "epoch": 0.41853400939318175,
+      "grad_norm": 0.3636428415775299,
+      "learning_rate": 1.2531421430202875e-05,
+      "loss": 0.2814,
+      "step": 11195
+    },
+    {
+      "epoch": 0.4187209383835316,
+      "grad_norm": 0.43354684114456177,
+      "learning_rate": 1.2525739694223436e-05,
+      "loss": 0.3003,
+      "step": 11200
+    },
+    {
+      "epoch": 0.41890786737388136,
+      "grad_norm": 0.3936502933502197,
+      "learning_rate": 1.2520057087188748e-05,
+      "loss": 0.2926,
+      "step": 11205
+    },
+    {
+      "epoch": 0.41909479636423114,
+      "grad_norm": 0.37949803471565247,
+      "learning_rate": 1.2514373611058578e-05,
+      "loss": 0.3242,
+      "step": 11210
+    },
+    {
+      "epoch": 0.4192817253545809,
+      "grad_norm": 0.2999209463596344,
+      "learning_rate": 1.2508689267792994e-05,
+      "loss": 0.2679,
+      "step": 11215
+    },
+    {
+      "epoch": 0.41946865434493075,
+      "grad_norm": 0.37106993794441223,
+      "learning_rate": 1.2503004059352369e-05,
+      "loss": 0.2887,
+      "step": 11220
+    },
+    {
+      "epoch": 0.4196555833352805,
+      "grad_norm": 0.286905437707901,
+      "learning_rate": 1.2497317987697359e-05,
+      "loss": 0.2298,
+      "step": 11225
+    },
+    {
+      "epoch": 0.4198425123256303,
+      "grad_norm": 0.5081347823143005,
+      "learning_rate": 1.2491631054788936e-05,
+      "loss": 0.305,
+      "step": 11230
+    },
+    {
+      "epoch": 0.4200294413159801,
+      "grad_norm": 0.43394070863723755,
+      "learning_rate": 1.2485943262588353e-05,
+      "loss": 0.2757,
+      "step": 11235
+    },
+    {
+      "epoch": 0.42021637030632986,
+      "grad_norm": 0.6610644459724426,
+      "learning_rate": 1.2480254613057172e-05,
+      "loss": 0.3212,
+      "step": 11240
+    },
+    {
+      "epoch": 0.4204032992966797,
+      "grad_norm": 0.6914903521537781,
+      "learning_rate": 1.2474565108157244e-05,
+      "loss": 0.3204,
+      "step": 11245
+    },
+    {
+      "epoch": 0.42059022828702947,
+      "grad_norm": 0.31729546189308167,
+      "learning_rate": 1.2468874749850715e-05,
+      "loss": 0.2934,
+      "step": 11250
+    },
+    {
+      "epoch": 0.42077715727737924,
+      "grad_norm": 0.40817737579345703,
+      "learning_rate": 1.2463183540100028e-05,
+      "loss": 0.2576,
+      "step": 11255
+    },
+    {
+      "epoch": 0.420964086267729,
+      "grad_norm": 0.48512113094329834,
+      "learning_rate": 1.2457491480867917e-05,
+      "loss": 0.3062,
+      "step": 11260
+    },
+    {
+      "epoch": 0.42115101525807885,
+      "grad_norm": 0.32723644375801086,
+      "learning_rate": 1.2451798574117406e-05,
+      "loss": 0.2556,
+      "step": 11265
+    },
+    {
+      "epoch": 0.42133794424842863,
+      "grad_norm": 0.3997679054737091,
+      "learning_rate": 1.2446104821811825e-05,
+      "loss": 0.2993,
+      "step": 11270
+    },
+    {
+      "epoch": 0.4215248732387784,
+      "grad_norm": 0.24848319590091705,
+      "learning_rate": 1.2440410225914779e-05,
+      "loss": 0.32,
+      "step": 11275
+    },
+    {
+      "epoch": 0.4217118022291282,
+      "grad_norm": 0.33824488520622253,
+      "learning_rate": 1.243471478839017e-05,
+      "loss": 0.3054,
+      "step": 11280
+    },
+    {
+      "epoch": 0.421898731219478,
+      "grad_norm": 0.3888397216796875,
+      "learning_rate": 1.2429018511202195e-05,
+      "loss": 0.2567,
+      "step": 11285
+    },
+    {
+      "epoch": 0.4220856602098278,
+      "grad_norm": 0.25221407413482666,
+      "learning_rate": 1.2423321396315338e-05,
+      "loss": 0.2519,
+      "step": 11290
+    },
+    {
+      "epoch": 0.4222725892001776,
+      "grad_norm": 0.4825991094112396,
+      "learning_rate": 1.2417623445694367e-05,
+      "loss": 0.2718,
+      "step": 11295
+    },
+    {
+      "epoch": 0.42245951819052735,
+      "grad_norm": 0.4877346456050873,
+      "learning_rate": 1.2411924661304346e-05,
+      "loss": 0.2421,
+      "step": 11300
+    },
+    {
+      "epoch": 0.4226464471808772,
+      "grad_norm": 0.734813392162323,
+      "learning_rate": 1.2406225045110617e-05,
+      "loss": 0.2855,
+      "step": 11305
+    },
+    {
+      "epoch": 0.42283337617122696,
+      "grad_norm": 0.4634631276130676,
+      "learning_rate": 1.2400524599078816e-05,
+      "loss": 0.3167,
+      "step": 11310
+    },
+    {
+      "epoch": 0.42302030516157674,
+      "grad_norm": 0.2482909858226776,
+      "learning_rate": 1.2394823325174866e-05,
+      "loss": 0.333,
+      "step": 11315
+    },
+    {
+      "epoch": 0.4232072341519265,
+      "grad_norm": 0.31496381759643555,
+      "learning_rate": 1.2389121225364968e-05,
+      "loss": 0.3233,
+      "step": 11320
+    },
+    {
+      "epoch": 0.42339416314227635,
+      "grad_norm": 0.46915552020072937,
+      "learning_rate": 1.2383418301615622e-05,
+      "loss": 0.33,
+      "step": 11325
+    },
+    {
+      "epoch": 0.4235810921326261,
+      "grad_norm": 0.3510735332965851,
+      "learning_rate": 1.2377714555893595e-05,
+      "loss": 0.2903,
+      "step": 11330
+    },
+    {
+      "epoch": 0.4237680211229759,
+      "grad_norm": 0.7571256756782532,
+      "learning_rate": 1.2372009990165948e-05,
+      "loss": 0.3083,
+      "step": 11335
+    },
+    {
+      "epoch": 0.4239549501133257,
+      "grad_norm": 0.37375253438949585,
+      "learning_rate": 1.2366304606400021e-05,
+      "loss": 0.2759,
+      "step": 11340
+    },
+    {
+      "epoch": 0.4241418791036755,
+      "grad_norm": 0.4373493790626526,
+      "learning_rate": 1.236059840656344e-05,
+      "loss": 0.2503,
+      "step": 11345
+    },
+    {
+      "epoch": 0.4243288080940253,
+      "grad_norm": 0.284297376871109,
+      "learning_rate": 1.235489139262411e-05,
+      "loss": 0.2491,
+      "step": 11350
+    },
+    {
+      "epoch": 0.42451573708437507,
+      "grad_norm": 0.24776756763458252,
+      "learning_rate": 1.2349183566550212e-05,
+      "loss": 0.2494,
+      "step": 11355
+    },
+    {
+      "epoch": 0.42470266607472484,
+      "grad_norm": 1.0645568370819092,
+      "learning_rate": 1.2343474930310213e-05,
+      "loss": 0.2053,
+      "step": 11360
+    },
+    {
+      "epoch": 0.4248895950650747,
+      "grad_norm": 0.325014591217041,
+      "learning_rate": 1.2337765485872859e-05,
+      "loss": 0.2916,
+      "step": 11365
+    },
+    {
+      "epoch": 0.42507652405542445,
+      "grad_norm": 0.9632283449172974,
+      "learning_rate": 1.2332055235207179e-05,
+      "loss": 0.3179,
+      "step": 11370
+    },
+    {
+      "epoch": 0.42526345304577423,
+      "grad_norm": 1.18338143825531,
+      "learning_rate": 1.2326344180282468e-05,
+      "loss": 0.3688,
+      "step": 11375
+    },
+    {
+      "epoch": 0.425450382036124,
+      "grad_norm": 0.4803732633590698,
+      "learning_rate": 1.2320632323068306e-05,
+      "loss": 0.3574,
+      "step": 11380
+    },
+    {
+      "epoch": 0.42563731102647384,
+      "grad_norm": 0.38758915662765503,
+      "learning_rate": 1.2314919665534552e-05,
+      "loss": 0.3133,
+      "step": 11385
+    },
+    {
+      "epoch": 0.4258242400168236,
+      "grad_norm": 0.4360613524913788,
+      "learning_rate": 1.2309206209651336e-05,
+      "loss": 0.2571,
+      "step": 11390
+    },
+    {
+      "epoch": 0.4260111690071734,
+      "grad_norm": 0.39218610525131226,
+      "learning_rate": 1.2303491957389069e-05,
+      "loss": 0.2504,
+      "step": 11395
+    },
+    {
+      "epoch": 0.4261980979975232,
+      "grad_norm": 0.4993368685245514,
+      "learning_rate": 1.2297776910718424e-05,
+      "loss": 0.3063,
+      "step": 11400
+    },
+    {
+      "epoch": 0.426385026987873,
+      "grad_norm": 0.5808417201042175,
+      "learning_rate": 1.2292061071610365e-05,
+      "loss": 0.2726,
+      "step": 11405
+    },
+    {
+      "epoch": 0.4265719559782228,
+      "grad_norm": 0.31993383169174194,
+      "learning_rate": 1.2286344442036114e-05,
+      "loss": 0.386,
+      "step": 11410
+    },
+    {
+      "epoch": 0.42675888496857256,
+      "grad_norm": 0.17474965751171112,
+      "learning_rate": 1.228062702396718e-05,
+      "loss": 0.2783,
+      "step": 11415
+    },
+    {
+      "epoch": 0.42694581395892234,
+      "grad_norm": 0.5042828917503357,
+      "learning_rate": 1.2274908819375335e-05,
+      "loss": 0.2297,
+      "step": 11420
+    },
+    {
+      "epoch": 0.42713274294927217,
+      "grad_norm": 0.4278635084629059,
+      "learning_rate": 1.2269189830232622e-05,
+      "loss": 0.2978,
+      "step": 11425
+    },
+    {
+      "epoch": 0.42731967193962195,
+      "grad_norm": 0.5413123369216919,
+      "learning_rate": 1.2263470058511355e-05,
+      "loss": 0.3018,
+      "step": 11430
+    },
+    {
+      "epoch": 0.4275066009299717,
+      "grad_norm": 0.3124532401561737,
+      "learning_rate": 1.2257749506184125e-05,
+      "loss": 0.3438,
+      "step": 11435
+    },
+    {
+      "epoch": 0.4276935299203215,
+      "grad_norm": 0.3359132707118988,
+      "learning_rate": 1.2252028175223778e-05,
+      "loss": 0.2612,
+      "step": 11440
+    },
+    {
+      "epoch": 0.42788045891067134,
+      "grad_norm": 0.5123247504234314,
+      "learning_rate": 1.224630606760345e-05,
+      "loss": 0.288,
+      "step": 11445
+    },
+    {
+      "epoch": 0.4280673879010211,
+      "grad_norm": 0.48125380277633667,
+      "learning_rate": 1.2240583185296517e-05,
+      "loss": 0.3077,
+      "step": 11450
+    },
+    {
+      "epoch": 0.4282543168913709,
+      "grad_norm": 0.24318033456802368,
+      "learning_rate": 1.2234859530276647e-05,
+      "loss": 0.2642,
+      "step": 11455
+    },
+    {
+      "epoch": 0.42844124588172067,
+      "grad_norm": 0.3051791489124298,
+      "learning_rate": 1.2229135104517757e-05,
+      "loss": 0.3298,
+      "step": 11460
+    },
+    {
+      "epoch": 0.4286281748720705,
+      "grad_norm": 0.5335453152656555,
+      "learning_rate": 1.2223409909994048e-05,
+      "loss": 0.3253,
+      "step": 11465
+    },
+    {
+      "epoch": 0.4288151038624203,
+      "grad_norm": 0.7249236702919006,
+      "learning_rate": 1.221768394867997e-05,
+      "loss": 0.3483,
+      "step": 11470
+    },
+    {
+      "epoch": 0.42900203285277005,
+      "grad_norm": 0.16120702028274536,
+      "learning_rate": 1.221195722255024e-05,
+      "loss": 0.2528,
+      "step": 11475
+    },
+    {
+      "epoch": 0.42918896184311983,
+      "grad_norm": 0.6941366791725159,
+      "learning_rate": 1.2206229733579846e-05,
+      "loss": 0.2684,
+      "step": 11480
+    },
+    {
+      "epoch": 0.4293758908334696,
+      "grad_norm": 0.4987410306930542,
+      "learning_rate": 1.2200501483744032e-05,
+      "loss": 0.3327,
+      "step": 11485
+    },
+    {
+      "epoch": 0.42956281982381944,
+      "grad_norm": 0.20793752372264862,
+      "learning_rate": 1.2194772475018309e-05,
+      "loss": 0.2491,
+      "step": 11490
+    },
+    {
+      "epoch": 0.4297497488141692,
+      "grad_norm": 0.4314498007297516,
+      "learning_rate": 1.218904270937845e-05,
+      "loss": 0.2784,
+      "step": 11495
+    },
+    {
+      "epoch": 0.429936677804519,
+      "grad_norm": 0.5785399079322815,
+      "learning_rate": 1.2183312188800483e-05,
+      "loss": 0.2836,
+      "step": 11500
+    },
+    {
+      "epoch": 0.4301236067948688,
+      "grad_norm": 0.4460437297821045,
+      "learning_rate": 1.2177580915260698e-05,
+      "loss": 0.3361,
+      "step": 11505
+    },
+    {
+      "epoch": 0.4303105357852186,
+      "grad_norm": 0.34010326862335205,
+      "learning_rate": 1.2171848890735655e-05,
+      "loss": 0.2738,
+      "step": 11510
+    },
+    {
+      "epoch": 0.4304974647755684,
+      "grad_norm": 0.2805013060569763,
+      "learning_rate": 1.2166116117202162e-05,
+      "loss": 0.2981,
+      "step": 11515
+    },
+    {
+      "epoch": 0.43068439376591816,
+      "grad_norm": 0.3375371992588043,
+      "learning_rate": 1.2160382596637286e-05,
+      "loss": 0.2759,
+      "step": 11520
+    },
+    {
+      "epoch": 0.43087132275626794,
+      "grad_norm": 0.5983409881591797,
+      "learning_rate": 1.2154648331018355e-05,
+      "loss": 0.2989,
+      "step": 11525
+    },
+    {
+      "epoch": 0.43105825174661777,
+      "grad_norm": 0.4123179614543915,
+      "learning_rate": 1.2148913322322952e-05,
+      "loss": 0.3054,
+      "step": 11530
+    },
+    {
+      "epoch": 0.43124518073696755,
+      "grad_norm": 0.6426059603691101,
+      "learning_rate": 1.214317757252892e-05,
+      "loss": 0.3205,
+      "step": 11535
+    },
+    {
+      "epoch": 0.4314321097273173,
+      "grad_norm": 0.35368436574935913,
+      "learning_rate": 1.2137441083614351e-05,
+      "loss": 0.297,
+      "step": 11540
+    },
+    {
+      "epoch": 0.4316190387176671,
+      "grad_norm": 0.3455871343612671,
+      "learning_rate": 1.21317038575576e-05,
+      "loss": 0.3098,
+      "step": 11545
+    },
+    {
+      "epoch": 0.43180596770801694,
+      "grad_norm": 0.34755730628967285,
+      "learning_rate": 1.2125965896337266e-05,
+      "loss": 0.3382,
+      "step": 11550
+    },
+    {
+      "epoch": 0.4319928966983667,
+      "grad_norm": 0.5450564622879028,
+      "learning_rate": 1.2120227201932213e-05,
+      "loss": 0.2472,
+      "step": 11555
+    },
+    {
+      "epoch": 0.4321798256887165,
+      "grad_norm": 0.28686073422431946,
+      "learning_rate": 1.2114487776321553e-05,
+      "loss": 0.2876,
+      "step": 11560
+    },
+    {
+      "epoch": 0.43236675467906627,
+      "grad_norm": 0.33250150084495544,
+      "learning_rate": 1.2108747621484645e-05,
+      "loss": 0.3267,
+      "step": 11565
+    },
+    {
+      "epoch": 0.4325536836694161,
+      "grad_norm": 0.5938117504119873,
+      "learning_rate": 1.2103006739401105e-05,
+      "loss": 0.319,
+      "step": 11570
+    },
+    {
+      "epoch": 0.4327406126597659,
+      "grad_norm": 0.32513856887817383,
+      "learning_rate": 1.20972651320508e-05,
+      "loss": 0.2741,
+      "step": 11575
+    },
+    {
+      "epoch": 0.43292754165011565,
+      "grad_norm": 0.39806506037712097,
+      "learning_rate": 1.2091522801413844e-05,
+      "loss": 0.3497,
+      "step": 11580
+    },
+    {
+      "epoch": 0.43311447064046543,
+      "grad_norm": 0.41519683599472046,
+      "learning_rate": 1.2085779749470603e-05,
+      "loss": 0.3165,
+      "step": 11585
+    },
+    {
+      "epoch": 0.43330139963081526,
+      "grad_norm": 0.5625133514404297,
+      "learning_rate": 1.2080035978201695e-05,
+      "loss": 0.2961,
+      "step": 11590
+    },
+    {
+      "epoch": 0.43348832862116504,
+      "grad_norm": 0.3750125765800476,
+      "learning_rate": 1.2074291489587972e-05,
+      "loss": 0.3051,
+      "step": 11595
+    },
+    {
+      "epoch": 0.4336752576115148,
+      "grad_norm": 0.43589699268341064,
+      "learning_rate": 1.2068546285610556e-05,
+      "loss": 0.3055,
+      "step": 11600
+    },
+    {
+      "epoch": 0.4338621866018646,
+      "grad_norm": 0.29387718439102173,
+      "learning_rate": 1.2062800368250795e-05,
+      "loss": 0.2602,
+      "step": 11605
+    },
+    {
+      "epoch": 0.43404911559221443,
+      "grad_norm": 0.4198211133480072,
+      "learning_rate": 1.2057053739490297e-05,
+      "loss": 0.2962,
+      "step": 11610
+    },
+    {
+      "epoch": 0.4342360445825642,
+      "grad_norm": 0.38284847140312195,
+      "learning_rate": 1.2051306401310904e-05,
+      "loss": 0.3328,
+      "step": 11615
+    },
+    {
+      "epoch": 0.434422973572914,
+      "grad_norm": 0.49905702471733093,
+      "learning_rate": 1.204555835569471e-05,
+      "loss": 0.2479,
+      "step": 11620
+    },
+    {
+      "epoch": 0.43460990256326376,
+      "grad_norm": 0.2864576578140259,
+      "learning_rate": 1.2039809604624053e-05,
+      "loss": 0.2545,
+      "step": 11625
+    },
+    {
+      "epoch": 0.4347968315536136,
+      "grad_norm": 0.3651660084724426,
+      "learning_rate": 1.203406015008151e-05,
+      "loss": 0.2738,
+      "step": 11630
+    },
+    {
+      "epoch": 0.43498376054396337,
+      "grad_norm": 0.5334829688072205,
+      "learning_rate": 1.2028309994049907e-05,
+      "loss": 0.2902,
+      "step": 11635
+    },
+    {
+      "epoch": 0.43517068953431315,
+      "grad_norm": 0.4753320813179016,
+      "learning_rate": 1.20225591385123e-05,
+      "loss": 0.3009,
+      "step": 11640
+    },
+    {
+      "epoch": 0.4353576185246629,
+      "grad_norm": 0.3971807360649109,
+      "learning_rate": 1.2016807585452004e-05,
+      "loss": 0.4801,
+      "step": 11645
+    },
+    {
+      "epoch": 0.43554454751501276,
+      "grad_norm": 0.3786313831806183,
+      "learning_rate": 1.201105533685256e-05,
+      "loss": 0.2662,
+      "step": 11650
+    },
+    {
+      "epoch": 0.43573147650536254,
+      "grad_norm": 0.5463781356811523,
+      "learning_rate": 1.2005302394697755e-05,
+      "loss": 0.2264,
+      "step": 11655
+    },
+    {
+      "epoch": 0.4359184054957123,
+      "grad_norm": 0.29028788208961487,
+      "learning_rate": 1.1999548760971614e-05,
+      "loss": 0.2305,
+      "step": 11660
+    },
+    {
+      "epoch": 0.4361053344860621,
+      "grad_norm": 0.30021342635154724,
+      "learning_rate": 1.1993794437658397e-05,
+      "loss": 0.3219,
+      "step": 11665
+    },
+    {
+      "epoch": 0.4362922634764119,
+      "grad_norm": 0.3362751007080078,
+      "learning_rate": 1.1988039426742608e-05,
+      "loss": 0.3006,
+      "step": 11670
+    },
+    {
+      "epoch": 0.4364791924667617,
+      "grad_norm": 0.25194594264030457,
+      "learning_rate": 1.1982283730208988e-05,
+      "loss": 0.2612,
+      "step": 11675
+    },
+    {
+      "epoch": 0.4366661214571115,
+      "grad_norm": 0.5895513296127319,
+      "learning_rate": 1.1976527350042507e-05,
+      "loss": 0.2559,
+      "step": 11680
+    },
+    {
+      "epoch": 0.43685305044746126,
+      "grad_norm": 0.6330528855323792,
+      "learning_rate": 1.1970770288228377e-05,
+      "loss": 0.2361,
+      "step": 11685
+    },
+    {
+      "epoch": 0.4370399794378111,
+      "grad_norm": 0.37626221776008606,
+      "learning_rate": 1.1965012546752047e-05,
+      "loss": 0.2266,
+      "step": 11690
+    },
+    {
+      "epoch": 0.43722690842816087,
+      "grad_norm": 0.3309417963027954,
+      "learning_rate": 1.1959254127599191e-05,
+      "loss": 0.265,
+      "step": 11695
+    },
+    {
+      "epoch": 0.43741383741851064,
+      "grad_norm": 0.5689538717269897,
+      "learning_rate": 1.1953495032755726e-05,
+      "loss": 0.2866,
+      "step": 11700
+    },
+    {
+      "epoch": 0.4376007664088604,
+      "grad_norm": 0.4492878019809723,
+      "learning_rate": 1.1947735264207804e-05,
+      "loss": 0.2061,
+      "step": 11705
+    },
+    {
+      "epoch": 0.43778769539921025,
+      "grad_norm": 0.40549832582473755,
+      "learning_rate": 1.1941974823941795e-05,
+      "loss": 0.3215,
+      "step": 11710
+    },
+    {
+      "epoch": 0.43797462438956003,
+      "grad_norm": 0.667451798915863,
+      "learning_rate": 1.1936213713944315e-05,
+      "loss": 0.292,
+      "step": 11715
+    },
+    {
+      "epoch": 0.4381615533799098,
+      "grad_norm": 0.349060982465744,
+      "learning_rate": 1.1930451936202203e-05,
+      "loss": 0.3204,
+      "step": 11720
+    },
+    {
+      "epoch": 0.4383484823702596,
+      "grad_norm": 0.3580839931964874,
+      "learning_rate": 1.1924689492702534e-05,
+      "loss": 0.2576,
+      "step": 11725
+    },
+    {
+      "epoch": 0.43853541136060936,
+      "grad_norm": 0.49423980712890625,
+      "learning_rate": 1.1918926385432608e-05,
+      "loss": 0.2562,
+      "step": 11730
+    },
+    {
+      "epoch": 0.4387223403509592,
+      "grad_norm": 0.2973886728286743,
+      "learning_rate": 1.1913162616379956e-05,
+      "loss": 0.3051,
+      "step": 11735
+    },
+    {
+      "epoch": 0.43890926934130897,
+      "grad_norm": 0.50880366563797,
+      "learning_rate": 1.1907398187532337e-05,
+      "loss": 0.2851,
+      "step": 11740
+    },
+    {
+      "epoch": 0.43909619833165875,
+      "grad_norm": 0.4905702471733093,
+      "learning_rate": 1.1901633100877736e-05,
+      "loss": 0.2635,
+      "step": 11745
+    },
+    {
+      "epoch": 0.4392831273220085,
+      "grad_norm": 0.21061702072620392,
+      "learning_rate": 1.1895867358404369e-05,
+      "loss": 0.2665,
+      "step": 11750
+    },
+    {
+      "epoch": 0.43947005631235836,
+      "grad_norm": 0.32583504915237427,
+      "learning_rate": 1.1890100962100672e-05,
+      "loss": 0.2656,
+      "step": 11755
+    },
+    {
+      "epoch": 0.43965698530270814,
+      "grad_norm": 0.2895315885543823,
+      "learning_rate": 1.1884333913955312e-05,
+      "loss": 0.2501,
+      "step": 11760
+    },
+    {
+      "epoch": 0.4398439142930579,
+      "grad_norm": 0.40925106406211853,
+      "learning_rate": 1.187856621595718e-05,
+      "loss": 0.2625,
+      "step": 11765
+    },
+    {
+      "epoch": 0.4400308432834077,
+      "grad_norm": 0.1345365196466446,
+      "learning_rate": 1.1872797870095385e-05,
+      "loss": 0.3494,
+      "step": 11770
+    },
+    {
+      "epoch": 0.4402177722737575,
+      "grad_norm": 0.5240778923034668,
+      "learning_rate": 1.1867028878359266e-05,
+      "loss": 0.2615,
+      "step": 11775
+    },
+    {
+      "epoch": 0.4404047012641073,
+      "grad_norm": 0.3817775547504425,
+      "learning_rate": 1.1861259242738386e-05,
+      "loss": 0.272,
+      "step": 11780
+    },
+    {
+      "epoch": 0.4405916302544571,
+      "grad_norm": 0.3994652032852173,
+      "learning_rate": 1.1855488965222526e-05,
+      "loss": 0.2841,
+      "step": 11785
+    },
+    {
+      "epoch": 0.44077855924480686,
+      "grad_norm": 0.461319237947464,
+      "learning_rate": 1.1849718047801686e-05,
+      "loss": 0.3008,
+      "step": 11790
+    },
+    {
+      "epoch": 0.4409654882351567,
+      "grad_norm": 0.5307036638259888,
+      "learning_rate": 1.1843946492466093e-05,
+      "loss": 0.2497,
+      "step": 11795
+    },
+    {
+      "epoch": 0.44115241722550647,
+      "grad_norm": 0.37287020683288574,
+      "learning_rate": 1.1838174301206194e-05,
+      "loss": 0.2826,
+      "step": 11800
+    },
+    {
+      "epoch": 0.44133934621585624,
+      "grad_norm": 0.538077712059021,
+      "learning_rate": 1.1832401476012645e-05,
+      "loss": 0.2475,
+      "step": 11805
+    },
+    {
+      "epoch": 0.441526275206206,
+      "grad_norm": 0.3736201524734497,
+      "learning_rate": 1.1826628018876334e-05,
+      "loss": 0.265,
+      "step": 11810
+    },
+    {
+      "epoch": 0.44171320419655585,
+      "grad_norm": 0.272097647190094,
+      "learning_rate": 1.1820853931788357e-05,
+      "loss": 0.3053,
+      "step": 11815
+    },
+    {
+      "epoch": 0.44190013318690563,
+      "grad_norm": 0.5018362998962402,
+      "learning_rate": 1.1815079216740033e-05,
+      "loss": 0.3002,
+      "step": 11820
+    },
+    {
+      "epoch": 0.4420870621772554,
+      "grad_norm": 0.5243871808052063,
+      "learning_rate": 1.1809303875722896e-05,
+      "loss": 0.3192,
+      "step": 11825
+    },
+    {
+      "epoch": 0.4422739911676052,
+      "grad_norm": 0.20126721262931824,
+      "learning_rate": 1.1803527910728695e-05,
+      "loss": 0.3049,
+      "step": 11830
+    },
+    {
+      "epoch": 0.442460920157955,
+      "grad_norm": 0.824718177318573,
+      "learning_rate": 1.1797751323749396e-05,
+      "loss": 0.3462,
+      "step": 11835
+    },
+    {
+      "epoch": 0.4426478491483048,
+      "grad_norm": 0.6242169737815857,
+      "learning_rate": 1.1791974116777179e-05,
+      "loss": 0.2373,
+      "step": 11840
+    },
+    {
+      "epoch": 0.4428347781386546,
+      "grad_norm": 0.23284803330898285,
+      "learning_rate": 1.1786196291804432e-05,
+      "loss": 0.2514,
+      "step": 11845
+    },
+    {
+      "epoch": 0.44302170712900435,
+      "grad_norm": 0.5042092800140381,
+      "learning_rate": 1.1780417850823768e-05,
+      "loss": 0.2548,
+      "step": 11850
+    },
+    {
+      "epoch": 0.4432086361193542,
+      "grad_norm": 0.28074219822883606,
+      "learning_rate": 1.1774638795828004e-05,
+      "loss": 0.2807,
+      "step": 11855
+    },
+    {
+      "epoch": 0.44339556510970396,
+      "grad_norm": 0.36168691515922546,
+      "learning_rate": 1.1768859128810167e-05,
+      "loss": 0.2616,
+      "step": 11860
+    },
+    {
+      "epoch": 0.44358249410005374,
+      "grad_norm": 0.6940954923629761,
+      "learning_rate": 1.1763078851763497e-05,
+      "loss": 0.3262,
+      "step": 11865
+    },
+    {
+      "epoch": 0.4437694230904035,
+      "grad_norm": 0.29294854402542114,
+      "learning_rate": 1.1757297966681455e-05,
+      "loss": 0.2751,
+      "step": 11870
+    },
+    {
+      "epoch": 0.44395635208075335,
+      "grad_norm": 0.2644307017326355,
+      "learning_rate": 1.1751516475557696e-05,
+      "loss": 0.2402,
+      "step": 11875
+    },
+    {
+      "epoch": 0.4441432810711031,
+      "grad_norm": 1.0284477472305298,
+      "learning_rate": 1.1745734380386091e-05,
+      "loss": 0.3128,
+      "step": 11880
+    },
+    {
+      "epoch": 0.4443302100614529,
+      "grad_norm": 0.3318440616130829,
+      "learning_rate": 1.1739951683160719e-05,
+      "loss": 0.2859,
+      "step": 11885
+    },
+    {
+      "epoch": 0.4445171390518027,
+      "grad_norm": 0.49884122610092163,
+      "learning_rate": 1.173416838587587e-05,
+      "loss": 0.3245,
+      "step": 11890
+    },
+    {
+      "epoch": 0.4447040680421525,
+      "grad_norm": 0.12781549990177155,
+      "learning_rate": 1.1728384490526035e-05,
+      "loss": 0.2808,
+      "step": 11895
+    },
+    {
+      "epoch": 0.4448909970325023,
+      "grad_norm": 0.4701847434043884,
+      "learning_rate": 1.1722599999105913e-05,
+      "loss": 0.3074,
+      "step": 11900
+    },
+    {
+      "epoch": 0.44507792602285207,
+      "grad_norm": 0.28572845458984375,
+      "learning_rate": 1.1716814913610409e-05,
+      "loss": 0.3912,
+      "step": 11905
+    },
+    {
+      "epoch": 0.44526485501320184,
+      "grad_norm": 0.36786168813705444,
+      "learning_rate": 1.1711029236034633e-05,
+      "loss": 0.2701,
+      "step": 11910
+    },
+    {
+      "epoch": 0.4454517840035517,
+      "grad_norm": 0.49527883529663086,
+      "learning_rate": 1.1705242968373903e-05,
+      "loss": 0.2451,
+      "step": 11915
+    },
+    {
+      "epoch": 0.44563871299390145,
+      "grad_norm": 0.46449634432792664,
+      "learning_rate": 1.1699456112623733e-05,
+      "loss": 0.2992,
+      "step": 11920
+    },
+    {
+      "epoch": 0.44582564198425123,
+      "grad_norm": 0.4579066038131714,
+      "learning_rate": 1.1693668670779847e-05,
+      "loss": 0.2084,
+      "step": 11925
+    },
+    {
+      "epoch": 0.446012570974601,
+      "grad_norm": 0.1956116259098053,
+      "learning_rate": 1.1687880644838164e-05,
+      "loss": 0.3574,
+      "step": 11930
+    },
+    {
+      "epoch": 0.44619949996495084,
+      "grad_norm": 0.37438297271728516,
+      "learning_rate": 1.1682092036794812e-05,
+      "loss": 0.3066,
+      "step": 11935
+    },
+    {
+      "epoch": 0.4463864289553006,
+      "grad_norm": 0.37410229444503784,
+      "learning_rate": 1.1676302848646116e-05,
+      "loss": 0.3174,
+      "step": 11940
+    },
+    {
+      "epoch": 0.4465733579456504,
+      "grad_norm": 0.23035752773284912,
+      "learning_rate": 1.16705130823886e-05,
+      "loss": 0.2645,
+      "step": 11945
+    },
+    {
+      "epoch": 0.4467602869360002,
+      "grad_norm": 0.21039049327373505,
+      "learning_rate": 1.1664722740018983e-05,
+      "loss": 0.3209,
+      "step": 11950
+    },
+    {
+      "epoch": 0.44694721592635,
+      "grad_norm": 0.4789251387119293,
+      "learning_rate": 1.1658931823534196e-05,
+      "loss": 0.3161,
+      "step": 11955
+    },
+    {
+      "epoch": 0.4471341449166998,
+      "grad_norm": 0.33394578099250793,
+      "learning_rate": 1.1653140334931357e-05,
+      "loss": 0.2549,
+      "step": 11960
+    },
+    {
+      "epoch": 0.44732107390704956,
+      "grad_norm": 0.5714621543884277,
+      "learning_rate": 1.1647348276207789e-05,
+      "loss": 0.3469,
+      "step": 11965
+    },
+    {
+      "epoch": 0.44750800289739934,
+      "grad_norm": 0.8689758777618408,
+      "learning_rate": 1.1641555649360998e-05,
+      "loss": 0.4199,
+      "step": 11970
+    },
+    {
+      "epoch": 0.4476949318877491,
+      "grad_norm": 0.1597885936498642,
+      "learning_rate": 1.1635762456388702e-05,
+      "loss": 0.3961,
+      "step": 11975
+    },
+    {
+      "epoch": 0.44788186087809895,
+      "grad_norm": 0.46579235792160034,
+      "learning_rate": 1.1629968699288805e-05,
+      "loss": 0.2853,
+      "step": 11980
+    },
+    {
+      "epoch": 0.4480687898684487,
+      "grad_norm": 0.34550490975379944,
+      "learning_rate": 1.162417438005941e-05,
+      "loss": 0.2943,
+      "step": 11985
+    },
+    {
+      "epoch": 0.4482557188587985,
+      "grad_norm": 0.4065869450569153,
+      "learning_rate": 1.1618379500698808e-05,
+      "loss": 0.3224,
+      "step": 11990
+    },
+    {
+      "epoch": 0.4484426478491483,
+      "grad_norm": 0.19699564576148987,
+      "learning_rate": 1.161258406320549e-05,
+      "loss": 0.3538,
+      "step": 11995
+    },
+    {
+      "epoch": 0.4486295768394981,
+      "grad_norm": 0.851565420627594,
+      "learning_rate": 1.1606788069578132e-05,
+      "loss": 0.3719,
+      "step": 12000
+    },
+    {
+      "epoch": 0.4488165058298479,
+      "grad_norm": 0.36260199546813965,
+      "learning_rate": 1.160099152181561e-05,
+      "loss": 0.3062,
+      "step": 12005
+    },
+    {
+      "epoch": 0.44900343482019767,
+      "grad_norm": 0.36545851826667786,
+      "learning_rate": 1.1595194421916988e-05,
+      "loss": 0.329,
+      "step": 12010
+    },
+    {
+      "epoch": 0.44919036381054744,
+      "grad_norm": 0.339874804019928,
+      "learning_rate": 1.1589396771881518e-05,
+      "loss": 0.2977,
+      "step": 12015
+    },
+    {
+      "epoch": 0.4493772928008973,
+      "grad_norm": 0.4972774386405945,
+      "learning_rate": 1.1583598573708642e-05,
+      "loss": 0.385,
+      "step": 12020
+    },
+    {
+      "epoch": 0.44956422179124705,
+      "grad_norm": 0.41324669122695923,
+      "learning_rate": 1.1577799829397996e-05,
+      "loss": 0.2967,
+      "step": 12025
+    },
+    {
+      "epoch": 0.44975115078159683,
+      "grad_norm": 0.44996240735054016,
+      "learning_rate": 1.1572000540949398e-05,
+      "loss": 0.2517,
+      "step": 12030
+    },
+    {
+      "epoch": 0.4499380797719466,
+      "grad_norm": 0.4872210621833801,
+      "learning_rate": 1.1566200710362854e-05,
+      "loss": 0.2805,
+      "step": 12035
+    },
+    {
+      "epoch": 0.45012500876229644,
+      "grad_norm": 0.3203275799751282,
+      "learning_rate": 1.1560400339638567e-05,
+      "loss": 0.2458,
+      "step": 12040
+    },
+    {
+      "epoch": 0.4503119377526462,
+      "grad_norm": 0.733971118927002,
+      "learning_rate": 1.1554599430776906e-05,
+      "loss": 0.2399,
+      "step": 12045
+    },
+    {
+      "epoch": 0.450498866742996,
+      "grad_norm": 0.38670045137405396,
+      "learning_rate": 1.1548797985778452e-05,
+      "loss": 0.2588,
+      "step": 12050
+    },
+    {
+      "epoch": 0.4506857957333458,
+      "grad_norm": 0.33743932843208313,
+      "learning_rate": 1.1542996006643952e-05,
+      "loss": 0.2816,
+      "step": 12055
+    },
+    {
+      "epoch": 0.4508727247236956,
+      "grad_norm": 0.23398756980895996,
+      "learning_rate": 1.1537193495374342e-05,
+      "loss": 0.2784,
+      "step": 12060
+    },
+    {
+      "epoch": 0.4510596537140454,
+      "grad_norm": 0.5807647705078125,
+      "learning_rate": 1.153139045397074e-05,
+      "loss": 0.3134,
+      "step": 12065
+    },
+    {
+      "epoch": 0.45124658270439516,
+      "grad_norm": 0.4573014974594116,
+      "learning_rate": 1.152558688443445e-05,
+      "loss": 0.2615,
+      "step": 12070
+    },
+    {
+      "epoch": 0.45143351169474494,
+      "grad_norm": 0.31440410017967224,
+      "learning_rate": 1.1519782788766957e-05,
+      "loss": 0.3578,
+      "step": 12075
+    },
+    {
+      "epoch": 0.45162044068509477,
+      "grad_norm": 0.19732816517353058,
+      "learning_rate": 1.1513978168969929e-05,
+      "loss": 0.2974,
+      "step": 12080
+    },
+    {
+      "epoch": 0.45180736967544455,
+      "grad_norm": 0.17791102826595306,
+      "learning_rate": 1.1508173027045214e-05,
+      "loss": 0.2286,
+      "step": 12085
+    },
+    {
+      "epoch": 0.4519942986657943,
+      "grad_norm": 0.35120418667793274,
+      "learning_rate": 1.1502367364994832e-05,
+      "loss": 0.3173,
+      "step": 12090
+    },
+    {
+      "epoch": 0.4521812276561441,
+      "grad_norm": 0.5374488234519958,
+      "learning_rate": 1.1496561184820991e-05,
+      "loss": 0.2903,
+      "step": 12095
+    },
+    {
+      "epoch": 0.45236815664649394,
+      "grad_norm": 0.37684205174446106,
+      "learning_rate": 1.1490754488526084e-05,
+      "loss": 0.3131,
+      "step": 12100
+    },
+    {
+      "epoch": 0.4525550856368437,
+      "grad_norm": 0.3075244426727295,
+      "learning_rate": 1.1484947278112673e-05,
+      "loss": 0.2625,
+      "step": 12105
+    },
+    {
+      "epoch": 0.4527420146271935,
+      "grad_norm": 0.5102051496505737,
+      "learning_rate": 1.147913955558349e-05,
+      "loss": 0.3038,
+      "step": 12110
+    },
+    {
+      "epoch": 0.45292894361754327,
+      "grad_norm": 0.2178102284669876,
+      "learning_rate": 1.1473331322941457e-05,
+      "loss": 0.2539,
+      "step": 12115
+    },
+    {
+      "epoch": 0.4531158726078931,
+      "grad_norm": 0.20256425440311432,
+      "learning_rate": 1.1467522582189667e-05,
+      "loss": 0.2735,
+      "step": 12120
+    },
+    {
+      "epoch": 0.4533028015982429,
+      "grad_norm": 0.5971643924713135,
+      "learning_rate": 1.1461713335331389e-05,
+      "loss": 0.4114,
+      "step": 12125
+    },
+    {
+      "epoch": 0.45348973058859265,
+      "grad_norm": 0.5443209409713745,
+      "learning_rate": 1.1455903584370065e-05,
+      "loss": 0.3009,
+      "step": 12130
+    },
+    {
+      "epoch": 0.45367665957894243,
+      "grad_norm": 0.6024529933929443,
+      "learning_rate": 1.1450093331309314e-05,
+      "loss": 0.2964,
+      "step": 12135
+    },
+    {
+      "epoch": 0.45386358856929226,
+      "grad_norm": 0.4719208776950836,
+      "learning_rate": 1.1444282578152918e-05,
+      "loss": 0.3716,
+      "step": 12140
+    },
+    {
+      "epoch": 0.45405051755964204,
+      "grad_norm": 0.3875749707221985,
+      "learning_rate": 1.1438471326904847e-05,
+      "loss": 0.2173,
+      "step": 12145
+    },
+    {
+      "epoch": 0.4542374465499918,
+      "grad_norm": 0.40976300835609436,
+      "learning_rate": 1.1432659579569234e-05,
+      "loss": 0.2978,
+      "step": 12150
+    },
+    {
+      "epoch": 0.4544243755403416,
+      "grad_norm": 0.32276174426078796,
+      "learning_rate": 1.1426847338150386e-05,
+      "loss": 0.3428,
+      "step": 12155
+    },
+    {
+      "epoch": 0.45461130453069143,
+      "grad_norm": 0.2936963140964508,
+      "learning_rate": 1.1421034604652771e-05,
+      "loss": 0.3055,
+      "step": 12160
+    },
+    {
+      "epoch": 0.4547982335210412,
+      "grad_norm": 0.24648843705654144,
+      "learning_rate": 1.1415221381081039e-05,
+      "loss": 0.3737,
+      "step": 12165
+    },
+    {
+      "epoch": 0.454985162511391,
+      "grad_norm": 0.3906302750110626,
+      "learning_rate": 1.1409407669440005e-05,
+      "loss": 0.2709,
+      "step": 12170
+    },
+    {
+      "epoch": 0.45517209150174076,
+      "grad_norm": 0.23113980889320374,
+      "learning_rate": 1.140359347173465e-05,
+      "loss": 0.2874,
+      "step": 12175
+    },
+    {
+      "epoch": 0.4553590204920906,
+      "grad_norm": 0.4913095533847809,
+      "learning_rate": 1.1397778789970126e-05,
+      "loss": 0.253,
+      "step": 12180
+    },
+    {
+      "epoch": 0.45554594948244037,
+      "grad_norm": 0.34004727005958557,
+      "learning_rate": 1.1391963626151745e-05,
+      "loss": 0.2647,
+      "step": 12185
+    },
+    {
+      "epoch": 0.45573287847279015,
+      "grad_norm": 0.41701698303222656,
+      "learning_rate": 1.1386147982284996e-05,
+      "loss": 0.2685,
+      "step": 12190
+    },
+    {
+      "epoch": 0.4559198074631399,
+      "grad_norm": 0.12583187222480774,
+      "learning_rate": 1.1380331860375527e-05,
+      "loss": 0.2764,
+      "step": 12195
+    },
+    {
+      "epoch": 0.4561067364534897,
+      "grad_norm": 0.29735517501831055,
+      "learning_rate": 1.1374515262429154e-05,
+      "loss": 0.2472,
+      "step": 12200
+    },
+    {
+      "epoch": 0.45629366544383954,
+      "grad_norm": 0.5592828989028931,
+      "learning_rate": 1.1368698190451848e-05,
+      "loss": 0.3773,
+      "step": 12205
+    },
+    {
+      "epoch": 0.4564805944341893,
+      "grad_norm": 0.5007039904594421,
+      "learning_rate": 1.1362880646449755e-05,
+      "loss": 0.2827,
+      "step": 12210
+    },
+    {
+      "epoch": 0.4566675234245391,
+      "grad_norm": 0.5219624042510986,
+      "learning_rate": 1.1357062632429177e-05,
+      "loss": 0.3227,
+      "step": 12215
+    },
+    {
+      "epoch": 0.45685445241488887,
+      "grad_norm": 0.45227476954460144,
+      "learning_rate": 1.1351244150396581e-05,
+      "loss": 0.329,
+      "step": 12220
+    },
+    {
+      "epoch": 0.4570413814052387,
+      "grad_norm": 0.5245965123176575,
+      "learning_rate": 1.1345425202358597e-05,
+      "loss": 0.3021,
+      "step": 12225
+    },
+    {
+      "epoch": 0.4572283103955885,
+      "grad_norm": 0.43462806940078735,
+      "learning_rate": 1.1339605790322016e-05,
+      "loss": 0.3904,
+      "step": 12230
+    },
+    {
+      "epoch": 0.45741523938593825,
+      "grad_norm": 0.44027209281921387,
+      "learning_rate": 1.1333785916293776e-05,
+      "loss": 0.2519,
+      "step": 12235
+    },
+    {
+      "epoch": 0.45760216837628803,
+      "grad_norm": 0.7230022549629211,
+      "learning_rate": 1.1327965582280995e-05,
+      "loss": 0.3982,
+      "step": 12240
+    },
+    {
+      "epoch": 0.45778909736663786,
+      "grad_norm": 0.5000762939453125,
+      "learning_rate": 1.1322144790290935e-05,
+      "loss": 0.3101,
+      "step": 12245
+    },
+    {
+      "epoch": 0.45797602635698764,
+      "grad_norm": 0.3863626718521118,
+      "learning_rate": 1.1316323542331022e-05,
+      "loss": 0.3012,
+      "step": 12250
+    },
+    {
+      "epoch": 0.4581629553473374,
+      "grad_norm": 0.34956860542297363,
+      "learning_rate": 1.1310501840408837e-05,
+      "loss": 0.3426,
+      "step": 12255
+    },
+    {
+      "epoch": 0.4583498843376872,
+      "grad_norm": 0.3766677975654602,
+      "learning_rate": 1.1304679686532116e-05,
+      "loss": 0.2716,
+      "step": 12260
+    },
+    {
+      "epoch": 0.45853681332803703,
+      "grad_norm": 0.5153632164001465,
+      "learning_rate": 1.1298857082708754e-05,
+      "loss": 0.2259,
+      "step": 12265
+    },
+    {
+      "epoch": 0.4587237423183868,
+      "grad_norm": 0.4910812973976135,
+      "learning_rate": 1.1293034030946804e-05,
+      "loss": 0.3003,
+      "step": 12270
+    },
+    {
+      "epoch": 0.4589106713087366,
+      "grad_norm": 0.43328529596328735,
+      "learning_rate": 1.1287210533254464e-05,
+      "loss": 0.3041,
+      "step": 12275
+    },
+    {
+      "epoch": 0.45909760029908636,
+      "grad_norm": 0.24060022830963135,
+      "learning_rate": 1.1281386591640096e-05,
+      "loss": 0.3299,
+      "step": 12280
+    },
+    {
+      "epoch": 0.4592845292894362,
+      "grad_norm": 0.2626323997974396,
+      "learning_rate": 1.127556220811221e-05,
+      "loss": 0.3169,
+      "step": 12285
+    },
+    {
+      "epoch": 0.45947145827978597,
+      "grad_norm": 0.27137619256973267,
+      "learning_rate": 1.1269737384679465e-05,
+      "loss": 0.2479,
+      "step": 12290
+    },
+    {
+      "epoch": 0.45965838727013575,
+      "grad_norm": 0.3326461911201477,
+      "learning_rate": 1.1263912123350679e-05,
+      "loss": 0.2324,
+      "step": 12295
+    },
+    {
+      "epoch": 0.4598453162604855,
+      "grad_norm": 0.20982854068279266,
+      "learning_rate": 1.1258086426134822e-05,
+      "loss": 0.2547,
+      "step": 12300
+    },
+    {
+      "epoch": 0.46003224525083536,
+      "grad_norm": 0.3248259127140045,
+      "learning_rate": 1.1252260295041003e-05,
+      "loss": 0.2574,
+      "step": 12305
+    },
+    {
+      "epoch": 0.46021917424118514,
+      "grad_norm": 0.35238370299339294,
+      "learning_rate": 1.1246433732078487e-05,
+      "loss": 0.2612,
+      "step": 12310
+    },
+    {
+      "epoch": 0.4604061032315349,
+      "grad_norm": 0.5635769367218018,
+      "learning_rate": 1.1240606739256694e-05,
+      "loss": 0.3106,
+      "step": 12315
+    },
+    {
+      "epoch": 0.4605930322218847,
+      "grad_norm": 0.33287879824638367,
+      "learning_rate": 1.1234779318585182e-05,
+      "loss": 0.24,
+      "step": 12320
+    },
+    {
+      "epoch": 0.4607799612122345,
+      "grad_norm": 0.5921958684921265,
+      "learning_rate": 1.1228951472073669e-05,
+      "loss": 0.2757,
+      "step": 12325
+    },
+    {
+      "epoch": 0.4609668902025843,
+      "grad_norm": 0.32586291432380676,
+      "learning_rate": 1.1223123201732002e-05,
+      "loss": 0.3202,
+      "step": 12330
+    },
+    {
+      "epoch": 0.4611538191929341,
+      "grad_norm": 0.4058232605457306,
+      "learning_rate": 1.1217294509570193e-05,
+      "loss": 0.2255,
+      "step": 12335
+    },
+    {
+      "epoch": 0.46134074818328386,
+      "grad_norm": 0.5720903873443604,
+      "learning_rate": 1.1211465397598387e-05,
+      "loss": 0.3099,
+      "step": 12340
+    },
+    {
+      "epoch": 0.4615276771736337,
+      "grad_norm": 0.43414196372032166,
+      "learning_rate": 1.1205635867826878e-05,
+      "loss": 0.2535,
+      "step": 12345
+    },
+    {
+      "epoch": 0.46171460616398347,
+      "grad_norm": 0.5103111863136292,
+      "learning_rate": 1.1199805922266105e-05,
+      "loss": 0.2551,
+      "step": 12350
+    },
+    {
+      "epoch": 0.46190153515433324,
+      "grad_norm": 0.4893711805343628,
+      "learning_rate": 1.1193975562926646e-05,
+      "loss": 0.2794,
+      "step": 12355
+    },
+    {
+      "epoch": 0.462088464144683,
+      "grad_norm": 0.48441797494888306,
+      "learning_rate": 1.1188144791819226e-05,
+      "loss": 0.4306,
+      "step": 12360
+    },
+    {
+      "epoch": 0.46227539313503285,
+      "grad_norm": 0.2794596254825592,
+      "learning_rate": 1.1182313610954716e-05,
+      "loss": 0.2915,
+      "step": 12365
+    },
+    {
+      "epoch": 0.46246232212538263,
+      "grad_norm": 0.5436064004898071,
+      "learning_rate": 1.1176482022344115e-05,
+      "loss": 0.297,
+      "step": 12370
+    },
+    {
+      "epoch": 0.4626492511157324,
+      "grad_norm": 0.40309974551200867,
+      "learning_rate": 1.1170650027998577e-05,
+      "loss": 0.4027,
+      "step": 12375
+    },
+    {
+      "epoch": 0.4628361801060822,
+      "grad_norm": 0.1574869155883789,
+      "learning_rate": 1.116481762992939e-05,
+      "loss": 0.2339,
+      "step": 12380
+    },
+    {
+      "epoch": 0.463023109096432,
+      "grad_norm": 0.43917763233184814,
+      "learning_rate": 1.1158984830147975e-05,
+      "loss": 0.3213,
+      "step": 12385
+    },
+    {
+      "epoch": 0.4632100380867818,
+      "grad_norm": 0.34497371315956116,
+      "learning_rate": 1.1153151630665902e-05,
+      "loss": 0.2609,
+      "step": 12390
+    },
+    {
+      "epoch": 0.46339696707713157,
+      "grad_norm": 0.3034207224845886,
+      "learning_rate": 1.1147318033494876e-05,
+      "loss": 0.2787,
+      "step": 12395
+    },
+    {
+      "epoch": 0.46358389606748135,
+      "grad_norm": 0.28706076741218567,
+      "learning_rate": 1.1141484040646732e-05,
+      "loss": 0.2624,
+      "step": 12400
+    },
+    {
+      "epoch": 0.4637708250578312,
+      "grad_norm": 0.3962991237640381,
+      "learning_rate": 1.1135649654133453e-05,
+      "loss": 0.2194,
+      "step": 12405
+    },
+    {
+      "epoch": 0.46395775404818096,
+      "grad_norm": 0.6293958425521851,
+      "learning_rate": 1.1129814875967143e-05,
+      "loss": 0.3141,
+      "step": 12410
+    },
+    {
+      "epoch": 0.46414468303853074,
+      "grad_norm": 0.4316769242286682,
+      "learning_rate": 1.1123979708160064e-05,
+      "loss": 0.2896,
+      "step": 12415
+    },
+    {
+      "epoch": 0.4643316120288805,
+      "grad_norm": 0.4425215423107147,
+      "learning_rate": 1.1118144152724584e-05,
+      "loss": 0.3304,
+      "step": 12420
+    },
+    {
+      "epoch": 0.46451854101923035,
+      "grad_norm": 0.43086567521095276,
+      "learning_rate": 1.1112308211673226e-05,
+      "loss": 0.3048,
+      "step": 12425
+    },
+    {
+      "epoch": 0.4647054700095801,
+      "grad_norm": 0.5925473570823669,
+      "learning_rate": 1.1106471887018637e-05,
+      "loss": 0.3241,
+      "step": 12430
+    },
+    {
+      "epoch": 0.4648923989999299,
+      "grad_norm": 0.4298711121082306,
+      "learning_rate": 1.11006351807736e-05,
+      "loss": 0.3052,
+      "step": 12435
+    },
+    {
+      "epoch": 0.4650793279902797,
+      "grad_norm": 0.40438607335090637,
+      "learning_rate": 1.1094798094951027e-05,
+      "loss": 0.3083,
+      "step": 12440
+    },
+    {
+      "epoch": 0.46526625698062946,
+      "grad_norm": 0.1780761182308197,
+      "learning_rate": 1.1088960631563958e-05,
+      "loss": 0.3185,
+      "step": 12445
+    },
+    {
+      "epoch": 0.4654531859709793,
+      "grad_norm": 0.5064865946769714,
+      "learning_rate": 1.108312279262557e-05,
+      "loss": 0.2251,
+      "step": 12450
+    },
+    {
+      "epoch": 0.46564011496132907,
+      "grad_norm": 0.49286672472953796,
+      "learning_rate": 1.1077284580149169e-05,
+      "loss": 0.2824,
+      "step": 12455
+    },
+    {
+      "epoch": 0.46582704395167884,
+      "grad_norm": 0.35339727997779846,
+      "learning_rate": 1.1071445996148182e-05,
+      "loss": 0.3037,
+      "step": 12460
+    },
+    {
+      "epoch": 0.4660139729420286,
+      "grad_norm": 0.5857782959938049,
+      "learning_rate": 1.1065607042636173e-05,
+      "loss": 0.3552,
+      "step": 12465
+    },
+    {
+      "epoch": 0.46620090193237845,
+      "grad_norm": 0.4888255000114441,
+      "learning_rate": 1.1059767721626828e-05,
+      "loss": 0.3064,
+      "step": 12470
+    },
+    {
+      "epoch": 0.46638783092272823,
+      "grad_norm": 0.2896244525909424,
+      "learning_rate": 1.1053928035133964e-05,
+      "loss": 0.2304,
+      "step": 12475
+    },
+    {
+      "epoch": 0.466574759913078,
+      "grad_norm": 0.26369529962539673,
+      "learning_rate": 1.1048087985171517e-05,
+      "loss": 0.2565,
+      "step": 12480
+    },
+    {
+      "epoch": 0.4667616889034278,
+      "grad_norm": 0.4036164879798889,
+      "learning_rate": 1.1042247573753555e-05,
+      "loss": 0.2744,
+      "step": 12485
+    },
+    {
+      "epoch": 0.4669486178937776,
+      "grad_norm": 0.03450818732380867,
+      "learning_rate": 1.103640680289427e-05,
+      "loss": 0.2852,
+      "step": 12490
+    },
+    {
+      "epoch": 0.4671355468841274,
+      "grad_norm": 0.39615771174430847,
+      "learning_rate": 1.1030565674607976e-05,
+      "loss": 0.2949,
+      "step": 12495
+    },
+    {
+      "epoch": 0.4673224758744772,
+      "grad_norm": 0.26349276304244995,
+      "learning_rate": 1.1024724190909109e-05,
+      "loss": 0.2381,
+      "step": 12500
+    },
+    {
+      "epoch": 0.46750940486482695,
+      "grad_norm": 0.6820989847183228,
+      "learning_rate": 1.101888235381223e-05,
+      "loss": 0.2836,
+      "step": 12505
+    },
+    {
+      "epoch": 0.4676963338551768,
+      "grad_norm": 0.30709752440452576,
+      "learning_rate": 1.1013040165332024e-05,
+      "loss": 0.3014,
+      "step": 12510
+    },
+    {
+      "epoch": 0.46788326284552656,
+      "grad_norm": 0.31798475980758667,
+      "learning_rate": 1.1007197627483292e-05,
+      "loss": 0.2755,
+      "step": 12515
+    },
+    {
+      "epoch": 0.46807019183587634,
+      "grad_norm": 0.4482247233390808,
+      "learning_rate": 1.1001354742280959e-05,
+      "loss": 0.2384,
+      "step": 12520
+    },
+    {
+      "epoch": 0.4682571208262261,
+      "grad_norm": 0.503452479839325,
+      "learning_rate": 1.0995511511740066e-05,
+      "loss": 0.3171,
+      "step": 12525
+    },
+    {
+      "epoch": 0.46844404981657595,
+      "grad_norm": 0.3593085706233978,
+      "learning_rate": 1.0989667937875778e-05,
+      "loss": 0.372,
+      "step": 12530
+    },
+    {
+      "epoch": 0.4686309788069257,
+      "grad_norm": 0.4590424597263336,
+      "learning_rate": 1.0983824022703377e-05,
+      "loss": 0.297,
+      "step": 12535
+    },
+    {
+      "epoch": 0.4688179077972755,
+      "grad_norm": 0.49093323945999146,
+      "learning_rate": 1.0977979768238261e-05,
+      "loss": 0.2479,
+      "step": 12540
+    },
+    {
+      "epoch": 0.4690048367876253,
+      "grad_norm": 0.29465728998184204,
+      "learning_rate": 1.0972135176495942e-05,
+      "loss": 0.2881,
+      "step": 12545
+    },
+    {
+      "epoch": 0.4691917657779751,
+      "grad_norm": 0.16378015279769897,
+      "learning_rate": 1.0966290249492057e-05,
+      "loss": 0.2464,
+      "step": 12550
+    },
+    {
+      "epoch": 0.4693786947683249,
+      "grad_norm": 0.3140955865383148,
+      "learning_rate": 1.0960444989242355e-05,
+      "loss": 0.1831,
+      "step": 12555
+    },
+    {
+      "epoch": 0.46956562375867467,
+      "grad_norm": 0.3941444158554077,
+      "learning_rate": 1.0954599397762695e-05,
+      "loss": 0.2958,
+      "step": 12560
+    },
+    {
+      "epoch": 0.46975255274902444,
+      "grad_norm": 0.3791537582874298,
+      "learning_rate": 1.0948753477069057e-05,
+      "loss": 0.3068,
+      "step": 12565
+    },
+    {
+      "epoch": 0.4699394817393743,
+      "grad_norm": 0.12725694477558136,
+      "learning_rate": 1.0942907229177526e-05,
+      "loss": 0.2489,
+      "step": 12570
+    },
+    {
+      "epoch": 0.47012641072972405,
+      "grad_norm": 0.6161131858825684,
+      "learning_rate": 1.0937060656104312e-05,
+      "loss": 0.3096,
+      "step": 12575
+    },
+    {
+      "epoch": 0.47031333972007383,
+      "grad_norm": 0.40583592653274536,
+      "learning_rate": 1.0931213759865729e-05,
+      "loss": 0.3163,
+      "step": 12580
+    },
+    {
+      "epoch": 0.4705002687104236,
+      "grad_norm": 3.2883963584899902,
+      "learning_rate": 1.0925366542478205e-05,
+      "loss": 0.2901,
+      "step": 12585
+    },
+    {
+      "epoch": 0.47068719770077344,
+      "grad_norm": 0.5457201600074768,
+      "learning_rate": 1.0919519005958268e-05,
+      "loss": 0.332,
+      "step": 12590
+    },
+    {
+      "epoch": 0.4708741266911232,
+      "grad_norm": 0.42853930592536926,
+      "learning_rate": 1.0913671152322578e-05,
+      "loss": 0.2741,
+      "step": 12595
+    },
+    {
+      "epoch": 0.471061055681473,
+      "grad_norm": 0.43397367000579834,
+      "learning_rate": 1.0907822983587888e-05,
+      "loss": 0.366,
+      "step": 12600
+    },
+    {
+      "epoch": 0.4712479846718228,
+      "grad_norm": 0.6656988263130188,
+      "learning_rate": 1.0901974501771065e-05,
+      "loss": 0.2831,
+      "step": 12605
+    },
+    {
+      "epoch": 0.4714349136621726,
+      "grad_norm": 0.3851011097431183,
+      "learning_rate": 1.0896125708889077e-05,
+      "loss": 0.2331,
+      "step": 12610
+    },
+    {
+      "epoch": 0.4716218426525224,
+      "grad_norm": 0.25610148906707764,
+      "learning_rate": 1.0890276606959011e-05,
+      "loss": 0.2575,
+      "step": 12615
+    },
+    {
+      "epoch": 0.47180877164287216,
+      "grad_norm": 0.4998786449432373,
+      "learning_rate": 1.0884427197998054e-05,
+      "loss": 0.2438,
+      "step": 12620
+    },
+    {
+      "epoch": 0.47199570063322194,
+      "grad_norm": 0.21021483838558197,
+      "learning_rate": 1.0878577484023496e-05,
+      "loss": 0.2544,
+      "step": 12625
+    },
+    {
+      "epoch": 0.47218262962357177,
+      "grad_norm": 0.3968941271305084,
+      "learning_rate": 1.0872727467052741e-05,
+      "loss": 0.2302,
+      "step": 12630
+    },
+    {
+      "epoch": 0.47236955861392155,
+      "grad_norm": 0.4154810905456543,
+      "learning_rate": 1.0866877149103286e-05,
+      "loss": 0.2776,
+      "step": 12635
+    },
+    {
+      "epoch": 0.4725564876042713,
+      "grad_norm": 0.4162626564502716,
+      "learning_rate": 1.0861026532192745e-05,
+      "loss": 0.2388,
+      "step": 12640
+    },
+    {
+      "epoch": 0.4727434165946211,
+      "grad_norm": 0.4286978244781494,
+      "learning_rate": 1.0855175618338823e-05,
+      "loss": 0.2645,
+      "step": 12645
+    },
+    {
+      "epoch": 0.47293034558497093,
+      "grad_norm": 0.3265318274497986,
+      "learning_rate": 1.0849324409559334e-05,
+      "loss": 0.2832,
+      "step": 12650
+    },
+    {
+      "epoch": 0.4731172745753207,
+      "grad_norm": 0.3723618686199188,
+      "learning_rate": 1.0843472907872192e-05,
+      "loss": 0.2681,
+      "step": 12655
+    },
+    {
+      "epoch": 0.4733042035656705,
+      "grad_norm": 0.47997230291366577,
+      "learning_rate": 1.0837621115295414e-05,
+      "loss": 0.4914,
+      "step": 12660
+    },
+    {
+      "epoch": 0.47349113255602027,
+      "grad_norm": 0.3690032362937927,
+      "learning_rate": 1.0831769033847113e-05,
+      "loss": 0.2536,
+      "step": 12665
+    },
+    {
+      "epoch": 0.4736780615463701,
+      "grad_norm": 0.44639065861701965,
+      "learning_rate": 1.0825916665545506e-05,
+      "loss": 0.3222,
+      "step": 12670
+    },
+    {
+      "epoch": 0.4738649905367199,
+      "grad_norm": 0.3845580816268921,
+      "learning_rate": 1.0820064012408905e-05,
+      "loss": 0.3047,
+      "step": 12675
+    },
+    {
+      "epoch": 0.47405191952706965,
+      "grad_norm": 0.5291163921356201,
+      "learning_rate": 1.0814211076455727e-05,
+      "loss": 0.2676,
+      "step": 12680
+    },
+    {
+      "epoch": 0.47423884851741943,
+      "grad_norm": 0.20399098098278046,
+      "learning_rate": 1.0808357859704478e-05,
+      "loss": 0.2618,
+      "step": 12685
+    },
+    {
+      "epoch": 0.4744257775077692,
+      "grad_norm": 0.5159106254577637,
+      "learning_rate": 1.0802504364173763e-05,
+      "loss": 0.3227,
+      "step": 12690
+    },
+    {
+      "epoch": 0.47461270649811904,
+      "grad_norm": 0.24743987619876862,
+      "learning_rate": 1.079665059188229e-05,
+      "loss": 0.2674,
+      "step": 12695
+    },
+    {
+      "epoch": 0.4747996354884688,
+      "grad_norm": 0.41947150230407715,
+      "learning_rate": 1.0790796544848853e-05,
+      "loss": 0.297,
+      "step": 12700
+    },
+    {
+      "epoch": 0.4749865644788186,
+      "grad_norm": 0.5480782389640808,
+      "learning_rate": 1.078494222509235e-05,
+      "loss": 0.2761,
+      "step": 12705
+    },
+    {
+      "epoch": 0.4751734934691684,
+      "grad_norm": 0.2428130805492401,
+      "learning_rate": 1.0779087634631763e-05,
+      "loss": 0.2324,
+      "step": 12710
+    },
+    {
+      "epoch": 0.4753604224595182,
+      "grad_norm": 0.2542586624622345,
+      "learning_rate": 1.0773232775486173e-05,
+      "loss": 0.3582,
+      "step": 12715
+    },
+    {
+      "epoch": 0.475547351449868,
+      "grad_norm": 0.341899573802948,
+      "learning_rate": 1.0767377649674755e-05,
+      "loss": 0.2975,
+      "step": 12720
+    },
+    {
+      "epoch": 0.47573428044021776,
+      "grad_norm": 0.6047893762588501,
+      "learning_rate": 1.0761522259216777e-05,
+      "loss": 0.2313,
+      "step": 12725
+    },
+    {
+      "epoch": 0.47592120943056754,
+      "grad_norm": 0.3485582768917084,
+      "learning_rate": 1.0755666606131588e-05,
+      "loss": 0.2606,
+      "step": 12730
+    },
+    {
+      "epoch": 0.47610813842091737,
+      "grad_norm": 0.38747602701187134,
+      "learning_rate": 1.074981069243864e-05,
+      "loss": 0.3315,
+      "step": 12735
+    },
+    {
+      "epoch": 0.47629506741126715,
+      "grad_norm": 0.5207949876785278,
+      "learning_rate": 1.0743954520157471e-05,
+      "loss": 0.34,
+      "step": 12740
+    },
+    {
+      "epoch": 0.4764819964016169,
+      "grad_norm": 0.6631015539169312,
+      "learning_rate": 1.0738098091307703e-05,
+      "loss": 0.2457,
+      "step": 12745
+    },
+    {
+      "epoch": 0.4766689253919667,
+      "grad_norm": 0.2515263855457306,
+      "learning_rate": 1.0732241407909057e-05,
+      "loss": 0.2445,
+      "step": 12750
+    },
+    {
+      "epoch": 0.47685585438231654,
+      "grad_norm": 0.27167659997940063,
+      "learning_rate": 1.0726384471981326e-05,
+      "loss": 0.2514,
+      "step": 12755
+    },
+    {
+      "epoch": 0.4770427833726663,
+      "grad_norm": 0.44566571712493896,
+      "learning_rate": 1.0720527285544406e-05,
+      "loss": 0.2548,
+      "step": 12760
+    },
+    {
+      "epoch": 0.4772297123630161,
+      "grad_norm": 0.5117403268814087,
+      "learning_rate": 1.071466985061827e-05,
+      "loss": 0.2884,
+      "step": 12765
+    },
+    {
+      "epoch": 0.47741664135336587,
+      "grad_norm": 0.36228641867637634,
+      "learning_rate": 1.070881216922298e-05,
+      "loss": 0.3223,
+      "step": 12770
+    },
+    {
+      "epoch": 0.4776035703437157,
+      "grad_norm": 0.35445940494537354,
+      "learning_rate": 1.0702954243378685e-05,
+      "loss": 0.2761,
+      "step": 12775
+    },
+    {
+      "epoch": 0.4777904993340655,
+      "grad_norm": 0.2719997763633728,
+      "learning_rate": 1.069709607510561e-05,
+      "loss": 0.3556,
+      "step": 12780
+    },
+    {
+      "epoch": 0.47797742832441525,
+      "grad_norm": 0.33862072229385376,
+      "learning_rate": 1.0691237666424077e-05,
+      "loss": 0.2534,
+      "step": 12785
+    },
+    {
+      "epoch": 0.47816435731476503,
+      "grad_norm": 0.2040008008480072,
+      "learning_rate": 1.0685379019354476e-05,
+      "loss": 0.2666,
+      "step": 12790
+    },
+    {
+      "epoch": 0.47835128630511486,
+      "grad_norm": 0.3832045793533325,
+      "learning_rate": 1.0679520135917293e-05,
+      "loss": 0.3088,
+      "step": 12795
+    },
+    {
+      "epoch": 0.47853821529546464,
+      "grad_norm": 0.4770766496658325,
+      "learning_rate": 1.0673661018133086e-05,
+      "loss": 0.2701,
+      "step": 12800
+    },
+    {
+      "epoch": 0.4787251442858144,
+      "grad_norm": 1.6293139457702637,
+      "learning_rate": 1.0667801668022496e-05,
+      "loss": 0.2655,
+      "step": 12805
+    },
+    {
+      "epoch": 0.4789120732761642,
+      "grad_norm": 0.44511693716049194,
+      "learning_rate": 1.0661942087606243e-05,
+      "loss": 0.354,
+      "step": 12810
+    },
+    {
+      "epoch": 0.47909900226651403,
+      "grad_norm": 0.22285427153110504,
+      "learning_rate": 1.065608227890513e-05,
+      "loss": 0.2925,
+      "step": 12815
+    },
+    {
+      "epoch": 0.4792859312568638,
+      "grad_norm": 0.34608471393585205,
+      "learning_rate": 1.0650222243940043e-05,
+      "loss": 0.245,
+      "step": 12820
+    },
+    {
+      "epoch": 0.4794728602472136,
+      "grad_norm": 0.42804983258247375,
+      "learning_rate": 1.0644361984731932e-05,
+      "loss": 0.288,
+      "step": 12825
+    },
+    {
+      "epoch": 0.47965978923756336,
+      "grad_norm": 0.4652557373046875,
+      "learning_rate": 1.0638501503301837e-05,
+      "loss": 0.3612,
+      "step": 12830
+    },
+    {
+      "epoch": 0.4798467182279132,
+      "grad_norm": 0.24128127098083496,
+      "learning_rate": 1.0632640801670868e-05,
+      "loss": 0.2583,
+      "step": 12835
+    },
+    {
+      "epoch": 0.48003364721826297,
+      "grad_norm": 0.45303890109062195,
+      "learning_rate": 1.0626779881860213e-05,
+      "loss": 0.2702,
+      "step": 12840
+    },
+    {
+      "epoch": 0.48022057620861275,
+      "grad_norm": 0.33776313066482544,
+      "learning_rate": 1.0620918745891143e-05,
+      "loss": 0.3176,
+      "step": 12845
+    },
+    {
+      "epoch": 0.4804075051989625,
+      "grad_norm": 0.304514616727829,
+      "learning_rate": 1.0615057395784983e-05,
+      "loss": 0.2317,
+      "step": 12850
+    },
+    {
+      "epoch": 0.48059443418931236,
+      "grad_norm": 0.3652326762676239,
+      "learning_rate": 1.0609195833563153e-05,
+      "loss": 0.271,
+      "step": 12855
+    },
+    {
+      "epoch": 0.48078136317966214,
+      "grad_norm": 0.397659033536911,
+      "learning_rate": 1.0603334061247133e-05,
+      "loss": 0.3553,
+      "step": 12860
+    },
+    {
+      "epoch": 0.4809682921700119,
+      "grad_norm": 0.33267614245414734,
+      "learning_rate": 1.0597472080858485e-05,
+      "loss": 0.3367,
+      "step": 12865
+    },
+    {
+      "epoch": 0.4811552211603617,
+      "grad_norm": 0.35334113240242004,
+      "learning_rate": 1.0591609894418835e-05,
+      "loss": 0.3743,
+      "step": 12870
+    },
+    {
+      "epoch": 0.4813421501507115,
+      "grad_norm": 0.7049282789230347,
+      "learning_rate": 1.0585747503949883e-05,
+      "loss": 0.3031,
+      "step": 12875
+    },
+    {
+      "epoch": 0.4815290791410613,
+      "grad_norm": 0.504492998123169,
+      "learning_rate": 1.0579884911473404e-05,
+      "loss": 0.2968,
+      "step": 12880
+    },
+    {
+      "epoch": 0.4817160081314111,
+      "grad_norm": 0.21734552085399628,
+      "learning_rate": 1.0574022119011234e-05,
+      "loss": 0.304,
+      "step": 12885
+    },
+    {
+      "epoch": 0.48190293712176085,
+      "grad_norm": 0.5968211889266968,
+      "learning_rate": 1.0568159128585283e-05,
+      "loss": 0.2402,
+      "step": 12890
+    },
+    {
+      "epoch": 0.4820898661121107,
+      "grad_norm": 0.5871456265449524,
+      "learning_rate": 1.056229594221753e-05,
+      "loss": 0.2916,
+      "step": 12895
+    },
+    {
+      "epoch": 0.48227679510246046,
+      "grad_norm": 0.4001029431819916,
+      "learning_rate": 1.0556432561930014e-05,
+      "loss": 0.2675,
+      "step": 12900
+    },
+    {
+      "epoch": 0.48246372409281024,
+      "grad_norm": 0.40792810916900635,
+      "learning_rate": 1.0550568989744852e-05,
+      "loss": 0.2857,
+      "step": 12905
+    },
+    {
+      "epoch": 0.48265065308316,
+      "grad_norm": 0.23005032539367676,
+      "learning_rate": 1.0544705227684223e-05,
+      "loss": 0.3099,
+      "step": 12910
+    },
+    {
+      "epoch": 0.48283758207350985,
+      "grad_norm": 0.4021388292312622,
+      "learning_rate": 1.053884127777037e-05,
+      "loss": 0.3509,
+      "step": 12915
+    },
+    {
+      "epoch": 0.48302451106385963,
+      "grad_norm": 0.40135905146598816,
+      "learning_rate": 1.0532977142025595e-05,
+      "loss": 0.3563,
+      "step": 12920
+    },
+    {
+      "epoch": 0.4832114400542094,
+      "grad_norm": 1.3269644975662231,
+      "learning_rate": 1.0527112822472279e-05,
+      "loss": 0.34,
+      "step": 12925
+    },
+    {
+      "epoch": 0.4833983690445592,
+      "grad_norm": 0.28477320075035095,
+      "learning_rate": 1.0521248321132853e-05,
+      "loss": 0.3246,
+      "step": 12930
+    },
+    {
+      "epoch": 0.48358529803490896,
+      "grad_norm": 0.38668492436408997,
+      "learning_rate": 1.0515383640029819e-05,
+      "loss": 0.348,
+      "step": 12935
+    },
+    {
+      "epoch": 0.4837722270252588,
+      "grad_norm": 0.5541387796401978,
+      "learning_rate": 1.0509518781185735e-05,
+      "loss": 0.2915,
+      "step": 12940
+    },
+    {
+      "epoch": 0.48395915601560857,
+      "grad_norm": 0.43491193652153015,
+      "learning_rate": 1.0503653746623221e-05,
+      "loss": 0.2804,
+      "step": 12945
+    },
+    {
+      "epoch": 0.48414608500595835,
+      "grad_norm": 0.3402020037174225,
+      "learning_rate": 1.0497788538364961e-05,
+      "loss": 0.3074,
+      "step": 12950
+    },
+    {
+      "epoch": 0.4843330139963081,
+      "grad_norm": 0.4688038229942322,
+      "learning_rate": 1.0491923158433696e-05,
+      "loss": 0.328,
+      "step": 12955
+    },
+    {
+      "epoch": 0.48451994298665796,
+      "grad_norm": 0.2162652462720871,
+      "learning_rate": 1.0486057608852236e-05,
+      "loss": 0.2978,
+      "step": 12960
+    },
+    {
+      "epoch": 0.48470687197700774,
+      "grad_norm": 0.33628934621810913,
+      "learning_rate": 1.0480191891643427e-05,
+      "loss": 0.2905,
+      "step": 12965
+    },
+    {
+      "epoch": 0.4848938009673575,
+      "grad_norm": 0.30055081844329834,
+      "learning_rate": 1.0474326008830198e-05,
+      "loss": 0.288,
+      "step": 12970
+    },
+    {
+      "epoch": 0.4850807299577073,
+      "grad_norm": 0.30387982726097107,
+      "learning_rate": 1.0468459962435517e-05,
+      "loss": 0.2517,
+      "step": 12975
+    },
+    {
+      "epoch": 0.4852676589480571,
+      "grad_norm": 0.3027805685997009,
+      "learning_rate": 1.046259375448242e-05,
+      "loss": 0.2283,
+      "step": 12980
+    },
+    {
+      "epoch": 0.4854545879384069,
+      "grad_norm": 0.6029101014137268,
+      "learning_rate": 1.0456727386993992e-05,
+      "loss": 0.313,
+      "step": 12985
+    },
+    {
+      "epoch": 0.4856415169287567,
+      "grad_norm": 0.23959915339946747,
+      "learning_rate": 1.0450860861993374e-05,
+      "loss": 0.2493,
+      "step": 12990
+    },
+    {
+      "epoch": 0.48582844591910646,
+      "grad_norm": 0.6326263546943665,
+      "learning_rate": 1.0444994181503764e-05,
+      "loss": 0.31,
+      "step": 12995
+    },
+    {
+      "epoch": 0.4860153749094563,
+      "grad_norm": 0.3467031419277191,
+      "learning_rate": 1.043912734754841e-05,
+      "loss": 0.2751,
+      "step": 13000
+    },
+    {
+      "epoch": 0.48620230389980607,
+      "grad_norm": 0.24758736789226532,
+      "learning_rate": 1.0433260362150618e-05,
+      "loss": 0.3156,
+      "step": 13005
+    },
+    {
+      "epoch": 0.48638923289015584,
+      "grad_norm": 0.4706723093986511,
+      "learning_rate": 1.0427393227333742e-05,
+      "loss": 0.2606,
+      "step": 13010
+    },
+    {
+      "epoch": 0.4865761618805056,
+      "grad_norm": 0.4599349796772003,
+      "learning_rate": 1.0421525945121187e-05,
+      "loss": 0.257,
+      "step": 13015
+    },
+    {
+      "epoch": 0.48676309087085545,
+      "grad_norm": 0.3387102484703064,
+      "learning_rate": 1.0415658517536414e-05,
+      "loss": 0.2991,
+      "step": 13020
+    },
+    {
+      "epoch": 0.48695001986120523,
+      "grad_norm": 0.4700387120246887,
+      "learning_rate": 1.0409790946602926e-05,
+      "loss": 0.2578,
+      "step": 13025
+    },
+    {
+      "epoch": 0.487136948851555,
+      "grad_norm": 0.5611972808837891,
+      "learning_rate": 1.0403923234344282e-05,
+      "loss": 0.3931,
+      "step": 13030
+    },
+    {
+      "epoch": 0.4873238778419048,
+      "grad_norm": 0.3460509479045868,
+      "learning_rate": 1.0398055382784094e-05,
+      "loss": 0.2514,
+      "step": 13035
+    },
+    {
+      "epoch": 0.4875108068322546,
+      "grad_norm": 0.45094433426856995,
+      "learning_rate": 1.0392187393946004e-05,
+      "loss": 0.2802,
+      "step": 13040
+    },
+    {
+      "epoch": 0.4876977358226044,
+      "grad_norm": 0.44121870398521423,
+      "learning_rate": 1.0386319269853719e-05,
+      "loss": 0.3061,
+      "step": 13045
+    },
+    {
+      "epoch": 0.48788466481295417,
+      "grad_norm": 0.35361143946647644,
+      "learning_rate": 1.038045101253099e-05,
+      "loss": 0.2768,
+      "step": 13050
+    },
+    {
+      "epoch": 0.48807159380330395,
+      "grad_norm": 1.3324558734893799,
+      "learning_rate": 1.0374582624001608e-05,
+      "loss": 0.3261,
+      "step": 13055
+    },
+    {
+      "epoch": 0.4882585227936538,
+      "grad_norm": 0.31909024715423584,
+      "learning_rate": 1.0368714106289412e-05,
+      "loss": 0.2433,
+      "step": 13060
+    },
+    {
+      "epoch": 0.48844545178400356,
+      "grad_norm": 0.8846080899238586,
+      "learning_rate": 1.0362845461418286e-05,
+      "loss": 0.3087,
+      "step": 13065
+    },
+    {
+      "epoch": 0.48863238077435334,
+      "grad_norm": 0.2497347593307495,
+      "learning_rate": 1.0356976691412156e-05,
+      "loss": 0.2936,
+      "step": 13070
+    },
+    {
+      "epoch": 0.4888193097647031,
+      "grad_norm": 0.33820840716362,
+      "learning_rate": 1.0351107798294994e-05,
+      "loss": 0.3137,
+      "step": 13075
+    },
+    {
+      "epoch": 0.48900623875505295,
+      "grad_norm": 0.23961906135082245,
+      "learning_rate": 1.0345238784090816e-05,
+      "loss": 0.2316,
+      "step": 13080
+    },
+    {
+      "epoch": 0.4891931677454027,
+      "grad_norm": 0.172140970826149,
+      "learning_rate": 1.0339369650823672e-05,
+      "loss": 0.3032,
+      "step": 13085
+    },
+    {
+      "epoch": 0.4893800967357525,
+      "grad_norm": 0.21862946450710297,
+      "learning_rate": 1.0333500400517656e-05,
+      "loss": 0.2501,
+      "step": 13090
+    },
+    {
+      "epoch": 0.4895670257261023,
+      "grad_norm": 0.35493218898773193,
+      "learning_rate": 1.032763103519691e-05,
+      "loss": 0.2768,
+      "step": 13095
+    },
+    {
+      "epoch": 0.4897539547164521,
+      "grad_norm": 0.2946685254573822,
+      "learning_rate": 1.0321761556885608e-05,
+      "loss": 0.3169,
+      "step": 13100
+    },
+    {
+      "epoch": 0.4899408837068019,
+      "grad_norm": 0.39835453033447266,
+      "learning_rate": 1.0315891967607968e-05,
+      "loss": 0.3315,
+      "step": 13105
+    },
+    {
+      "epoch": 0.49012781269715167,
+      "grad_norm": 0.4330670237541199,
+      "learning_rate": 1.0310022269388236e-05,
+      "loss": 0.3063,
+      "step": 13110
+    },
+    {
+      "epoch": 0.49031474168750144,
+      "grad_norm": 0.3474862277507782,
+      "learning_rate": 1.0304152464250707e-05,
+      "loss": 0.3044,
+      "step": 13115
+    },
+    {
+      "epoch": 0.4905016706778513,
+      "grad_norm": 0.3377613425254822,
+      "learning_rate": 1.0298282554219707e-05,
+      "loss": 0.2859,
+      "step": 13120
+    },
+    {
+      "epoch": 0.49068859966820105,
+      "grad_norm": 0.2067660242319107,
+      "learning_rate": 1.0292412541319603e-05,
+      "loss": 0.2821,
+      "step": 13125
+    },
+    {
+      "epoch": 0.49087552865855083,
+      "grad_norm": 0.5813116431236267,
+      "learning_rate": 1.0286542427574794e-05,
+      "loss": 0.3219,
+      "step": 13130
+    },
+    {
+      "epoch": 0.4910624576489006,
+      "grad_norm": 0.5601276159286499,
+      "learning_rate": 1.0280672215009706e-05,
+      "loss": 0.2605,
+      "step": 13135
+    },
+    {
+      "epoch": 0.49124938663925044,
+      "grad_norm": 0.35725921392440796,
+      "learning_rate": 1.0274801905648816e-05,
+      "loss": 0.2353,
+      "step": 13140
+    },
+    {
+      "epoch": 0.4914363156296002,
+      "grad_norm": 0.3625008165836334,
+      "learning_rate": 1.0268931501516626e-05,
+      "loss": 0.3755,
+      "step": 13145
+    },
+    {
+      "epoch": 0.49162324461995,
+      "grad_norm": 0.2937968671321869,
+      "learning_rate": 1.0263061004637666e-05,
+      "loss": 0.2857,
+      "step": 13150
+    },
+    {
+      "epoch": 0.4918101736102998,
+      "grad_norm": 0.48958075046539307,
+      "learning_rate": 1.0257190417036502e-05,
+      "loss": 0.289,
+      "step": 13155
+    },
+    {
+      "epoch": 0.4919971026006496,
+      "grad_norm": 0.4188752770423889,
+      "learning_rate": 1.0251319740737732e-05,
+      "loss": 0.2968,
+      "step": 13160
+    },
+    {
+      "epoch": 0.4921840315909994,
+      "grad_norm": 0.32891568541526794,
+      "learning_rate": 1.0245448977765986e-05,
+      "loss": 0.299,
+      "step": 13165
+    },
+    {
+      "epoch": 0.49237096058134916,
+      "grad_norm": 0.4842504858970642,
+      "learning_rate": 1.023957813014592e-05,
+      "loss": 0.212,
+      "step": 13170
+    },
+    {
+      "epoch": 0.49255788957169894,
+      "grad_norm": 0.5407660007476807,
+      "learning_rate": 1.0233707199902223e-05,
+      "loss": 0.24,
+      "step": 13175
+    },
+    {
+      "epoch": 0.4927448185620487,
+      "grad_norm": 0.268076092004776,
+      "learning_rate": 1.0227836189059606e-05,
+      "loss": 0.229,
+      "step": 13180
+    },
+    {
+      "epoch": 0.49293174755239855,
+      "grad_norm": 0.6369917392730713,
+      "learning_rate": 1.0221965099642817e-05,
+      "loss": 0.2855,
+      "step": 13185
+    },
+    {
+      "epoch": 0.4931186765427483,
+      "grad_norm": 0.39120930433273315,
+      "learning_rate": 1.0216093933676625e-05,
+      "loss": 0.2355,
+      "step": 13190
+    },
+    {
+      "epoch": 0.4933056055330981,
+      "grad_norm": 0.31261909008026123,
+      "learning_rate": 1.0210222693185829e-05,
+      "loss": 0.2723,
+      "step": 13195
+    },
+    {
+      "epoch": 0.4934925345234479,
+      "grad_norm": 0.5893165469169617,
+      "learning_rate": 1.0204351380195249e-05,
+      "loss": 0.2922,
+      "step": 13200
+    },
+    {
+      "epoch": 0.4936794635137977,
+      "grad_norm": 0.4819238781929016,
+      "learning_rate": 1.0198479996729736e-05,
+      "loss": 0.3932,
+      "step": 13205
+    },
+    {
+      "epoch": 0.4938663925041475,
+      "grad_norm": 0.35270988941192627,
+      "learning_rate": 1.0192608544814155e-05,
+      "loss": 0.35,
+      "step": 13210
+    },
+    {
+      "epoch": 0.49405332149449727,
+      "grad_norm": 0.6773056387901306,
+      "learning_rate": 1.0186737026473408e-05,
+      "loss": 0.3297,
+      "step": 13215
+    },
+    {
+      "epoch": 0.49424025048484704,
+      "grad_norm": 0.28301501274108887,
+      "learning_rate": 1.0180865443732408e-05,
+      "loss": 0.3117,
+      "step": 13220
+    },
+    {
+      "epoch": 0.4944271794751969,
+      "grad_norm": 0.3076423406600952,
+      "learning_rate": 1.0174993798616101e-05,
+      "loss": 0.2898,
+      "step": 13225
+    },
+    {
+      "epoch": 0.49461410846554665,
+      "grad_norm": 0.6145200133323669,
+      "learning_rate": 1.0169122093149449e-05,
+      "loss": 0.271,
+      "step": 13230
+    },
+    {
+      "epoch": 0.49480103745589643,
+      "grad_norm": 0.6317368745803833,
+      "learning_rate": 1.016325032935743e-05,
+      "loss": 0.3003,
+      "step": 13235
+    },
+    {
+      "epoch": 0.4949879664462462,
+      "grad_norm": 0.5369362235069275,
+      "learning_rate": 1.0157378509265053e-05,
+      "loss": 0.2718,
+      "step": 13240
+    },
+    {
+      "epoch": 0.49517489543659604,
+      "grad_norm": 0.3894631564617157,
+      "learning_rate": 1.015150663489734e-05,
+      "loss": 0.3061,
+      "step": 13245
+    },
+    {
+      "epoch": 0.4953618244269458,
+      "grad_norm": 0.35104823112487793,
+      "learning_rate": 1.0145634708279324e-05,
+      "loss": 0.3223,
+      "step": 13250
+    },
+    {
+      "epoch": 0.4955487534172956,
+      "grad_norm": 0.5245153903961182,
+      "learning_rate": 1.013976273143607e-05,
+      "loss": 0.3233,
+      "step": 13255
+    },
+    {
+      "epoch": 0.4957356824076454,
+      "grad_norm": 0.4358454942703247,
+      "learning_rate": 1.013389070639266e-05,
+      "loss": 0.344,
+      "step": 13260
+    },
+    {
+      "epoch": 0.4959226113979952,
+      "grad_norm": 0.3596687912940979,
+      "learning_rate": 1.0128018635174177e-05,
+      "loss": 0.2782,
+      "step": 13265
+    },
+    {
+      "epoch": 0.496109540388345,
+      "grad_norm": 0.314059317111969,
+      "learning_rate": 1.0122146519805736e-05,
+      "loss": 0.27,
+      "step": 13270
+    },
+    {
+      "epoch": 0.49629646937869476,
+      "grad_norm": 0.49434518814086914,
+      "learning_rate": 1.0116274362312462e-05,
+      "loss": 0.2295,
+      "step": 13275
+    },
+    {
+      "epoch": 0.49648339836904454,
+      "grad_norm": 0.42047810554504395,
+      "learning_rate": 1.011040216471949e-05,
+      "loss": 0.2461,
+      "step": 13280
+    },
+    {
+      "epoch": 0.49667032735939437,
+      "grad_norm": 0.44239863753318787,
+      "learning_rate": 1.0104529929051977e-05,
+      "loss": 0.3447,
+      "step": 13285
+    },
+    {
+      "epoch": 0.49685725634974415,
+      "grad_norm": 0.4856170117855072,
+      "learning_rate": 1.0098657657335083e-05,
+      "loss": 0.3261,
+      "step": 13290
+    },
+    {
+      "epoch": 0.4970441853400939,
+      "grad_norm": 0.7703909277915955,
+      "learning_rate": 1.0092785351593995e-05,
+      "loss": 0.2374,
+      "step": 13295
+    },
+    {
+      "epoch": 0.4972311143304437,
+      "grad_norm": 0.44741690158843994,
+      "learning_rate": 1.0086913013853894e-05,
+      "loss": 0.2753,
+      "step": 13300
+    },
+    {
+      "epoch": 0.49741804332079353,
+      "grad_norm": 0.38895756006240845,
+      "learning_rate": 1.0081040646139985e-05,
+      "loss": 0.2716,
+      "step": 13305
+    },
+    {
+      "epoch": 0.4976049723111433,
+      "grad_norm": 0.7058769464492798,
+      "learning_rate": 1.0075168250477482e-05,
+      "loss": 0.3592,
+      "step": 13310
+    },
+    {
+      "epoch": 0.4977919013014931,
+      "grad_norm": 0.34860917925834656,
+      "learning_rate": 1.00692958288916e-05,
+      "loss": 0.3461,
+      "step": 13315
+    },
+    {
+      "epoch": 0.49797883029184287,
+      "grad_norm": 0.3462105691432953,
+      "learning_rate": 1.0063423383407575e-05,
+      "loss": 0.2995,
+      "step": 13320
+    },
+    {
+      "epoch": 0.4981657592821927,
+      "grad_norm": 0.351172536611557,
+      "learning_rate": 1.005755091605064e-05,
+      "loss": 0.4439,
+      "step": 13325
+    },
+    {
+      "epoch": 0.4983526882725425,
+      "grad_norm": 0.16182276606559753,
+      "learning_rate": 1.0051678428846046e-05,
+      "loss": 0.2951,
+      "step": 13330
+    },
+    {
+      "epoch": 0.49853961726289225,
+      "grad_norm": 0.41749125719070435,
+      "learning_rate": 1.0045805923819039e-05,
+      "loss": 0.2411,
+      "step": 13335
+    },
+    {
+      "epoch": 0.49872654625324203,
+      "grad_norm": 0.3556321859359741,
+      "learning_rate": 1.0039933402994885e-05,
+      "loss": 0.2667,
+      "step": 13340
+    },
+    {
+      "epoch": 0.49891347524359186,
+      "grad_norm": 0.5114247798919678,
+      "learning_rate": 1.0034060868398843e-05,
+      "loss": 0.248,
+      "step": 13345
+    },
+    {
+      "epoch": 0.49910040423394164,
+      "grad_norm": 0.29534754157066345,
+      "learning_rate": 1.0028188322056183e-05,
+      "loss": 0.2799,
+      "step": 13350
+    },
+    {
+      "epoch": 0.4992873332242914,
+      "grad_norm": 0.5048911571502686,
+      "learning_rate": 1.0022315765992179e-05,
+      "loss": 0.2801,
+      "step": 13355
+    },
+    {
+      "epoch": 0.4994742622146412,
+      "grad_norm": 0.3704678416252136,
+      "learning_rate": 1.0016443202232107e-05,
+      "loss": 0.2554,
+      "step": 13360
+    },
+    {
+      "epoch": 0.49966119120499103,
+      "grad_norm": 0.4027446210384369,
+      "learning_rate": 1.0010570632801244e-05,
+      "loss": 0.2369,
+      "step": 13365
+    },
+    {
+      "epoch": 0.4998481201953408,
+      "grad_norm": 0.4065941572189331,
+      "learning_rate": 1.0004698059724873e-05,
+      "loss": 0.3427,
+      "step": 13370
+    },
+    {
+      "epoch": 0.5000350491856906,
+      "grad_norm": 0.1492055058479309,
+      "learning_rate": 9.998825485028277e-06,
+      "loss": 0.3667,
+      "step": 13375
+    },
+    {
+      "epoch": 0.5002219781760404,
+      "grad_norm": 0.7147864103317261,
+      "learning_rate": 9.99295291073674e-06,
+      "loss": 0.3257,
+      "step": 13380
+    },
+    {
+      "epoch": 0.5004089071663902,
+      "grad_norm": 0.355305552482605,
+      "learning_rate": 9.987080338875537e-06,
+      "loss": 0.384,
+      "step": 13385
+    },
+    {
+      "epoch": 0.50059583615674,
+      "grad_norm": 0.24693630635738373,
+      "learning_rate": 9.981207771469956e-06,
+      "loss": 0.3332,
+      "step": 13390
+    },
+    {
+      "epoch": 0.5007827651470897,
+      "grad_norm": 0.4253278076648712,
+      "learning_rate": 9.975335210545279e-06,
+      "loss": 0.2696,
+      "step": 13395
+    },
+    {
+      "epoch": 0.5009696941374395,
+      "grad_norm": 0.22588220238685608,
+      "learning_rate": 9.969462658126778e-06,
+      "loss": 0.3232,
+      "step": 13400
+    },
+    {
+      "epoch": 0.5011566231277893,
+      "grad_norm": 0.6063287854194641,
+      "learning_rate": 9.963590116239734e-06,
+      "loss": 0.2805,
+      "step": 13405
+    },
+    {
+      "epoch": 0.5013435521181391,
+      "grad_norm": 0.5377234816551208,
+      "learning_rate": 9.957717586909415e-06,
+      "loss": 0.3082,
+      "step": 13410
+    },
+    {
+      "epoch": 0.501530481108489,
+      "grad_norm": 0.6075440645217896,
+      "learning_rate": 9.95184507216109e-06,
+      "loss": 0.2732,
+      "step": 13415
+    },
+    {
+      "epoch": 0.5017174100988387,
+      "grad_norm": 0.6097599864006042,
+      "learning_rate": 9.945972574020015e-06,
+      "loss": 0.3988,
+      "step": 13420
+    },
+    {
+      "epoch": 0.5019043390891885,
+      "grad_norm": 0.3699340224266052,
+      "learning_rate": 9.940100094511457e-06,
+      "loss": 0.3371,
+      "step": 13425
+    },
+    {
+      "epoch": 0.5020912680795383,
+      "grad_norm": 0.30382204055786133,
+      "learning_rate": 9.934227635660654e-06,
+      "loss": 0.2644,
+      "step": 13430
+    },
+    {
+      "epoch": 0.5022781970698881,
+      "grad_norm": 0.38782837986946106,
+      "learning_rate": 9.928355199492859e-06,
+      "loss": 0.2467,
+      "step": 13435
+    },
+    {
+      "epoch": 0.5024651260602379,
+      "grad_norm": 0.22586065530776978,
+      "learning_rate": 9.922482788033304e-06,
+      "loss": 0.2594,
+      "step": 13440
+    },
+    {
+      "epoch": 0.5026520550505876,
+      "grad_norm": 0.38370853662490845,
+      "learning_rate": 9.91661040330721e-06,
+      "loss": 0.2202,
+      "step": 13445
+    },
+    {
+      "epoch": 0.5028389840409374,
+      "grad_norm": 0.3413105010986328,
+      "learning_rate": 9.910738047339801e-06,
+      "loss": 0.2385,
+      "step": 13450
+    },
+    {
+      "epoch": 0.5030259130312872,
+      "grad_norm": 0.5473100543022156,
+      "learning_rate": 9.90486572215628e-06,
+      "loss": 0.2837,
+      "step": 13455
+    },
+    {
+      "epoch": 0.5032128420216371,
+      "grad_norm": 0.5079526305198669,
+      "learning_rate": 9.898993429781848e-06,
+      "loss": 0.328,
+      "step": 13460
+    },
+    {
+      "epoch": 0.5033997710119869,
+      "grad_norm": 0.23653025925159454,
+      "learning_rate": 9.893121172241686e-06,
+      "loss": 0.2817,
+      "step": 13465
+    },
+    {
+      "epoch": 0.5035867000023366,
+      "grad_norm": 0.3025575280189514,
+      "learning_rate": 9.887248951560972e-06,
+      "loss": 0.3381,
+      "step": 13470
+    },
+    {
+      "epoch": 0.5037736289926864,
+      "grad_norm": 0.3230903148651123,
+      "learning_rate": 9.88137676976486e-06,
+      "loss": 0.2964,
+      "step": 13475
+    },
+    {
+      "epoch": 0.5039605579830362,
+      "grad_norm": 0.8575633764266968,
+      "learning_rate": 9.875504628878502e-06,
+      "loss": 0.3539,
+      "step": 13480
+    },
+    {
+      "epoch": 0.504147486973386,
+      "grad_norm": 0.5326083302497864,
+      "learning_rate": 9.869632530927033e-06,
+      "loss": 0.3046,
+      "step": 13485
+    },
+    {
+      "epoch": 0.5043344159637357,
+      "grad_norm": 0.5689330697059631,
+      "learning_rate": 9.863760477935565e-06,
+      "loss": 0.2426,
+      "step": 13490
+    },
+    {
+      "epoch": 0.5045213449540855,
+      "grad_norm": 0.4615536034107208,
+      "learning_rate": 9.857888471929207e-06,
+      "loss": 0.3155,
+      "step": 13495
+    },
+    {
+      "epoch": 0.5047082739444354,
+      "grad_norm": 0.284390926361084,
+      "learning_rate": 9.85201651493304e-06,
+      "loss": 0.2493,
+      "step": 13500
+    },
+    {
+      "epoch": 0.5048952029347852,
+      "grad_norm": 0.382242351770401,
+      "learning_rate": 9.846144608972141e-06,
+      "loss": 0.3991,
+      "step": 13505
+    },
+    {
+      "epoch": 0.505082131925135,
+      "grad_norm": 0.3869224190711975,
+      "learning_rate": 9.840272756071556e-06,
+      "loss": 0.2395,
+      "step": 13510
+    },
+    {
+      "epoch": 0.5052690609154847,
+      "grad_norm": 0.2770150601863861,
+      "learning_rate": 9.834400958256322e-06,
+      "loss": 0.2365,
+      "step": 13515
+    },
+    {
+      "epoch": 0.5054559899058345,
+      "grad_norm": 0.1850321739912033,
+      "learning_rate": 9.828529217551448e-06,
+      "loss": 0.4006,
+      "step": 13520
+    },
+    {
+      "epoch": 0.5056429188961843,
+      "grad_norm": 0.5925412178039551,
+      "learning_rate": 9.822657535981936e-06,
+      "loss": 0.3216,
+      "step": 13525
+    },
+    {
+      "epoch": 0.5058298478865341,
+      "grad_norm": 0.4952634572982788,
+      "learning_rate": 9.816785915572762e-06,
+      "loss": 0.3249,
+      "step": 13530
+    },
+    {
+      "epoch": 0.5060167768768838,
+      "grad_norm": 0.4701783061027527,
+      "learning_rate": 9.81091435834887e-06,
+      "loss": 0.2652,
+      "step": 13535
+    },
+    {
+      "epoch": 0.5062037058672337,
+      "grad_norm": 0.8160406351089478,
+      "learning_rate": 9.805042866335202e-06,
+      "loss": 0.2646,
+      "step": 13540
+    },
+    {
+      "epoch": 0.5063906348575835,
+      "grad_norm": 0.32033684849739075,
+      "learning_rate": 9.79917144155666e-06,
+      "loss": 0.3069,
+      "step": 13545
+    },
+    {
+      "epoch": 0.5065775638479333,
+      "grad_norm": 0.39233770966529846,
+      "learning_rate": 9.793300086038137e-06,
+      "loss": 0.2425,
+      "step": 13550
+    },
+    {
+      "epoch": 0.5067644928382831,
+      "grad_norm": 0.4731293022632599,
+      "learning_rate": 9.78742880180449e-06,
+      "loss": 0.2729,
+      "step": 13555
+    },
+    {
+      "epoch": 0.5069514218286328,
+      "grad_norm": 0.45634377002716064,
+      "learning_rate": 9.781557590880559e-06,
+      "loss": 0.2657,
+      "step": 13560
+    },
+    {
+      "epoch": 0.5071383508189826,
+      "grad_norm": 0.22461484372615814,
+      "learning_rate": 9.775686455291153e-06,
+      "loss": 0.3041,
+      "step": 13565
+    },
+    {
+      "epoch": 0.5073252798093324,
+      "grad_norm": 0.19289721548557281,
+      "learning_rate": 9.769815397061062e-06,
+      "loss": 0.2881,
+      "step": 13570
+    },
+    {
+      "epoch": 0.5075122087996822,
+      "grad_norm": 0.5333814024925232,
+      "learning_rate": 9.763944418215047e-06,
+      "loss": 0.3384,
+      "step": 13575
+    },
+    {
+      "epoch": 0.5076991377900321,
+      "grad_norm": 0.3757210373878479,
+      "learning_rate": 9.758073520777837e-06,
+      "loss": 0.2283,
+      "step": 13580
+    },
+    {
+      "epoch": 0.5078860667803818,
+      "grad_norm": 0.2797081470489502,
+      "learning_rate": 9.75220270677414e-06,
+      "loss": 0.2902,
+      "step": 13585
+    },
+    {
+      "epoch": 0.5080729957707316,
+      "grad_norm": 0.22128711640834808,
+      "learning_rate": 9.746331978228623e-06,
+      "loss": 0.2467,
+      "step": 13590
+    },
+    {
+      "epoch": 0.5082599247610814,
+      "grad_norm": 0.6138377785682678,
+      "learning_rate": 9.740461337165945e-06,
+      "loss": 0.2697,
+      "step": 13595
+    },
+    {
+      "epoch": 0.5084468537514312,
+      "grad_norm": 0.43696799874305725,
+      "learning_rate": 9.734590785610713e-06,
+      "loss": 0.2734,
+      "step": 13600
+    },
+    {
+      "epoch": 0.508633782741781,
+      "grad_norm": 0.38263434171676636,
+      "learning_rate": 9.728720325587515e-06,
+      "loss": 0.2258,
+      "step": 13605
+    },
+    {
+      "epoch": 0.5088207117321307,
+      "grad_norm": 0.634128749370575,
+      "learning_rate": 9.722849959120899e-06,
+      "loss": 0.3134,
+      "step": 13610
+    },
+    {
+      "epoch": 0.5090076407224805,
+      "grad_norm": 0.3166172504425049,
+      "learning_rate": 9.716979688235392e-06,
+      "loss": 0.309,
+      "step": 13615
+    },
+    {
+      "epoch": 0.5091945697128304,
+      "grad_norm": 0.29206401109695435,
+      "learning_rate": 9.711109514955485e-06,
+      "loss": 0.2797,
+      "step": 13620
+    },
+    {
+      "epoch": 0.5093814987031802,
+      "grad_norm": 0.648830235004425,
+      "learning_rate": 9.705239441305626e-06,
+      "loss": 0.2557,
+      "step": 13625
+    },
+    {
+      "epoch": 0.50956842769353,
+      "grad_norm": 0.601845383644104,
+      "learning_rate": 9.699369469310238e-06,
+      "loss": 0.3296,
+      "step": 13630
+    },
+    {
+      "epoch": 0.5097553566838797,
+      "grad_norm": 0.20830629765987396,
+      "learning_rate": 9.693499600993705e-06,
+      "loss": 0.2889,
+      "step": 13635
+    },
+    {
+      "epoch": 0.5099422856742295,
+      "grad_norm": 0.1208687499165535,
+      "learning_rate": 9.68762983838038e-06,
+      "loss": 0.2389,
+      "step": 13640
+    },
+    {
+      "epoch": 0.5101292146645793,
+      "grad_norm": 0.38630300760269165,
+      "learning_rate": 9.681760183494568e-06,
+      "loss": 0.3048,
+      "step": 13645
+    },
+    {
+      "epoch": 0.510316143654929,
+      "grad_norm": 0.2682333290576935,
+      "learning_rate": 9.675890638360556e-06,
+      "loss": 0.3013,
+      "step": 13650
+    },
+    {
+      "epoch": 0.5105030726452788,
+      "grad_norm": 0.46546033024787903,
+      "learning_rate": 9.670021205002573e-06,
+      "loss": 0.3102,
+      "step": 13655
+    },
+    {
+      "epoch": 0.5106900016356287,
+      "grad_norm": 0.2961997985839844,
+      "learning_rate": 9.66415188544482e-06,
+      "loss": 0.2771,
+      "step": 13660
+    },
+    {
+      "epoch": 0.5108769306259785,
+      "grad_norm": 0.4869229197502136,
+      "learning_rate": 9.65828268171146e-06,
+      "loss": 0.292,
+      "step": 13665
+    },
+    {
+      "epoch": 0.5110638596163283,
+      "grad_norm": 0.42666685581207275,
+      "learning_rate": 9.652413595826612e-06,
+      "loss": 0.3738,
+      "step": 13670
+    },
+    {
+      "epoch": 0.511250788606678,
+      "grad_norm": 0.5058145523071289,
+      "learning_rate": 9.646544629814357e-06,
+      "loss": 0.3383,
+      "step": 13675
+    },
+    {
+      "epoch": 0.5114377175970278,
+      "grad_norm": 0.4218425154685974,
+      "learning_rate": 9.640675785698726e-06,
+      "loss": 0.2804,
+      "step": 13680
+    },
+    {
+      "epoch": 0.5116246465873776,
+      "grad_norm": 0.5010436177253723,
+      "learning_rate": 9.634807065503726e-06,
+      "loss": 0.2663,
+      "step": 13685
+    },
+    {
+      "epoch": 0.5118115755777274,
+      "grad_norm": 0.44729965925216675,
+      "learning_rate": 9.628938471253302e-06,
+      "loss": 0.347,
+      "step": 13690
+    },
+    {
+      "epoch": 0.5119985045680772,
+      "grad_norm": 0.4109884202480316,
+      "learning_rate": 9.62307000497137e-06,
+      "loss": 0.2371,
+      "step": 13695
+    },
+    {
+      "epoch": 0.5121854335584269,
+      "grad_norm": 0.35468733310699463,
+      "learning_rate": 9.61720166868179e-06,
+      "loss": 0.2445,
+      "step": 13700
+    },
+    {
+      "epoch": 0.5123723625487768,
+      "grad_norm": 0.31252321600914,
+      "learning_rate": 9.611333464408383e-06,
+      "loss": 0.3187,
+      "step": 13705
+    },
+    {
+      "epoch": 0.5125592915391266,
+      "grad_norm": 0.31071966886520386,
+      "learning_rate": 9.605465394174933e-06,
+      "loss": 0.235,
+      "step": 13710
+    },
+    {
+      "epoch": 0.5127462205294764,
+      "grad_norm": 0.4738657474517822,
+      "learning_rate": 9.599597460005161e-06,
+      "loss": 0.2461,
+      "step": 13715
+    },
+    {
+      "epoch": 0.5129331495198262,
+      "grad_norm": 0.35220175981521606,
+      "learning_rate": 9.593729663922752e-06,
+      "loss": 0.3399,
+      "step": 13720
+    },
+    {
+      "epoch": 0.5131200785101759,
+      "grad_norm": 0.3583666682243347,
+      "learning_rate": 9.587862007951343e-06,
+      "loss": 0.307,
+      "step": 13725
+    },
+    {
+      "epoch": 0.5133070075005257,
+      "grad_norm": 0.5065263509750366,
+      "learning_rate": 9.581994494114518e-06,
+      "loss": 0.3031,
+      "step": 13730
+    },
+    {
+      "epoch": 0.5134939364908755,
+      "grad_norm": 0.5432529449462891,
+      "learning_rate": 9.576127124435811e-06,
+      "loss": 0.2919,
+      "step": 13735
+    },
+    {
+      "epoch": 0.5136808654812253,
+      "grad_norm": 0.36350008845329285,
+      "learning_rate": 9.570259900938717e-06,
+      "loss": 0.284,
+      "step": 13740
+    },
+    {
+      "epoch": 0.5138677944715752,
+      "grad_norm": 0.3522621691226959,
+      "learning_rate": 9.564392825646669e-06,
+      "loss": 0.2702,
+      "step": 13745
+    },
+    {
+      "epoch": 0.5140547234619249,
+      "grad_norm": 0.24228136241436005,
+      "learning_rate": 9.55852590058305e-06,
+      "loss": 0.2595,
+      "step": 13750
+    },
+    {
+      "epoch": 0.5142416524522747,
+      "grad_norm": 0.40059664845466614,
+      "learning_rate": 9.552659127771204e-06,
+      "loss": 0.3595,
+      "step": 13755
+    },
+    {
+      "epoch": 0.5144285814426245,
+      "grad_norm": 0.3901945948600769,
+      "learning_rate": 9.5467925092344e-06,
+      "loss": 0.2465,
+      "step": 13760
+    },
+    {
+      "epoch": 0.5146155104329743,
+      "grad_norm": 0.4316410422325134,
+      "learning_rate": 9.54092604699588e-06,
+      "loss": 0.3,
+      "step": 13765
+    },
+    {
+      "epoch": 0.514802439423324,
+      "grad_norm": 0.2531914710998535,
+      "learning_rate": 9.53505974307881e-06,
+      "loss": 0.275,
+      "step": 13770
+    },
+    {
+      "epoch": 0.5149893684136738,
+      "grad_norm": 1.1252632141113281,
+      "learning_rate": 9.529193599506313e-06,
+      "loss": 0.3713,
+      "step": 13775
+    },
+    {
+      "epoch": 0.5151762974040236,
+      "grad_norm": 0.3661787807941437,
+      "learning_rate": 9.52332761830145e-06,
+      "loss": 0.2687,
+      "step": 13780
+    },
+    {
+      "epoch": 0.5153632263943735,
+      "grad_norm": 0.234545037150383,
+      "learning_rate": 9.517461801487239e-06,
+      "loss": 0.3,
+      "step": 13785
+    },
+    {
+      "epoch": 0.5155501553847233,
+      "grad_norm": 0.29935866594314575,
+      "learning_rate": 9.51159615108662e-06,
+      "loss": 0.3166,
+      "step": 13790
+    },
+    {
+      "epoch": 0.515737084375073,
+      "grad_norm": 0.5030335783958435,
+      "learning_rate": 9.505730669122494e-06,
+      "loss": 0.3111,
+      "step": 13795
+    },
+    {
+      "epoch": 0.5159240133654228,
+      "grad_norm": 0.33726051449775696,
+      "learning_rate": 9.499865357617703e-06,
+      "loss": 0.2607,
+      "step": 13800
+    },
+    {
+      "epoch": 0.5161109423557726,
+      "grad_norm": 0.4101554751396179,
+      "learning_rate": 9.494000218595015e-06,
+      "loss": 0.297,
+      "step": 13805
+    },
+    {
+      "epoch": 0.5162978713461224,
+      "grad_norm": 0.3463868796825409,
+      "learning_rate": 9.488135254077155e-06,
+      "loss": 0.2925,
+      "step": 13810
+    },
+    {
+      "epoch": 0.5164848003364721,
+      "grad_norm": 0.15302151441574097,
+      "learning_rate": 9.482270466086778e-06,
+      "loss": 0.3581,
+      "step": 13815
+    },
+    {
+      "epoch": 0.5166717293268219,
+      "grad_norm": 0.2611302435398102,
+      "learning_rate": 9.476405856646485e-06,
+      "loss": 0.2782,
+      "step": 13820
+    },
+    {
+      "epoch": 0.5168586583171718,
+      "grad_norm": 0.6067391633987427,
+      "learning_rate": 9.470541427778805e-06,
+      "loss": 0.2788,
+      "step": 13825
+    },
+    {
+      "epoch": 0.5170455873075216,
+      "grad_norm": 0.2857629954814911,
+      "learning_rate": 9.46467718150622e-06,
+      "loss": 0.3126,
+      "step": 13830
+    },
+    {
+      "epoch": 0.5172325162978714,
+      "grad_norm": 0.45087161660194397,
+      "learning_rate": 9.45881311985113e-06,
+      "loss": 0.3074,
+      "step": 13835
+    },
+    {
+      "epoch": 0.5174194452882211,
+      "grad_norm": 0.29453983902931213,
+      "learning_rate": 9.452949244835893e-06,
+      "loss": 0.2423,
+      "step": 13840
+    },
+    {
+      "epoch": 0.5176063742785709,
+      "grad_norm": 0.4951644241809845,
+      "learning_rate": 9.447085558482787e-06,
+      "loss": 0.2675,
+      "step": 13845
+    },
+    {
+      "epoch": 0.5177933032689207,
+      "grad_norm": 0.43756482005119324,
+      "learning_rate": 9.441222062814024e-06,
+      "loss": 0.2897,
+      "step": 13850
+    },
+    {
+      "epoch": 0.5179802322592705,
+      "grad_norm": 0.6385819911956787,
+      "learning_rate": 9.435358759851767e-06,
+      "loss": 0.2524,
+      "step": 13855
+    },
+    {
+      "epoch": 0.5181671612496203,
+      "grad_norm": 0.3710041046142578,
+      "learning_rate": 9.42949565161809e-06,
+      "loss": 0.2996,
+      "step": 13860
+    },
+    {
+      "epoch": 0.5183540902399701,
+      "grad_norm": 0.47704362869262695,
+      "learning_rate": 9.423632740135021e-06,
+      "loss": 0.2501,
+      "step": 13865
+    },
+    {
+      "epoch": 0.5185410192303199,
+      "grad_norm": 0.30794888734817505,
+      "learning_rate": 9.417770027424499e-06,
+      "loss": 0.2924,
+      "step": 13870
+    },
+    {
+      "epoch": 0.5187279482206697,
+      "grad_norm": 0.567466139793396,
+      "learning_rate": 9.411907515508415e-06,
+      "loss": 0.3012,
+      "step": 13875
+    },
+    {
+      "epoch": 0.5189148772110195,
+      "grad_norm": 0.5275660157203674,
+      "learning_rate": 9.406045206408574e-06,
+      "loss": 0.3539,
+      "step": 13880
+    },
+    {
+      "epoch": 0.5191018062013693,
+      "grad_norm": 0.17727385461330414,
+      "learning_rate": 9.400183102146726e-06,
+      "loss": 0.2714,
+      "step": 13885
+    },
+    {
+      "epoch": 0.519288735191719,
+      "grad_norm": 0.5720701813697815,
+      "learning_rate": 9.394321204744538e-06,
+      "loss": 0.2234,
+      "step": 13890
+    },
+    {
+      "epoch": 0.5194756641820688,
+      "grad_norm": 0.19189219176769257,
+      "learning_rate": 9.388459516223611e-06,
+      "loss": 0.3371,
+      "step": 13895
+    },
+    {
+      "epoch": 0.5196625931724186,
+      "grad_norm": 0.7133250832557678,
+      "learning_rate": 9.382598038605477e-06,
+      "loss": 0.3986,
+      "step": 13900
+    },
+    {
+      "epoch": 0.5198495221627685,
+      "grad_norm": 0.3994240164756775,
+      "learning_rate": 9.376736773911583e-06,
+      "loss": 0.3639,
+      "step": 13905
+    },
+    {
+      "epoch": 0.5200364511531183,
+      "grad_norm": 0.27001604437828064,
+      "learning_rate": 9.370875724163322e-06,
+      "loss": 0.2863,
+      "step": 13910
+    },
+    {
+      "epoch": 0.520223380143468,
+      "grad_norm": 0.5072554349899292,
+      "learning_rate": 9.365014891381996e-06,
+      "loss": 0.2816,
+      "step": 13915
+    },
+    {
+      "epoch": 0.5204103091338178,
+      "grad_norm": 0.4427657723426819,
+      "learning_rate": 9.35915427758884e-06,
+      "loss": 0.257,
+      "step": 13920
+    },
+    {
+      "epoch": 0.5205972381241676,
+      "grad_norm": 0.3535575270652771,
+      "learning_rate": 9.353293884805008e-06,
+      "loss": 0.2465,
+      "step": 13925
+    },
+    {
+      "epoch": 0.5207841671145174,
+      "grad_norm": 0.2692049443721771,
+      "learning_rate": 9.347433715051585e-06,
+      "loss": 0.2309,
+      "step": 13930
+    },
+    {
+      "epoch": 0.5209710961048671,
+      "grad_norm": 0.316593199968338,
+      "learning_rate": 9.341573770349579e-06,
+      "loss": 0.2596,
+      "step": 13935
+    },
+    {
+      "epoch": 0.5211580250952169,
+      "grad_norm": 0.5187430381774902,
+      "learning_rate": 9.33571405271991e-06,
+      "loss": 0.3025,
+      "step": 13940
+    },
+    {
+      "epoch": 0.5213449540855667,
+      "grad_norm": 0.4760133922100067,
+      "learning_rate": 9.329854564183433e-06,
+      "loss": 0.2937,
+      "step": 13945
+    },
+    {
+      "epoch": 0.5215318830759166,
+      "grad_norm": 0.36547884345054626,
+      "learning_rate": 9.323995306760909e-06,
+      "loss": 0.2883,
+      "step": 13950
+    },
+    {
+      "epoch": 0.5217188120662664,
+      "grad_norm": 0.3512476980686188,
+      "learning_rate": 9.31813628247304e-06,
+      "loss": 0.283,
+      "step": 13955
+    },
+    {
+      "epoch": 0.5219057410566161,
+      "grad_norm": 0.40813568234443665,
+      "learning_rate": 9.312277493340428e-06,
+      "loss": 0.278,
+      "step": 13960
+    },
+    {
+      "epoch": 0.5220926700469659,
+      "grad_norm": 0.35714733600616455,
+      "learning_rate": 9.306418941383602e-06,
+      "loss": 0.2797,
+      "step": 13965
+    },
+    {
+      "epoch": 0.5222795990373157,
+      "grad_norm": 0.4371785521507263,
+      "learning_rate": 9.300560628623007e-06,
+      "loss": 0.2286,
+      "step": 13970
+    },
+    {
+      "epoch": 0.5224665280276655,
+      "grad_norm": 0.8280014991760254,
+      "learning_rate": 9.294702557079012e-06,
+      "loss": 0.3348,
+      "step": 13975
+    },
+    {
+      "epoch": 0.5226534570180152,
+      "grad_norm": 0.6949049234390259,
+      "learning_rate": 9.288844728771898e-06,
+      "loss": 0.327,
+      "step": 13980
+    },
+    {
+      "epoch": 0.522840386008365,
+      "grad_norm": 0.573320209980011,
+      "learning_rate": 9.282987145721853e-06,
+      "loss": 0.2654,
+      "step": 13985
+    },
+    {
+      "epoch": 0.5230273149987149,
+      "grad_norm": 0.28571873903274536,
+      "learning_rate": 9.277129809949004e-06,
+      "loss": 0.3184,
+      "step": 13990
+    },
+    {
+      "epoch": 0.5232142439890647,
+      "grad_norm": 0.49826252460479736,
+      "learning_rate": 9.271272723473365e-06,
+      "loss": 0.2586,
+      "step": 13995
+    },
+    {
+      "epoch": 0.5234011729794145,
+      "grad_norm": 0.33316782116889954,
+      "learning_rate": 9.265415888314887e-06,
+      "loss": 0.3419,
+      "step": 14000
+    },
+    {
+      "epoch": 0.5235881019697642,
+      "grad_norm": 0.589277446269989,
+      "learning_rate": 9.25955930649342e-06,
+      "loss": 0.2259,
+      "step": 14005
+    },
+    {
+      "epoch": 0.523775030960114,
+      "grad_norm": 0.48725539445877075,
+      "learning_rate": 9.253702980028732e-06,
+      "loss": 0.3722,
+      "step": 14010
+    },
+    {
+      "epoch": 0.5239619599504638,
+      "grad_norm": 0.7195634245872498,
+      "learning_rate": 9.2478469109405e-06,
+      "loss": 0.2598,
+      "step": 14015
+    },
+    {
+      "epoch": 0.5241488889408136,
+      "grad_norm": 0.23791752755641937,
+      "learning_rate": 9.241991101248314e-06,
+      "loss": 0.2484,
+      "step": 14020
+    },
+    {
+      "epoch": 0.5243358179311634,
+      "grad_norm": 0.32968446612358093,
+      "learning_rate": 9.236135552971684e-06,
+      "loss": 0.2552,
+      "step": 14025
+    },
+    {
+      "epoch": 0.5245227469215132,
+      "grad_norm": 0.32867518067359924,
+      "learning_rate": 9.230280268130011e-06,
+      "loss": 0.2843,
+      "step": 14030
+    },
+    {
+      "epoch": 0.524709675911863,
+      "grad_norm": 0.4443724453449249,
+      "learning_rate": 9.22442524874262e-06,
+      "loss": 0.313,
+      "step": 14035
+    },
+    {
+      "epoch": 0.5248966049022128,
+      "grad_norm": 0.38702741265296936,
+      "learning_rate": 9.218570496828733e-06,
+      "loss": 0.2381,
+      "step": 14040
+    },
+    {
+      "epoch": 0.5250835338925626,
+      "grad_norm": 0.27657172083854675,
+      "learning_rate": 9.212716014407498e-06,
+      "loss": 0.2563,
+      "step": 14045
+    },
+    {
+      "epoch": 0.5252704628829123,
+      "grad_norm": 0.247044637799263,
+      "learning_rate": 9.206861803497946e-06,
+      "loss": 0.2164,
+      "step": 14050
+    },
+    {
+      "epoch": 0.5254573918732621,
+      "grad_norm": 0.3950256407260895,
+      "learning_rate": 9.201007866119035e-06,
+      "loss": 0.3164,
+      "step": 14055
+    },
+    {
+      "epoch": 0.5256443208636119,
+      "grad_norm": 0.4747505784034729,
+      "learning_rate": 9.195154204289614e-06,
+      "loss": 0.2861,
+      "step": 14060
+    },
+    {
+      "epoch": 0.5258312498539617,
+      "grad_norm": 0.27522939443588257,
+      "learning_rate": 9.189300820028444e-06,
+      "loss": 0.3118,
+      "step": 14065
+    },
+    {
+      "epoch": 0.5260181788443116,
+      "grad_norm": 0.4004516303539276,
+      "learning_rate": 9.183447715354197e-06,
+      "loss": 0.2722,
+      "step": 14070
+    },
+    {
+      "epoch": 0.5262051078346613,
+      "grad_norm": 1.1275408267974854,
+      "learning_rate": 9.177594892285434e-06,
+      "loss": 0.2359,
+      "step": 14075
+    },
+    {
+      "epoch": 0.5263920368250111,
+      "grad_norm": 0.221548393368721,
+      "learning_rate": 9.171742352840628e-06,
+      "loss": 0.2919,
+      "step": 14080
+    },
+    {
+      "epoch": 0.5265789658153609,
+      "grad_norm": 0.3027479946613312,
+      "learning_rate": 9.165890099038149e-06,
+      "loss": 0.2954,
+      "step": 14085
+    },
+    {
+      "epoch": 0.5267658948057107,
+      "grad_norm": 0.39747804403305054,
+      "learning_rate": 9.160038132896279e-06,
+      "loss": 0.2939,
+      "step": 14090
+    },
+    {
+      "epoch": 0.5269528237960605,
+      "grad_norm": 0.3907001316547394,
+      "learning_rate": 9.154186456433185e-06,
+      "loss": 0.3942,
+      "step": 14095
+    },
+    {
+      "epoch": 0.5271397527864102,
+      "grad_norm": 0.4810253083705902,
+      "learning_rate": 9.148335071666949e-06,
+      "loss": 0.2717,
+      "step": 14100
+    },
+    {
+      "epoch": 0.52732668177676,
+      "grad_norm": 0.21729324758052826,
+      "learning_rate": 9.142483980615545e-06,
+      "loss": 0.2178,
+      "step": 14105
+    },
+    {
+      "epoch": 0.5275136107671099,
+      "grad_norm": 0.331342875957489,
+      "learning_rate": 9.13663318529684e-06,
+      "loss": 0.3857,
+      "step": 14110
+    },
+    {
+      "epoch": 0.5277005397574597,
+      "grad_norm": 0.44765812158584595,
+      "learning_rate": 9.130782687728615e-06,
+      "loss": 0.2857,
+      "step": 14115
+    },
+    {
+      "epoch": 0.5278874687478095,
+      "grad_norm": 0.4248161315917969,
+      "learning_rate": 9.124932489928535e-06,
+      "loss": 0.3475,
+      "step": 14120
+    },
+    {
+      "epoch": 0.5280743977381592,
+      "grad_norm": 0.2518801689147949,
+      "learning_rate": 9.119082593914164e-06,
+      "loss": 0.3148,
+      "step": 14125
+    },
+    {
+      "epoch": 0.528261326728509,
+      "grad_norm": 0.24737538397312164,
+      "learning_rate": 9.113233001702963e-06,
+      "loss": 0.2533,
+      "step": 14130
+    },
+    {
+      "epoch": 0.5284482557188588,
+      "grad_norm": 0.36829543113708496,
+      "learning_rate": 9.107383715312294e-06,
+      "loss": 0.3111,
+      "step": 14135
+    },
+    {
+      "epoch": 0.5286351847092086,
+      "grad_norm": 0.42858853936195374,
+      "learning_rate": 9.101534736759402e-06,
+      "loss": 0.2648,
+      "step": 14140
+    },
+    {
+      "epoch": 0.5288221136995583,
+      "grad_norm": 0.436758428812027,
+      "learning_rate": 9.095686068061439e-06,
+      "loss": 0.296,
+      "step": 14145
+    },
+    {
+      "epoch": 0.5290090426899082,
+      "grad_norm": 0.4839959442615509,
+      "learning_rate": 9.089837711235436e-06,
+      "loss": 0.3344,
+      "step": 14150
+    },
+    {
+      "epoch": 0.529195971680258,
+      "grad_norm": 0.3259483277797699,
+      "learning_rate": 9.083989668298326e-06,
+      "loss": 0.2556,
+      "step": 14155
+    },
+    {
+      "epoch": 0.5293829006706078,
+      "grad_norm": 0.3588860332965851,
+      "learning_rate": 9.078141941266934e-06,
+      "loss": 0.2971,
+      "step": 14160
+    },
+    {
+      "epoch": 0.5295698296609576,
+      "grad_norm": 0.33428633213043213,
+      "learning_rate": 9.072294532157973e-06,
+      "loss": 0.2645,
+      "step": 14165
+    },
+    {
+      "epoch": 0.5297567586513073,
+      "grad_norm": 0.7300418615341187,
+      "learning_rate": 9.066447442988044e-06,
+      "loss": 0.3597,
+      "step": 14170
+    },
+    {
+      "epoch": 0.5299436876416571,
+      "grad_norm": 0.3091430366039276,
+      "learning_rate": 9.060600675773644e-06,
+      "loss": 0.2246,
+      "step": 14175
+    },
+    {
+      "epoch": 0.5301306166320069,
+      "grad_norm": 0.47581833600997925,
+      "learning_rate": 9.054754232531153e-06,
+      "loss": 0.2581,
+      "step": 14180
+    },
+    {
+      "epoch": 0.5303175456223567,
+      "grad_norm": 0.4748400151729584,
+      "learning_rate": 9.04890811527684e-06,
+      "loss": 0.2788,
+      "step": 14185
+    },
+    {
+      "epoch": 0.5305044746127064,
+      "grad_norm": 0.21533802151679993,
+      "learning_rate": 9.04306232602687e-06,
+      "loss": 0.2591,
+      "step": 14190
+    },
+    {
+      "epoch": 0.5306914036030563,
+      "grad_norm": 0.3625734746456146,
+      "learning_rate": 9.037216866797281e-06,
+      "loss": 0.2998,
+      "step": 14195
+    },
+    {
+      "epoch": 0.5308783325934061,
+      "grad_norm": 0.6287603378295898,
+      "learning_rate": 9.031371739604006e-06,
+      "loss": 0.3302,
+      "step": 14200
+    },
+    {
+      "epoch": 0.5310652615837559,
+      "grad_norm": 0.28976184129714966,
+      "learning_rate": 9.025526946462868e-06,
+      "loss": 0.2641,
+      "step": 14205
+    },
+    {
+      "epoch": 0.5312521905741057,
+      "grad_norm": 0.4596756398677826,
+      "learning_rate": 9.01968248938956e-06,
+      "loss": 0.2452,
+      "step": 14210
+    },
+    {
+      "epoch": 0.5314391195644554,
+      "grad_norm": 0.3350590765476227,
+      "learning_rate": 9.013838370399675e-06,
+      "loss": 0.3145,
+      "step": 14215
+    },
+    {
+      "epoch": 0.5316260485548052,
+      "grad_norm": 0.31188416481018066,
+      "learning_rate": 9.007994591508677e-06,
+      "loss": 0.2753,
+      "step": 14220
+    },
+    {
+      "epoch": 0.531812977545155,
+      "grad_norm": 0.22735527157783508,
+      "learning_rate": 9.002151154731922e-06,
+      "loss": 0.3517,
+      "step": 14225
+    },
+    {
+      "epoch": 0.5319999065355048,
+      "grad_norm": 0.32361796498298645,
+      "learning_rate": 8.996308062084638e-06,
+      "loss": 0.3578,
+      "step": 14230
+    },
+    {
+      "epoch": 0.5321868355258547,
+      "grad_norm": 0.44661062955856323,
+      "learning_rate": 8.990465315581947e-06,
+      "loss": 0.3692,
+      "step": 14235
+    },
+    {
+      "epoch": 0.5323737645162044,
+      "grad_norm": 0.6878971457481384,
+      "learning_rate": 8.984622917238842e-06,
+      "loss": 0.3867,
+      "step": 14240
+    },
+    {
+      "epoch": 0.5325606935065542,
+      "grad_norm": 0.33011388778686523,
+      "learning_rate": 8.978780869070198e-06,
+      "loss": 0.2912,
+      "step": 14245
+    },
+    {
+      "epoch": 0.532747622496904,
+      "grad_norm": 0.38461020588874817,
+      "learning_rate": 8.972939173090768e-06,
+      "loss": 0.2551,
+      "step": 14250
+    },
+    {
+      "epoch": 0.5329345514872538,
+      "grad_norm": 0.6979146599769592,
+      "learning_rate": 8.967097831315188e-06,
+      "loss": 0.3248,
+      "step": 14255
+    },
+    {
+      "epoch": 0.5331214804776035,
+      "grad_norm": 0.2943269610404968,
+      "learning_rate": 8.961256845757973e-06,
+      "loss": 0.2884,
+      "step": 14260
+    },
+    {
+      "epoch": 0.5333084094679533,
+      "grad_norm": 0.2779577076435089,
+      "learning_rate": 8.955416218433506e-06,
+      "loss": 0.2509,
+      "step": 14265
+    },
+    {
+      "epoch": 0.5334953384583031,
+      "grad_norm": 0.16579632461071014,
+      "learning_rate": 8.949575951356057e-06,
+      "loss": 0.2584,
+      "step": 14270
+    },
+    {
+      "epoch": 0.533682267448653,
+      "grad_norm": 0.30751311779022217,
+      "learning_rate": 8.94373604653976e-06,
+      "loss": 0.2895,
+      "step": 14275
+    },
+    {
+      "epoch": 0.5338691964390028,
+      "grad_norm": 4.094731330871582,
+      "learning_rate": 8.937896505998638e-06,
+      "loss": 0.3417,
+      "step": 14280
+    },
+    {
+      "epoch": 0.5340561254293525,
+      "grad_norm": 0.3279981017112732,
+      "learning_rate": 8.932057331746576e-06,
+      "loss": 0.3156,
+      "step": 14285
+    },
+    {
+      "epoch": 0.5342430544197023,
+      "grad_norm": 0.6543190479278564,
+      "learning_rate": 8.926218525797342e-06,
+      "loss": 0.2528,
+      "step": 14290
+    },
+    {
+      "epoch": 0.5344299834100521,
+      "grad_norm": 0.28789252042770386,
+      "learning_rate": 8.920380090164569e-06,
+      "loss": 0.3179,
+      "step": 14295
+    },
+    {
+      "epoch": 0.5346169124004019,
+      "grad_norm": 0.4924599230289459,
+      "learning_rate": 8.914542026861765e-06,
+      "loss": 0.3752,
+      "step": 14300
+    },
+    {
+      "epoch": 0.5348038413907517,
+      "grad_norm": 0.33676034212112427,
+      "learning_rate": 8.908704337902318e-06,
+      "loss": 0.2342,
+      "step": 14305
+    },
+    {
+      "epoch": 0.5349907703811014,
+      "grad_norm": 0.3200925290584564,
+      "learning_rate": 8.902867025299475e-06,
+      "loss": 0.3312,
+      "step": 14310
+    },
+    {
+      "epoch": 0.5351776993714513,
+      "grad_norm": 0.39076340198516846,
+      "learning_rate": 8.897030091066359e-06,
+      "loss": 0.2409,
+      "step": 14315
+    },
+    {
+      "epoch": 0.5353646283618011,
+      "grad_norm": 0.1652345061302185,
+      "learning_rate": 8.891193537215956e-06,
+      "loss": 0.2935,
+      "step": 14320
+    },
+    {
+      "epoch": 0.5355515573521509,
+      "grad_norm": 0.33995094895362854,
+      "learning_rate": 8.885357365761136e-06,
+      "loss": 0.2829,
+      "step": 14325
+    },
+    {
+      "epoch": 0.5357384863425007,
+      "grad_norm": 0.48137718439102173,
+      "learning_rate": 8.879521578714617e-06,
+      "loss": 0.3732,
+      "step": 14330
+    },
+    {
+      "epoch": 0.5359254153328504,
+      "grad_norm": 0.3473806381225586,
+      "learning_rate": 8.873686178089004e-06,
+      "loss": 0.2582,
+      "step": 14335
+    },
+    {
+      "epoch": 0.5361123443232002,
+      "grad_norm": 0.9458356499671936,
+      "learning_rate": 8.867851165896752e-06,
+      "loss": 0.2783,
+      "step": 14340
+    },
+    {
+      "epoch": 0.53629927331355,
+      "grad_norm": 0.2733253538608551,
+      "learning_rate": 8.862016544150192e-06,
+      "loss": 0.3056,
+      "step": 14345
+    },
+    {
+      "epoch": 0.5364862023038998,
+      "grad_norm": 0.33390501141548157,
+      "learning_rate": 8.856182314861524e-06,
+      "loss": 0.2556,
+      "step": 14350
+    },
+    {
+      "epoch": 0.5366731312942497,
+      "grad_norm": 0.5112034678459167,
+      "learning_rate": 8.850348480042794e-06,
+      "loss": 0.3071,
+      "step": 14355
+    },
+    {
+      "epoch": 0.5368600602845994,
+      "grad_norm": 0.3586961328983307,
+      "learning_rate": 8.844515041705938e-06,
+      "loss": 0.242,
+      "step": 14360
+    },
+    {
+      "epoch": 0.5370469892749492,
+      "grad_norm": 0.4830436408519745,
+      "learning_rate": 8.838682001862732e-06,
+      "loss": 0.227,
+      "step": 14365
+    },
+    {
+      "epoch": 0.537233918265299,
+      "grad_norm": 0.34037473797798157,
+      "learning_rate": 8.83284936252483e-06,
+      "loss": 0.3112,
+      "step": 14370
+    },
+    {
+      "epoch": 0.5374208472556488,
+      "grad_norm": 0.21011002361774445,
+      "learning_rate": 8.827017125703735e-06,
+      "loss": 0.2921,
+      "step": 14375
+    },
+    {
+      "epoch": 0.5376077762459985,
+      "grad_norm": 0.42194753885269165,
+      "learning_rate": 8.821185293410827e-06,
+      "loss": 0.3428,
+      "step": 14380
+    },
+    {
+      "epoch": 0.5377947052363483,
+      "grad_norm": 0.48263490200042725,
+      "learning_rate": 8.815353867657334e-06,
+      "loss": 0.2702,
+      "step": 14385
+    },
+    {
+      "epoch": 0.5379816342266981,
+      "grad_norm": 0.46013012528419495,
+      "learning_rate": 8.809522850454343e-06,
+      "loss": 0.3728,
+      "step": 14390
+    },
+    {
+      "epoch": 0.538168563217048,
+      "grad_norm": 0.4305642247200012,
+      "learning_rate": 8.803692243812816e-06,
+      "loss": 0.2556,
+      "step": 14395
+    },
+    {
+      "epoch": 0.5383554922073978,
+      "grad_norm": 0.3640795052051544,
+      "learning_rate": 8.79786204974355e-06,
+      "loss": 0.3871,
+      "step": 14400
+    },
+    {
+      "epoch": 0.5385424211977475,
+      "grad_norm": 0.413748562335968,
+      "learning_rate": 8.792032270257223e-06,
+      "loss": 0.2976,
+      "step": 14405
+    },
+    {
+      "epoch": 0.5387293501880973,
+      "grad_norm": 0.28498637676239014,
+      "learning_rate": 8.786202907364349e-06,
+      "loss": 0.3093,
+      "step": 14410
+    },
+    {
+      "epoch": 0.5389162791784471,
+      "grad_norm": 0.3062063455581665,
+      "learning_rate": 8.780373963075315e-06,
+      "loss": 0.3509,
+      "step": 14415
+    },
+    {
+      "epoch": 0.5391032081687969,
+      "grad_norm": 0.34864166378974915,
+      "learning_rate": 8.774545439400352e-06,
+      "loss": 0.3229,
+      "step": 14420
+    },
+    {
+      "epoch": 0.5392901371591466,
+      "grad_norm": 0.27013546228408813,
+      "learning_rate": 8.768717338349557e-06,
+      "loss": 0.2775,
+      "step": 14425
+    },
+    {
+      "epoch": 0.5394770661494964,
+      "grad_norm": 0.23966774344444275,
+      "learning_rate": 8.762889661932869e-06,
+      "loss": 0.3353,
+      "step": 14430
+    },
+    {
+      "epoch": 0.5396639951398462,
+      "grad_norm": 0.3070327639579773,
+      "learning_rate": 8.757062412160085e-06,
+      "loss": 0.2272,
+      "step": 14435
+    },
+    {
+      "epoch": 0.5398509241301961,
+      "grad_norm": 0.9705345630645752,
+      "learning_rate": 8.751235591040867e-06,
+      "loss": 0.2292,
+      "step": 14440
+    },
+    {
+      "epoch": 0.5400378531205459,
+      "grad_norm": 0.2887144386768341,
+      "learning_rate": 8.745409200584707e-06,
+      "loss": 0.2384,
+      "step": 14445
+    },
+    {
+      "epoch": 0.5402247821108956,
+      "grad_norm": 0.22318506240844727,
+      "learning_rate": 8.739583242800968e-06,
+      "loss": 0.2734,
+      "step": 14450
+    },
+    {
+      "epoch": 0.5404117111012454,
+      "grad_norm": 0.4824393093585968,
+      "learning_rate": 8.733757719698854e-06,
+      "loss": 0.2736,
+      "step": 14455
+    },
+    {
+      "epoch": 0.5405986400915952,
+      "grad_norm": 0.677008867263794,
+      "learning_rate": 8.72793263328742e-06,
+      "loss": 0.2748,
+      "step": 14460
+    },
+    {
+      "epoch": 0.540785569081945,
+      "grad_norm": 0.32696858048439026,
+      "learning_rate": 8.722107985575567e-06,
+      "loss": 0.2841,
+      "step": 14465
+    },
+    {
+      "epoch": 0.5409724980722947,
+      "grad_norm": 0.35395094752311707,
+      "learning_rate": 8.716283778572058e-06,
+      "loss": 0.3439,
+      "step": 14470
+    },
+    {
+      "epoch": 0.5411594270626445,
+      "grad_norm": 0.27938494086265564,
+      "learning_rate": 8.710460014285486e-06,
+      "loss": 0.2786,
+      "step": 14475
+    },
+    {
+      "epoch": 0.5413463560529944,
+      "grad_norm": 0.4673710763454437,
+      "learning_rate": 8.704636694724309e-06,
+      "loss": 0.2452,
+      "step": 14480
+    },
+    {
+      "epoch": 0.5415332850433442,
+      "grad_norm": 0.4600588083267212,
+      "learning_rate": 8.69881382189682e-06,
+      "loss": 0.327,
+      "step": 14485
+    },
+    {
+      "epoch": 0.541720214033694,
+      "grad_norm": 0.33369168639183044,
+      "learning_rate": 8.692991397811157e-06,
+      "loss": 0.3612,
+      "step": 14490
+    },
+    {
+      "epoch": 0.5419071430240437,
+      "grad_norm": 0.4194375276565552,
+      "learning_rate": 8.687169424475312e-06,
+      "loss": 0.2282,
+      "step": 14495
+    },
+    {
+      "epoch": 0.5420940720143935,
+      "grad_norm": 0.3811163902282715,
+      "learning_rate": 8.681347903897115e-06,
+      "loss": 0.324,
+      "step": 14500
+    },
+    {
+      "epoch": 0.5422810010047433,
+      "grad_norm": 0.6883916854858398,
+      "learning_rate": 8.675526838084244e-06,
+      "loss": 0.3881,
+      "step": 14505
+    },
+    {
+      "epoch": 0.5424679299950931,
+      "grad_norm": 0.36151525378227234,
+      "learning_rate": 8.66970622904421e-06,
+      "loss": 0.2661,
+      "step": 14510
+    },
+    {
+      "epoch": 0.5426548589854429,
+      "grad_norm": 0.4361328184604645,
+      "learning_rate": 8.663886078784386e-06,
+      "loss": 0.3693,
+      "step": 14515
+    },
+    {
+      "epoch": 0.5428417879757927,
+      "grad_norm": 0.3427176773548126,
+      "learning_rate": 8.658066389311963e-06,
+      "loss": 0.3023,
+      "step": 14520
+    },
+    {
+      "epoch": 0.5430287169661425,
+      "grad_norm": 0.4090884029865265,
+      "learning_rate": 8.652247162633994e-06,
+      "loss": 0.2677,
+      "step": 14525
+    },
+    {
+      "epoch": 0.5432156459564923,
+      "grad_norm": 0.4226764440536499,
+      "learning_rate": 8.646428400757363e-06,
+      "loss": 0.294,
+      "step": 14530
+    },
+    {
+      "epoch": 0.5434025749468421,
+      "grad_norm": 0.5746899843215942,
+      "learning_rate": 8.640610105688787e-06,
+      "loss": 0.325,
+      "step": 14535
+    },
+    {
+      "epoch": 0.5435895039371919,
+      "grad_norm": 0.37049710750579834,
+      "learning_rate": 8.634792279434838e-06,
+      "loss": 0.2668,
+      "step": 14540
+    },
+    {
+      "epoch": 0.5437764329275416,
+      "grad_norm": 0.40856534242630005,
+      "learning_rate": 8.62897492400191e-06,
+      "loss": 0.3502,
+      "step": 14545
+    },
+    {
+      "epoch": 0.5439633619178914,
+      "grad_norm": 0.3843344449996948,
+      "learning_rate": 8.623158041396251e-06,
+      "loss": 0.317,
+      "step": 14550
+    },
+    {
+      "epoch": 0.5441502909082412,
+      "grad_norm": 0.4334682822227478,
+      "learning_rate": 8.617341633623928e-06,
+      "loss": 0.3195,
+      "step": 14555
+    },
+    {
+      "epoch": 0.5443372198985911,
+      "grad_norm": 0.55138099193573,
+      "learning_rate": 8.611525702690861e-06,
+      "loss": 0.2328,
+      "step": 14560
+    },
+    {
+      "epoch": 0.5445241488889409,
+      "grad_norm": 0.599747359752655,
+      "learning_rate": 8.60571025060279e-06,
+      "loss": 0.2468,
+      "step": 14565
+    },
+    {
+      "epoch": 0.5447110778792906,
+      "grad_norm": 0.6509507298469543,
+      "learning_rate": 8.599895279365303e-06,
+      "loss": 0.3409,
+      "step": 14570
+    },
+    {
+      "epoch": 0.5448980068696404,
+      "grad_norm": 0.2125466763973236,
+      "learning_rate": 8.59408079098382e-06,
+      "loss": 0.2632,
+      "step": 14575
+    },
+    {
+      "epoch": 0.5450849358599902,
+      "grad_norm": 0.3049337565898895,
+      "learning_rate": 8.588266787463582e-06,
+      "loss": 0.2482,
+      "step": 14580
+    },
+    {
+      "epoch": 0.54527186485034,
+      "grad_norm": 0.372995525598526,
+      "learning_rate": 8.582453270809682e-06,
+      "loss": 0.3199,
+      "step": 14585
+    },
+    {
+      "epoch": 0.5454587938406897,
+      "grad_norm": 0.3478299379348755,
+      "learning_rate": 8.576640243027027e-06,
+      "loss": 0.2878,
+      "step": 14590
+    },
+    {
+      "epoch": 0.5456457228310395,
+      "grad_norm": 0.32996904850006104,
+      "learning_rate": 8.570827706120373e-06,
+      "loss": 0.2564,
+      "step": 14595
+    },
+    {
+      "epoch": 0.5458326518213894,
+      "grad_norm": 0.33126890659332275,
+      "learning_rate": 8.565015662094289e-06,
+      "loss": 0.2473,
+      "step": 14600
+    },
+    {
+      "epoch": 0.5460195808117392,
+      "grad_norm": 0.6810954809188843,
+      "learning_rate": 8.559204112953187e-06,
+      "loss": 0.3061,
+      "step": 14605
+    },
+    {
+      "epoch": 0.546206509802089,
+      "grad_norm": 0.286606103181839,
+      "learning_rate": 8.5533930607013e-06,
+      "loss": 0.4564,
+      "step": 14610
+    },
+    {
+      "epoch": 0.5463934387924387,
+      "grad_norm": 0.23572850227355957,
+      "learning_rate": 8.547582507342696e-06,
+      "loss": 0.2598,
+      "step": 14615
+    },
+    {
+      "epoch": 0.5465803677827885,
+      "grad_norm": 0.35271137952804565,
+      "learning_rate": 8.54177245488127e-06,
+      "loss": 0.2811,
+      "step": 14620
+    },
+    {
+      "epoch": 0.5467672967731383,
+      "grad_norm": 0.3139854967594147,
+      "learning_rate": 8.535962905320739e-06,
+      "loss": 0.2576,
+      "step": 14625
+    },
+    {
+      "epoch": 0.5469542257634881,
+      "grad_norm": 0.5428050756454468,
+      "learning_rate": 8.530153860664657e-06,
+      "loss": 0.2428,
+      "step": 14630
+    },
+    {
+      "epoch": 0.5471411547538378,
+      "grad_norm": 0.5763229131698608,
+      "learning_rate": 8.524345322916383e-06,
+      "loss": 0.2484,
+      "step": 14635
+    },
+    {
+      "epoch": 0.5473280837441876,
+      "grad_norm": 0.3398786783218384,
+      "learning_rate": 8.518537294079132e-06,
+      "loss": 0.3553,
+      "step": 14640
+    },
+    {
+      "epoch": 0.5475150127345375,
+      "grad_norm": 0.40959233045578003,
+      "learning_rate": 8.512729776155917e-06,
+      "loss": 0.255,
+      "step": 14645
+    },
+    {
+      "epoch": 0.5477019417248873,
+      "grad_norm": 0.34932756423950195,
+      "learning_rate": 8.506922771149588e-06,
+      "loss": 0.3235,
+      "step": 14650
+    },
+    {
+      "epoch": 0.5478888707152371,
+      "grad_norm": 0.2500455379486084,
+      "learning_rate": 8.501116281062809e-06,
+      "loss": 0.2978,
+      "step": 14655
+    },
+    {
+      "epoch": 0.5480757997055868,
+      "grad_norm": 0.37845414876937866,
+      "learning_rate": 8.495310307898076e-06,
+      "loss": 0.3534,
+      "step": 14660
+    },
+    {
+      "epoch": 0.5482627286959366,
+      "grad_norm": 0.1609395295381546,
+      "learning_rate": 8.489504853657707e-06,
+      "loss": 0.2224,
+      "step": 14665
+    },
+    {
+      "epoch": 0.5484496576862864,
+      "grad_norm": 0.37711602449417114,
+      "learning_rate": 8.48369992034383e-06,
+      "loss": 0.2195,
+      "step": 14670
+    },
+    {
+      "epoch": 0.5486365866766362,
+      "grad_norm": 0.373844712972641,
+      "learning_rate": 8.477895509958407e-06,
+      "loss": 0.2513,
+      "step": 14675
+    },
+    {
+      "epoch": 0.548823515666986,
+      "grad_norm": 0.24388177692890167,
+      "learning_rate": 8.472091624503204e-06,
+      "loss": 0.3141,
+      "step": 14680
+    },
+    {
+      "epoch": 0.5490104446573358,
+      "grad_norm": 0.5337631106376648,
+      "learning_rate": 8.466288265979822e-06,
+      "loss": 0.3304,
+      "step": 14685
+    },
+    {
+      "epoch": 0.5491973736476856,
+      "grad_norm": 0.3623740077018738,
+      "learning_rate": 8.460485436389672e-06,
+      "loss": 0.2617,
+      "step": 14690
+    },
+    {
+      "epoch": 0.5493843026380354,
+      "grad_norm": 0.47312211990356445,
+      "learning_rate": 8.454683137733982e-06,
+      "loss": 0.2966,
+      "step": 14695
+    },
+    {
+      "epoch": 0.5495712316283852,
+      "grad_norm": 0.4503675401210785,
+      "learning_rate": 8.448881372013795e-06,
+      "loss": 0.299,
+      "step": 14700
+    },
+    {
+      "epoch": 0.549758160618735,
+      "grad_norm": 0.3221707344055176,
+      "learning_rate": 8.443080141229978e-06,
+      "loss": 0.2533,
+      "step": 14705
+    },
+    {
+      "epoch": 0.5499450896090847,
+      "grad_norm": 0.25598636269569397,
+      "learning_rate": 8.437279447383213e-06,
+      "loss": 0.2354,
+      "step": 14710
+    },
+    {
+      "epoch": 0.5501320185994345,
+      "grad_norm": 0.3973354399204254,
+      "learning_rate": 8.431479292473986e-06,
+      "loss": 0.3455,
+      "step": 14715
+    },
+    {
+      "epoch": 0.5503189475897843,
+      "grad_norm": 0.30457204580307007,
+      "learning_rate": 8.42567967850261e-06,
+      "loss": 0.2669,
+      "step": 14720
+    },
+    {
+      "epoch": 0.5505058765801342,
+      "grad_norm": 0.5336440205574036,
+      "learning_rate": 8.4198806074692e-06,
+      "loss": 0.3144,
+      "step": 14725
+    },
+    {
+      "epoch": 0.550692805570484,
+      "grad_norm": 0.39468759298324585,
+      "learning_rate": 8.414082081373695e-06,
+      "loss": 0.3424,
+      "step": 14730
+    },
+    {
+      "epoch": 0.5508797345608337,
+      "grad_norm": 0.5090770125389099,
+      "learning_rate": 8.408284102215833e-06,
+      "loss": 0.2587,
+      "step": 14735
+    },
+    {
+      "epoch": 0.5510666635511835,
+      "grad_norm": 0.2587001621723175,
+      "learning_rate": 8.40248667199518e-06,
+      "loss": 0.2465,
+      "step": 14740
+    },
+    {
+      "epoch": 0.5512535925415333,
+      "grad_norm": 0.4725680351257324,
+      "learning_rate": 8.396689792711098e-06,
+      "loss": 0.2382,
+      "step": 14745
+    },
+    {
+      "epoch": 0.551440521531883,
+      "grad_norm": 0.4414125084877014,
+      "learning_rate": 8.390893466362765e-06,
+      "loss": 0.2707,
+      "step": 14750
+    },
+    {
+      "epoch": 0.5516274505222328,
+      "grad_norm": 0.43323227763175964,
+      "learning_rate": 8.385097694949171e-06,
+      "loss": 0.2367,
+      "step": 14755
+    },
+    {
+      "epoch": 0.5518143795125826,
+      "grad_norm": 0.4801645874977112,
+      "learning_rate": 8.379302480469109e-06,
+      "loss": 0.2377,
+      "step": 14760
+    },
+    {
+      "epoch": 0.5520013085029325,
+      "grad_norm": 0.42729660868644714,
+      "learning_rate": 8.373507824921184e-06,
+      "loss": 0.311,
+      "step": 14765
+    },
+    {
+      "epoch": 0.5521882374932823,
+      "grad_norm": 0.3292931318283081,
+      "learning_rate": 8.3677137303038e-06,
+      "loss": 0.2404,
+      "step": 14770
+    },
+    {
+      "epoch": 0.552375166483632,
+      "grad_norm": 0.2796480655670166,
+      "learning_rate": 8.361920198615182e-06,
+      "loss": 0.2542,
+      "step": 14775
+    },
+    {
+      "epoch": 0.5525620954739818,
+      "grad_norm": 0.1017402783036232,
+      "learning_rate": 8.35612723185335e-06,
+      "loss": 0.3207,
+      "step": 14780
+    },
+    {
+      "epoch": 0.5527490244643316,
+      "grad_norm": 0.296451598405838,
+      "learning_rate": 8.350334832016136e-06,
+      "loss": 0.269,
+      "step": 14785
+    },
+    {
+      "epoch": 0.5529359534546814,
+      "grad_norm": 0.4688802659511566,
+      "learning_rate": 8.344543001101167e-06,
+      "loss": 0.2914,
+      "step": 14790
+    },
+    {
+      "epoch": 0.5531228824450312,
+      "grad_norm": 0.2229701429605484,
+      "learning_rate": 8.33875174110588e-06,
+      "loss": 0.3238,
+      "step": 14795
+    },
+    {
+      "epoch": 0.5533098114353809,
+      "grad_norm": 0.6549943685531616,
+      "learning_rate": 8.332961054027522e-06,
+      "loss": 0.2368,
+      "step": 14800
+    },
+    {
+      "epoch": 0.5534967404257308,
+      "grad_norm": 0.4109474718570709,
+      "learning_rate": 8.327170941863124e-06,
+      "loss": 0.2954,
+      "step": 14805
+    },
+    {
+      "epoch": 0.5536836694160806,
+      "grad_norm": 0.4711514711380005,
+      "learning_rate": 8.32138140660954e-06,
+      "loss": 0.2371,
+      "step": 14810
+    },
+    {
+      "epoch": 0.5538705984064304,
+      "grad_norm": 0.32290342450141907,
+      "learning_rate": 8.31559245026341e-06,
+      "loss": 0.2736,
+      "step": 14815
+    },
+    {
+      "epoch": 0.5540575273967802,
+      "grad_norm": 0.34553197026252747,
+      "learning_rate": 8.309804074821179e-06,
+      "loss": 0.3564,
+      "step": 14820
+    },
+    {
+      "epoch": 0.5542444563871299,
+      "grad_norm": 0.3491702377796173,
+      "learning_rate": 8.304016282279089e-06,
+      "loss": 0.281,
+      "step": 14825
+    },
+    {
+      "epoch": 0.5544313853774797,
+      "grad_norm": 0.2792864739894867,
+      "learning_rate": 8.298229074633192e-06,
+      "loss": 0.2732,
+      "step": 14830
+    },
+    {
+      "epoch": 0.5546183143678295,
+      "grad_norm": 0.48167872428894043,
+      "learning_rate": 8.292442453879324e-06,
+      "loss": 0.277,
+      "step": 14835
+    },
+    {
+      "epoch": 0.5548052433581793,
+      "grad_norm": 0.563044548034668,
+      "learning_rate": 8.286656422013122e-06,
+      "loss": 0.3111,
+      "step": 14840
+    },
+    {
+      "epoch": 0.5549921723485292,
+      "grad_norm": 0.3365761637687683,
+      "learning_rate": 8.280870981030031e-06,
+      "loss": 0.286,
+      "step": 14845
+    },
+    {
+      "epoch": 0.5551791013388789,
+      "grad_norm": 0.3140876591205597,
+      "learning_rate": 8.275086132925277e-06,
+      "loss": 0.3126,
+      "step": 14850
+    },
+    {
+      "epoch": 0.5553660303292287,
+      "grad_norm": 0.4990136921405792,
+      "learning_rate": 8.269301879693892e-06,
+      "loss": 0.2471,
+      "step": 14855
+    },
+    {
+      "epoch": 0.5555529593195785,
+      "grad_norm": 0.3173409402370453,
+      "learning_rate": 8.263518223330698e-06,
+      "loss": 0.3087,
+      "step": 14860
+    },
+    {
+      "epoch": 0.5557398883099283,
+      "grad_norm": 0.32432377338409424,
+      "learning_rate": 8.257735165830314e-06,
+      "loss": 0.1955,
+      "step": 14865
+    },
+    {
+      "epoch": 0.555926817300278,
+      "grad_norm": 0.41975530982017517,
+      "learning_rate": 8.251952709187145e-06,
+      "loss": 0.2308,
+      "step": 14870
+    },
+    {
+      "epoch": 0.5561137462906278,
+      "grad_norm": 0.22599612176418304,
+      "learning_rate": 8.2461708553954e-06,
+      "loss": 0.2998,
+      "step": 14875
+    },
+    {
+      "epoch": 0.5563006752809776,
+      "grad_norm": 0.2920520603656769,
+      "learning_rate": 8.240389606449075e-06,
+      "loss": 0.2895,
+      "step": 14880
+    },
+    {
+      "epoch": 0.5564876042713274,
+      "grad_norm": 0.2696673274040222,
+      "learning_rate": 8.234608964341953e-06,
+      "loss": 0.3093,
+      "step": 14885
+    },
+    {
+      "epoch": 0.5566745332616773,
+      "grad_norm": 0.37283971905708313,
+      "learning_rate": 8.228828931067618e-06,
+      "loss": 0.3073,
+      "step": 14890
+    },
+    {
+      "epoch": 0.556861462252027,
+      "grad_norm": 0.22317394614219666,
+      "learning_rate": 8.223049508619429e-06,
+      "loss": 0.3098,
+      "step": 14895
+    },
+    {
+      "epoch": 0.5570483912423768,
+      "grad_norm": 0.5522322058677673,
+      "learning_rate": 8.217270698990555e-06,
+      "loss": 0.3913,
+      "step": 14900
+    },
+    {
+      "epoch": 0.5572353202327266,
+      "grad_norm": 0.3208252787590027,
+      "learning_rate": 8.21149250417393e-06,
+      "loss": 0.2539,
+      "step": 14905
+    },
+    {
+      "epoch": 0.5574222492230764,
+      "grad_norm": 0.3736477494239807,
+      "learning_rate": 8.205714926162298e-06,
+      "loss": 0.4014,
+      "step": 14910
+    },
+    {
+      "epoch": 0.5576091782134261,
+      "grad_norm": 0.3176252543926239,
+      "learning_rate": 8.199937966948168e-06,
+      "loss": 0.3256,
+      "step": 14915
+    },
+    {
+      "epoch": 0.5577961072037759,
+      "grad_norm": 0.3628195822238922,
+      "learning_rate": 8.194161628523863e-06,
+      "loss": 0.3898,
+      "step": 14920
+    },
+    {
+      "epoch": 0.5579830361941257,
+      "grad_norm": 0.3258436620235443,
+      "learning_rate": 8.18838591288146e-06,
+      "loss": 0.2693,
+      "step": 14925
+    },
+    {
+      "epoch": 0.5581699651844756,
+      "grad_norm": 0.4312792420387268,
+      "learning_rate": 8.18261082201285e-06,
+      "loss": 0.2869,
+      "step": 14930
+    },
+    {
+      "epoch": 0.5583568941748254,
+      "grad_norm": 0.43801358342170715,
+      "learning_rate": 8.176836357909697e-06,
+      "loss": 0.3012,
+      "step": 14935
+    },
+    {
+      "epoch": 0.5585438231651751,
+      "grad_norm": 0.4672282934188843,
+      "learning_rate": 8.171062522563438e-06,
+      "loss": 0.2744,
+      "step": 14940
+    },
+    {
+      "epoch": 0.5587307521555249,
+      "grad_norm": 0.36364656686782837,
+      "learning_rate": 8.165289317965314e-06,
+      "loss": 0.3149,
+      "step": 14945
+    },
+    {
+      "epoch": 0.5589176811458747,
+      "grad_norm": 0.6358222961425781,
+      "learning_rate": 8.159516746106331e-06,
+      "loss": 0.3848,
+      "step": 14950
+    },
+    {
+      "epoch": 0.5591046101362245,
+      "grad_norm": 0.34080687165260315,
+      "learning_rate": 8.153744808977287e-06,
+      "loss": 0.287,
+      "step": 14955
+    },
+    {
+      "epoch": 0.5592915391265743,
+      "grad_norm": 0.4259713888168335,
+      "learning_rate": 8.147973508568753e-06,
+      "loss": 0.2668,
+      "step": 14960
+    },
+    {
+      "epoch": 0.559478468116924,
+      "grad_norm": 0.25825029611587524,
+      "learning_rate": 8.142202846871093e-06,
+      "loss": 0.2385,
+      "step": 14965
+    },
+    {
+      "epoch": 0.5596653971072739,
+      "grad_norm": 0.5594242215156555,
+      "learning_rate": 8.136432825874433e-06,
+      "loss": 0.3407,
+      "step": 14970
+    },
+    {
+      "epoch": 0.5598523260976237,
+      "grad_norm": 0.39524081349372864,
+      "learning_rate": 8.130663447568696e-06,
+      "loss": 0.2397,
+      "step": 14975
+    },
+    {
+      "epoch": 0.5600392550879735,
+      "grad_norm": 0.391757071018219,
+      "learning_rate": 8.124894713943576e-06,
+      "loss": 0.2447,
+      "step": 14980
+    },
+    {
+      "epoch": 0.5602261840783233,
+      "grad_norm": 0.4008239209651947,
+      "learning_rate": 8.119126626988535e-06,
+      "loss": 0.2966,
+      "step": 14985
+    },
+    {
+      "epoch": 0.560413113068673,
+      "grad_norm": 0.3706965148448944,
+      "learning_rate": 8.11335918869283e-06,
+      "loss": 0.3095,
+      "step": 14990
+    },
+    {
+      "epoch": 0.5606000420590228,
+      "grad_norm": 1.8738888502120972,
+      "learning_rate": 8.10759240104548e-06,
+      "loss": 0.3566,
+      "step": 14995
+    },
+    {
+      "epoch": 0.5607869710493726,
+      "grad_norm": 0.28992828726768494,
+      "learning_rate": 8.10182626603529e-06,
+      "loss": 0.2649,
+      "step": 15000
+    },
+    {
+      "epoch": 0.5609739000397224,
+      "grad_norm": 0.2337641566991806,
+      "learning_rate": 8.096060785650829e-06,
+      "loss": 0.2402,
+      "step": 15005
+    },
+    {
+      "epoch": 0.5611608290300723,
+      "grad_norm": 0.28632795810699463,
+      "learning_rate": 8.09029596188045e-06,
+      "loss": 0.25,
+      "step": 15010
+    },
+    {
+      "epoch": 0.561347758020422,
+      "grad_norm": 0.4603891670703888,
+      "learning_rate": 8.08453179671227e-06,
+      "loss": 0.3261,
+      "step": 15015
+    },
+    {
+      "epoch": 0.5615346870107718,
+      "grad_norm": 0.4253883361816406,
+      "learning_rate": 8.07876829213419e-06,
+      "loss": 0.3169,
+      "step": 15020
+    },
+    {
+      "epoch": 0.5617216160011216,
+      "grad_norm": 0.3845186233520508,
+      "learning_rate": 8.073005450133877e-06,
+      "loss": 0.2886,
+      "step": 15025
+    },
+    {
+      "epoch": 0.5619085449914714,
+      "grad_norm": 0.31487545371055603,
+      "learning_rate": 8.067243272698766e-06,
+      "loss": 0.2967,
+      "step": 15030
+    },
+    {
+      "epoch": 0.5620954739818211,
+      "grad_norm": 0.29043373465538025,
+      "learning_rate": 8.061481761816073e-06,
+      "loss": 0.2707,
+      "step": 15035
+    },
+    {
+      "epoch": 0.5622824029721709,
+      "grad_norm": 0.4541192054748535,
+      "learning_rate": 8.055720919472771e-06,
+      "loss": 0.3279,
+      "step": 15040
+    },
+    {
+      "epoch": 0.5624693319625207,
+      "grad_norm": 0.3087974190711975,
+      "learning_rate": 8.049960747655618e-06,
+      "loss": 0.231,
+      "step": 15045
+    },
+    {
+      "epoch": 0.5626562609528706,
+      "grad_norm": 1.4350837469100952,
+      "learning_rate": 8.044201248351125e-06,
+      "loss": 0.3562,
+      "step": 15050
+    },
+    {
+      "epoch": 0.5628431899432204,
+      "grad_norm": 0.5882908701896667,
+      "learning_rate": 8.038442423545583e-06,
+      "loss": 0.337,
+      "step": 15055
+    },
+    {
+      "epoch": 0.5630301189335701,
+      "grad_norm": 0.3661154806613922,
+      "learning_rate": 8.032684275225038e-06,
+      "loss": 0.4017,
+      "step": 15060
+    },
+    {
+      "epoch": 0.5632170479239199,
+      "grad_norm": 0.552562952041626,
+      "learning_rate": 8.026926805375319e-06,
+      "loss": 0.2177,
+      "step": 15065
+    },
+    {
+      "epoch": 0.5634039769142697,
+      "grad_norm": 0.5023968815803528,
+      "learning_rate": 8.021170015982009e-06,
+      "loss": 0.2301,
+      "step": 15070
+    },
+    {
+      "epoch": 0.5635909059046195,
+      "grad_norm": 0.22658932209014893,
+      "learning_rate": 8.01541390903046e-06,
+      "loss": 0.3225,
+      "step": 15075
+    },
+    {
+      "epoch": 0.5637778348949692,
+      "grad_norm": 0.32941967248916626,
+      "learning_rate": 8.00965848650579e-06,
+      "loss": 0.264,
+      "step": 15080
+    },
+    {
+      "epoch": 0.563964763885319,
+      "grad_norm": 0.5150809288024902,
+      "learning_rate": 8.003903750392872e-06,
+      "loss": 0.2672,
+      "step": 15085
+    },
+    {
+      "epoch": 0.5641516928756689,
+      "grad_norm": 0.2671123743057251,
+      "learning_rate": 7.99814970267636e-06,
+      "loss": 0.2327,
+      "step": 15090
+    },
+    {
+      "epoch": 0.5643386218660187,
+      "grad_norm": 0.4776862561702728,
+      "learning_rate": 7.992396345340654e-06,
+      "loss": 0.2703,
+      "step": 15095
+    },
+    {
+      "epoch": 0.5645255508563685,
+      "grad_norm": 0.34787583351135254,
+      "learning_rate": 7.986643680369925e-06,
+      "loss": 0.2732,
+      "step": 15100
+    },
+    {
+      "epoch": 0.5647124798467182,
+      "grad_norm": 0.34038421511650085,
+      "learning_rate": 7.980891709748097e-06,
+      "loss": 0.2566,
+      "step": 15105
+    },
+    {
+      "epoch": 0.564899408837068,
+      "grad_norm": 0.5559023022651672,
+      "learning_rate": 7.975140435458864e-06,
+      "loss": 0.3003,
+      "step": 15110
+    },
+    {
+      "epoch": 0.5650863378274178,
+      "grad_norm": 0.6152315139770508,
+      "learning_rate": 7.969389859485679e-06,
+      "loss": 0.3102,
+      "step": 15115
+    },
+    {
+      "epoch": 0.5652732668177676,
+      "grad_norm": 0.28499045968055725,
+      "learning_rate": 7.963639983811744e-06,
+      "loss": 0.3374,
+      "step": 15120
+    },
+    {
+      "epoch": 0.5654601958081173,
+      "grad_norm": 0.4572030007839203,
+      "learning_rate": 7.957890810420033e-06,
+      "loss": 0.2414,
+      "step": 15125
+    },
+    {
+      "epoch": 0.5656471247984671,
+      "grad_norm": 0.44276192784309387,
+      "learning_rate": 7.952142341293264e-06,
+      "loss": 0.3226,
+      "step": 15130
+    },
+    {
+      "epoch": 0.565834053788817,
+      "grad_norm": 0.44015127420425415,
+      "learning_rate": 7.946394578413923e-06,
+      "loss": 0.2718,
+      "step": 15135
+    },
+    {
+      "epoch": 0.5660209827791668,
+      "grad_norm": 0.4518749415874481,
+      "learning_rate": 7.940647523764251e-06,
+      "loss": 0.2729,
+      "step": 15140
+    },
+    {
+      "epoch": 0.5662079117695166,
+      "grad_norm": 0.13977321982383728,
+      "learning_rate": 7.93490117932624e-06,
+      "loss": 0.3054,
+      "step": 15145
+    },
+    {
+      "epoch": 0.5663948407598663,
+      "grad_norm": 0.3425142765045166,
+      "learning_rate": 7.929155547081637e-06,
+      "loss": 0.2925,
+      "step": 15150
+    },
+    {
+      "epoch": 0.5665817697502161,
+      "grad_norm": 1.439296841621399,
+      "learning_rate": 7.923410629011947e-06,
+      "loss": 0.2461,
+      "step": 15155
+    },
+    {
+      "epoch": 0.5667686987405659,
+      "grad_norm": 0.3877362310886383,
+      "learning_rate": 7.917666427098434e-06,
+      "loss": 0.2503,
+      "step": 15160
+    },
+    {
+      "epoch": 0.5669556277309157,
+      "grad_norm": 0.4661048352718353,
+      "learning_rate": 7.911922943322102e-06,
+      "loss": 0.3003,
+      "step": 15165
+    },
+    {
+      "epoch": 0.5671425567212655,
+      "grad_norm": 0.5539686679840088,
+      "learning_rate": 7.906180179663719e-06,
+      "loss": 0.3221,
+      "step": 15170
+    },
+    {
+      "epoch": 0.5673294857116153,
+      "grad_norm": 0.45953917503356934,
+      "learning_rate": 7.900438138103791e-06,
+      "loss": 0.2582,
+      "step": 15175
+    },
+    {
+      "epoch": 0.5675164147019651,
+      "grad_norm": 0.5768008828163147,
+      "learning_rate": 7.894696820622594e-06,
+      "loss": 0.239,
+      "step": 15180
+    },
+    {
+      "epoch": 0.5677033436923149,
+      "grad_norm": 0.3129447400569916,
+      "learning_rate": 7.888956229200134e-06,
+      "loss": 0.3238,
+      "step": 15185
+    },
+    {
+      "epoch": 0.5678902726826647,
+      "grad_norm": 0.3107287287712097,
+      "learning_rate": 7.883216365816186e-06,
+      "loss": 0.2606,
+      "step": 15190
+    },
+    {
+      "epoch": 0.5680772016730145,
+      "grad_norm": 0.35008296370506287,
+      "learning_rate": 7.877477232450258e-06,
+      "loss": 0.252,
+      "step": 15195
+    },
+    {
+      "epoch": 0.5682641306633642,
+      "grad_norm": 0.33052533864974976,
+      "learning_rate": 7.871738831081613e-06,
+      "loss": 0.2242,
+      "step": 15200
+    },
+    {
+      "epoch": 0.568451059653714,
+      "grad_norm": 0.36529335379600525,
+      "learning_rate": 7.866001163689264e-06,
+      "loss": 0.4166,
+      "step": 15205
+    },
+    {
+      "epoch": 0.5686379886440638,
+      "grad_norm": 0.5268134474754333,
+      "learning_rate": 7.860264232251968e-06,
+      "loss": 0.3337,
+      "step": 15210
+    },
+    {
+      "epoch": 0.5688249176344137,
+      "grad_norm": 0.5382173657417297,
+      "learning_rate": 7.85452803874823e-06,
+      "loss": 0.3197,
+      "step": 15215
+    },
+    {
+      "epoch": 0.5690118466247635,
+      "grad_norm": 0.29692742228507996,
+      "learning_rate": 7.84879258515629e-06,
+      "loss": 0.3722,
+      "step": 15220
+    },
+    {
+      "epoch": 0.5691987756151132,
+      "grad_norm": 0.31395187973976135,
+      "learning_rate": 7.843057873454151e-06,
+      "loss": 0.2898,
+      "step": 15225
+    },
+    {
+      "epoch": 0.569385704605463,
+      "grad_norm": 0.16513025760650635,
+      "learning_rate": 7.837323905619543e-06,
+      "loss": 0.2928,
+      "step": 15230
+    },
+    {
+      "epoch": 0.5695726335958128,
+      "grad_norm": 0.396751344203949,
+      "learning_rate": 7.831590683629957e-06,
+      "loss": 0.3757,
+      "step": 15235
+    },
+    {
+      "epoch": 0.5697595625861626,
+      "grad_norm": 0.6405811905860901,
+      "learning_rate": 7.825858209462609e-06,
+      "loss": 0.3309,
+      "step": 15240
+    },
+    {
+      "epoch": 0.5699464915765123,
+      "grad_norm": 0.3609001934528351,
+      "learning_rate": 7.820126485094465e-06,
+      "loss": 0.1863,
+      "step": 15245
+    },
+    {
+      "epoch": 0.5701334205668621,
+      "grad_norm": 0.3763851821422577,
+      "learning_rate": 7.814395512502239e-06,
+      "loss": 0.254,
+      "step": 15250
+    },
+    {
+      "epoch": 0.570320349557212,
+      "grad_norm": 0.37580713629722595,
+      "learning_rate": 7.80866529366237e-06,
+      "loss": 0.2708,
+      "step": 15255
+    },
+    {
+      "epoch": 0.5705072785475618,
+      "grad_norm": 0.4320635497570038,
+      "learning_rate": 7.802935830551058e-06,
+      "loss": 0.2882,
+      "step": 15260
+    },
+    {
+      "epoch": 0.5706942075379116,
+      "grad_norm": 0.4472927153110504,
+      "learning_rate": 7.797207125144222e-06,
+      "loss": 0.2909,
+      "step": 15265
+    },
+    {
+      "epoch": 0.5708811365282613,
+      "grad_norm": 0.32192474603652954,
+      "learning_rate": 7.791479179417532e-06,
+      "loss": 0.2998,
+      "step": 15270
+    },
+    {
+      "epoch": 0.5710680655186111,
+      "grad_norm": 0.39683884382247925,
+      "learning_rate": 7.785751995346385e-06,
+      "loss": 0.2981,
+      "step": 15275
+    },
+    {
+      "epoch": 0.5712549945089609,
+      "grad_norm": 0.5190780758857727,
+      "learning_rate": 7.780025574905935e-06,
+      "loss": 0.2564,
+      "step": 15280
+    },
+    {
+      "epoch": 0.5714419234993107,
+      "grad_norm": 0.41170167922973633,
+      "learning_rate": 7.774299920071052e-06,
+      "loss": 0.3135,
+      "step": 15285
+    },
+    {
+      "epoch": 0.5716288524896604,
+      "grad_norm": 0.5996455550193787,
+      "learning_rate": 7.768575032816347e-06,
+      "loss": 0.3116,
+      "step": 15290
+    },
+    {
+      "epoch": 0.5718157814800103,
+      "grad_norm": 0.40233224630355835,
+      "learning_rate": 7.762850915116183e-06,
+      "loss": 0.261,
+      "step": 15295
+    },
+    {
+      "epoch": 0.5720027104703601,
+      "grad_norm": 1.258935809135437,
+      "learning_rate": 7.757127568944629e-06,
+      "loss": 0.2613,
+      "step": 15300
+    },
+    {
+      "epoch": 0.5721896394607099,
+      "grad_norm": 0.4860970377922058,
+      "learning_rate": 7.751404996275515e-06,
+      "loss": 0.3349,
+      "step": 15305
+    },
+    {
+      "epoch": 0.5723765684510597,
+      "grad_norm": 0.6576589941978455,
+      "learning_rate": 7.745683199082385e-06,
+      "loss": 0.3026,
+      "step": 15310
+    },
+    {
+      "epoch": 0.5725634974414094,
+      "grad_norm": 0.39305242896080017,
+      "learning_rate": 7.739962179338528e-06,
+      "loss": 0.3438,
+      "step": 15315
+    },
+    {
+      "epoch": 0.5727504264317592,
+      "grad_norm": 0.23568610846996307,
+      "learning_rate": 7.734241939016953e-06,
+      "loss": 0.2488,
+      "step": 15320
+    },
+    {
+      "epoch": 0.572937355422109,
+      "grad_norm": 0.39915916323661804,
+      "learning_rate": 7.728522480090415e-06,
+      "loss": 0.3723,
+      "step": 15325
+    },
+    {
+      "epoch": 0.5731242844124588,
+      "grad_norm": 0.4517640173435211,
+      "learning_rate": 7.722803804531385e-06,
+      "loss": 0.2179,
+      "step": 15330
+    },
+    {
+      "epoch": 0.5733112134028087,
+      "grad_norm": 0.5729992985725403,
+      "learning_rate": 7.717085914312071e-06,
+      "loss": 0.2954,
+      "step": 15335
+    },
+    {
+      "epoch": 0.5734981423931584,
+      "grad_norm": 0.26858532428741455,
+      "learning_rate": 7.711368811404417e-06,
+      "loss": 0.3368,
+      "step": 15340
+    },
+    {
+      "epoch": 0.5736850713835082,
+      "grad_norm": 0.3362521231174469,
+      "learning_rate": 7.705652497780076e-06,
+      "loss": 0.2627,
+      "step": 15345
+    },
+    {
+      "epoch": 0.573872000373858,
+      "grad_norm": 0.5297291874885559,
+      "learning_rate": 7.699936975410452e-06,
+      "loss": 0.3659,
+      "step": 15350
+    },
+    {
+      "epoch": 0.5740589293642078,
+      "grad_norm": 0.8421899080276489,
+      "learning_rate": 7.694222246266659e-06,
+      "loss": 0.2792,
+      "step": 15355
+    },
+    {
+      "epoch": 0.5742458583545575,
+      "grad_norm": 0.3951667845249176,
+      "learning_rate": 7.688508312319545e-06,
+      "loss": 0.3122,
+      "step": 15360
+    },
+    {
+      "epoch": 0.5744327873449073,
+      "grad_norm": 0.23748940229415894,
+      "learning_rate": 7.682795175539677e-06,
+      "loss": 0.3775,
+      "step": 15365
+    },
+    {
+      "epoch": 0.5746197163352571,
+      "grad_norm": 0.42018768191337585,
+      "learning_rate": 7.677082837897362e-06,
+      "loss": 0.2916,
+      "step": 15370
+    },
+    {
+      "epoch": 0.5748066453256069,
+      "grad_norm": 0.21799716353416443,
+      "learning_rate": 7.671371301362613e-06,
+      "loss": 0.2834,
+      "step": 15375
+    },
+    {
+      "epoch": 0.5749935743159568,
+      "grad_norm": 0.3239682912826538,
+      "learning_rate": 7.66566056790518e-06,
+      "loss": 0.3178,
+      "step": 15380
+    },
+    {
+      "epoch": 0.5751805033063065,
+      "grad_norm": 0.2807358205318451,
+      "learning_rate": 7.659950639494531e-06,
+      "loss": 0.2459,
+      "step": 15385
+    },
+    {
+      "epoch": 0.5753674322966563,
+      "grad_norm": 0.39530667662620544,
+      "learning_rate": 7.654241518099851e-06,
+      "loss": 0.306,
+      "step": 15390
+    },
+    {
+      "epoch": 0.5755543612870061,
+      "grad_norm": 0.2550390362739563,
+      "learning_rate": 7.648533205690062e-06,
+      "loss": 0.3522,
+      "step": 15395
+    },
+    {
+      "epoch": 0.5757412902773559,
+      "grad_norm": 0.4049346148967743,
+      "learning_rate": 7.64282570423379e-06,
+      "loss": 0.3467,
+      "step": 15400
+    },
+    {
+      "epoch": 0.5759282192677057,
+      "grad_norm": 0.7592940926551819,
+      "learning_rate": 7.637119015699394e-06,
+      "loss": 0.2773,
+      "step": 15405
+    },
+    {
+      "epoch": 0.5761151482580554,
+      "grad_norm": 0.43101072311401367,
+      "learning_rate": 7.631413142054938e-06,
+      "loss": 0.2392,
+      "step": 15410
+    },
+    {
+      "epoch": 0.5763020772484052,
+      "grad_norm": 0.2688102126121521,
+      "learning_rate": 7.625708085268227e-06,
+      "loss": 0.2488,
+      "step": 15415
+    },
+    {
+      "epoch": 0.5764890062387551,
+      "grad_norm": 0.4400135576725006,
+      "learning_rate": 7.620003847306761e-06,
+      "loss": 0.3319,
+      "step": 15420
+    },
+    {
+      "epoch": 0.5766759352291049,
+      "grad_norm": 0.35837605595588684,
+      "learning_rate": 7.6143004301377735e-06,
+      "loss": 0.3054,
+      "step": 15425
+    },
+    {
+      "epoch": 0.5768628642194547,
+      "grad_norm": 0.42280635237693787,
+      "learning_rate": 7.6085978357282105e-06,
+      "loss": 0.31,
+      "step": 15430
+    },
+    {
+      "epoch": 0.5770497932098044,
+      "grad_norm": 0.36921992897987366,
+      "learning_rate": 7.60289606604473e-06,
+      "loss": 0.2359,
+      "step": 15435
+    },
+    {
+      "epoch": 0.5772367222001542,
+      "grad_norm": 0.676522433757782,
+      "learning_rate": 7.597195123053711e-06,
+      "loss": 0.2833,
+      "step": 15440
+    },
+    {
+      "epoch": 0.577423651190504,
+      "grad_norm": 0.4574277102947235,
+      "learning_rate": 7.591495008721243e-06,
+      "loss": 0.2164,
+      "step": 15445
+    },
+    {
+      "epoch": 0.5776105801808538,
+      "grad_norm": 0.3864579200744629,
+      "learning_rate": 7.585795725013138e-06,
+      "loss": 0.3748,
+      "step": 15450
+    },
+    {
+      "epoch": 0.5777975091712035,
+      "grad_norm": 0.5847076177597046,
+      "learning_rate": 7.580097273894911e-06,
+      "loss": 0.3585,
+      "step": 15455
+    },
+    {
+      "epoch": 0.5779844381615534,
+      "grad_norm": 0.2284291684627533,
+      "learning_rate": 7.574399657331796e-06,
+      "loss": 0.3382,
+      "step": 15460
+    },
+    {
+      "epoch": 0.5781713671519032,
+      "grad_norm": 0.5911282300949097,
+      "learning_rate": 7.568702877288732e-06,
+      "loss": 0.2877,
+      "step": 15465
+    },
+    {
+      "epoch": 0.578358296142253,
+      "grad_norm": 0.2202225774526596,
+      "learning_rate": 7.5630069357303835e-06,
+      "loss": 0.2816,
+      "step": 15470
+    },
+    {
+      "epoch": 0.5785452251326028,
+      "grad_norm": 0.4826430082321167,
+      "learning_rate": 7.557311834621116e-06,
+      "loss": 0.2662,
+      "step": 15475
+    },
+    {
+      "epoch": 0.5787321541229525,
+      "grad_norm": 0.23101915419101715,
+      "learning_rate": 7.551617575925001e-06,
+      "loss": 0.3696,
+      "step": 15480
+    },
+    {
+      "epoch": 0.5789190831133023,
+      "grad_norm": 0.4637044668197632,
+      "learning_rate": 7.545924161605832e-06,
+      "loss": 0.3342,
+      "step": 15485
+    },
+    {
+      "epoch": 0.5791060121036521,
+      "grad_norm": 0.3432970345020294,
+      "learning_rate": 7.540231593627098e-06,
+      "loss": 0.2257,
+      "step": 15490
+    },
+    {
+      "epoch": 0.5792929410940019,
+      "grad_norm": 0.18397098779678345,
+      "learning_rate": 7.5345398739520105e-06,
+      "loss": 0.2383,
+      "step": 15495
+    },
+    {
+      "epoch": 0.5794798700843518,
+      "grad_norm": 0.47786015272140503,
+      "learning_rate": 7.528849004543473e-06,
+      "loss": 0.3306,
+      "step": 15500
+    },
+    {
+      "epoch": 0.5796667990747015,
+      "grad_norm": 0.2156762033700943,
+      "learning_rate": 7.52315898736411e-06,
+      "loss": 0.3126,
+      "step": 15505
+    },
+    {
+      "epoch": 0.5798537280650513,
+      "grad_norm": 0.22711989283561707,
+      "learning_rate": 7.517469824376238e-06,
+      "loss": 0.2556,
+      "step": 15510
+    },
+    {
+      "epoch": 0.5800406570554011,
+      "grad_norm": 0.4607297480106354,
+      "learning_rate": 7.5117815175418914e-06,
+      "loss": 0.3023,
+      "step": 15515
+    },
+    {
+      "epoch": 0.5802275860457509,
+      "grad_norm": 0.43681520223617554,
+      "learning_rate": 7.506094068822801e-06,
+      "loss": 0.2141,
+      "step": 15520
+    },
+    {
+      "epoch": 0.5804145150361006,
+      "grad_norm": 0.33917364478111267,
+      "learning_rate": 7.50040748018041e-06,
+      "loss": 0.3088,
+      "step": 15525
+    },
+    {
+      "epoch": 0.5806014440264504,
+      "grad_norm": 0.2788265347480774,
+      "learning_rate": 7.494721753575856e-06,
+      "loss": 0.2605,
+      "step": 15530
+    },
+    {
+      "epoch": 0.5807883730168002,
+      "grad_norm": 0.34832385182380676,
+      "learning_rate": 7.489036890969981e-06,
+      "loss": 0.2741,
+      "step": 15535
+    },
+    {
+      "epoch": 0.5809753020071501,
+      "grad_norm": 0.3322335481643677,
+      "learning_rate": 7.483352894323339e-06,
+      "loss": 0.2696,
+      "step": 15540
+    },
+    {
+      "epoch": 0.5811622309974999,
+      "grad_norm": 0.27210044860839844,
+      "learning_rate": 7.47766976559617e-06,
+      "loss": 0.272,
+      "step": 15545
+    },
+    {
+      "epoch": 0.5813491599878496,
+      "grad_norm": 0.4030235707759857,
+      "learning_rate": 7.471987506748426e-06,
+      "loss": 0.3164,
+      "step": 15550
+    },
+    {
+      "epoch": 0.5815360889781994,
+      "grad_norm": 0.31630632281303406,
+      "learning_rate": 7.466306119739751e-06,
+      "loss": 0.2757,
+      "step": 15555
+    },
+    {
+      "epoch": 0.5817230179685492,
+      "grad_norm": 0.34428203105926514,
+      "learning_rate": 7.4606256065295e-06,
+      "loss": 0.3276,
+      "step": 15560
+    },
+    {
+      "epoch": 0.581909946958899,
+      "grad_norm": 0.41596320271492004,
+      "learning_rate": 7.4549459690767105e-06,
+      "loss": 0.3167,
+      "step": 15565
+    },
+    {
+      "epoch": 0.5820968759492487,
+      "grad_norm": 0.275072306394577,
+      "learning_rate": 7.4492672093401345e-06,
+      "loss": 0.2606,
+      "step": 15570
+    },
+    {
+      "epoch": 0.5822838049395985,
+      "grad_norm": 0.29760444164276123,
+      "learning_rate": 7.443589329278211e-06,
+      "loss": 0.3137,
+      "step": 15575
+    },
+    {
+      "epoch": 0.5824707339299484,
+      "grad_norm": 0.40786728262901306,
+      "learning_rate": 7.4379123308490735e-06,
+      "loss": 0.2841,
+      "step": 15580
+    },
+    {
+      "epoch": 0.5826576629202982,
+      "grad_norm": 0.40761783719062805,
+      "learning_rate": 7.432236216010564e-06,
+      "loss": 0.2717,
+      "step": 15585
+    },
+    {
+      "epoch": 0.582844591910648,
+      "grad_norm": 0.36162450909614563,
+      "learning_rate": 7.426560986720206e-06,
+      "loss": 0.314,
+      "step": 15590
+    },
+    {
+      "epoch": 0.5830315209009977,
+      "grad_norm": 0.36004146933555603,
+      "learning_rate": 7.4208866449352275e-06,
+      "loss": 0.3408,
+      "step": 15595
+    },
+    {
+      "epoch": 0.5832184498913475,
+      "grad_norm": 0.42730987071990967,
+      "learning_rate": 7.4152131926125405e-06,
+      "loss": 0.2853,
+      "step": 15600
+    },
+    {
+      "epoch": 0.5834053788816973,
+      "grad_norm": 0.4817928671836853,
+      "learning_rate": 7.409540631708763e-06,
+      "loss": 0.3038,
+      "step": 15605
+    },
+    {
+      "epoch": 0.5835923078720471,
+      "grad_norm": 0.20412497222423553,
+      "learning_rate": 7.403868964180192e-06,
+      "loss": 0.262,
+      "step": 15610
+    },
+    {
+      "epoch": 0.5837792368623969,
+      "grad_norm": 0.18026775121688843,
+      "learning_rate": 7.398198191982828e-06,
+      "loss": 0.2356,
+      "step": 15615
+    },
+    {
+      "epoch": 0.5839661658527466,
+      "grad_norm": 0.22581899166107178,
+      "learning_rate": 7.39252831707236e-06,
+      "loss": 0.3154,
+      "step": 15620
+    },
+    {
+      "epoch": 0.5841530948430965,
+      "grad_norm": 0.26380541920661926,
+      "learning_rate": 7.386859341404158e-06,
+      "loss": 0.3565,
+      "step": 15625
+    },
+    {
+      "epoch": 0.5843400238334463,
+      "grad_norm": 0.4216972291469574,
+      "learning_rate": 7.3811912669332965e-06,
+      "loss": 0.2868,
+      "step": 15630
+    },
+    {
+      "epoch": 0.5845269528237961,
+      "grad_norm": 0.4482162892818451,
+      "learning_rate": 7.375524095614524e-06,
+      "loss": 0.2696,
+      "step": 15635
+    },
+    {
+      "epoch": 0.5847138818141459,
+      "grad_norm": 0.34576284885406494,
+      "learning_rate": 7.369857829402294e-06,
+      "loss": 0.2722,
+      "step": 15640
+    },
+    {
+      "epoch": 0.5849008108044956,
+      "grad_norm": 0.35120487213134766,
+      "learning_rate": 7.364192470250735e-06,
+      "loss": 0.2365,
+      "step": 15645
+    },
+    {
+      "epoch": 0.5850877397948454,
+      "grad_norm": 0.5268846154212952,
+      "learning_rate": 7.358528020113669e-06,
+      "loss": 0.3953,
+      "step": 15650
+    },
+    {
+      "epoch": 0.5852746687851952,
+      "grad_norm": 0.5416049957275391,
+      "learning_rate": 7.352864480944597e-06,
+      "loss": 0.3447,
+      "step": 15655
+    },
+    {
+      "epoch": 0.585461597775545,
+      "grad_norm": 0.3339357376098633,
+      "learning_rate": 7.3472018546967175e-06,
+      "loss": 0.3417,
+      "step": 15660
+    },
+    {
+      "epoch": 0.5856485267658949,
+      "grad_norm": 0.4160725474357605,
+      "learning_rate": 7.341540143322907e-06,
+      "loss": 0.331,
+      "step": 15665
+    },
+    {
+      "epoch": 0.5858354557562446,
+      "grad_norm": 0.4474146366119385,
+      "learning_rate": 7.335879348775724e-06,
+      "loss": 0.2705,
+      "step": 15670
+    },
+    {
+      "epoch": 0.5860223847465944,
+      "grad_norm": 0.39802151918411255,
+      "learning_rate": 7.33021947300742e-06,
+      "loss": 0.2973,
+      "step": 15675
+    },
+    {
+      "epoch": 0.5862093137369442,
+      "grad_norm": 0.3424316346645355,
+      "learning_rate": 7.324560517969918e-06,
+      "loss": 0.2956,
+      "step": 15680
+    },
+    {
+      "epoch": 0.586396242727294,
+      "grad_norm": 0.31695395708084106,
+      "learning_rate": 7.318902485614836e-06,
+      "loss": 0.2828,
+      "step": 15685
+    },
+    {
+      "epoch": 0.5865831717176437,
+      "grad_norm": 0.30436766147613525,
+      "learning_rate": 7.313245377893461e-06,
+      "loss": 0.2704,
+      "step": 15690
+    },
+    {
+      "epoch": 0.5867701007079935,
+      "grad_norm": 0.349795401096344,
+      "learning_rate": 7.307589196756772e-06,
+      "loss": 0.3184,
+      "step": 15695
+    },
+    {
+      "epoch": 0.5869570296983433,
+      "grad_norm": 0.3413912057876587,
+      "learning_rate": 7.301933944155417e-06,
+      "loss": 0.2635,
+      "step": 15700
+    },
+    {
+      "epoch": 0.5871439586886932,
+      "grad_norm": 0.3070078194141388,
+      "learning_rate": 7.296279622039737e-06,
+      "loss": 0.36,
+      "step": 15705
+    },
+    {
+      "epoch": 0.587330887679043,
+      "grad_norm": 0.29595261812210083,
+      "learning_rate": 7.290626232359746e-06,
+      "loss": 0.2837,
+      "step": 15710
+    },
+    {
+      "epoch": 0.5875178166693927,
+      "grad_norm": 0.47354787588119507,
+      "learning_rate": 7.284973777065134e-06,
+      "loss": 0.2849,
+      "step": 15715
+    },
+    {
+      "epoch": 0.5877047456597425,
+      "grad_norm": 0.5047826766967773,
+      "learning_rate": 7.279322258105272e-06,
+      "loss": 0.2876,
+      "step": 15720
+    },
+    {
+      "epoch": 0.5878916746500923,
+      "grad_norm": 0.5029893517494202,
+      "learning_rate": 7.273671677429202e-06,
+      "loss": 0.3886,
+      "step": 15725
+    },
+    {
+      "epoch": 0.5880786036404421,
+      "grad_norm": 0.27061954140663147,
+      "learning_rate": 7.2680220369856546e-06,
+      "loss": 0.3005,
+      "step": 15730
+    },
+    {
+      "epoch": 0.5882655326307918,
+      "grad_norm": 0.5909748673439026,
+      "learning_rate": 7.2623733387230245e-06,
+      "loss": 0.2534,
+      "step": 15735
+    },
+    {
+      "epoch": 0.5884524616211416,
+      "grad_norm": 0.3261001408100128,
+      "learning_rate": 7.256725584589388e-06,
+      "loss": 0.2761,
+      "step": 15740
+    },
+    {
+      "epoch": 0.5886393906114915,
+      "grad_norm": 0.47955918312072754,
+      "learning_rate": 7.25107877653249e-06,
+      "loss": 0.2752,
+      "step": 15745
+    },
+    {
+      "epoch": 0.5888263196018413,
+      "grad_norm": 0.49793973565101624,
+      "learning_rate": 7.245432916499755e-06,
+      "loss": 0.3089,
+      "step": 15750
+    },
+    {
+      "epoch": 0.5890132485921911,
+      "grad_norm": 0.5249269604682922,
+      "learning_rate": 7.2397880064382816e-06,
+      "loss": 0.3462,
+      "step": 15755
+    },
+    {
+      "epoch": 0.5892001775825408,
+      "grad_norm": 0.7072009444236755,
+      "learning_rate": 7.234144048294833e-06,
+      "loss": 0.2022,
+      "step": 15760
+    },
+    {
+      "epoch": 0.5893871065728906,
+      "grad_norm": 0.22332626581192017,
+      "learning_rate": 7.228501044015854e-06,
+      "loss": 0.3619,
+      "step": 15765
+    },
+    {
+      "epoch": 0.5895740355632404,
+      "grad_norm": 0.6335933208465576,
+      "learning_rate": 7.222858995547446e-06,
+      "loss": 0.2363,
+      "step": 15770
+    },
+    {
+      "epoch": 0.5897609645535902,
+      "grad_norm": 0.40368562936782837,
+      "learning_rate": 7.2172179048354e-06,
+      "loss": 0.3412,
+      "step": 15775
+    },
+    {
+      "epoch": 0.58994789354394,
+      "grad_norm": 0.41344451904296875,
+      "learning_rate": 7.211577773825157e-06,
+      "loss": 0.3029,
+      "step": 15780
+    },
+    {
+      "epoch": 0.5901348225342898,
+      "grad_norm": 0.15166114270687103,
+      "learning_rate": 7.205938604461846e-06,
+      "loss": 0.2663,
+      "step": 15785
+    },
+    {
+      "epoch": 0.5903217515246396,
+      "grad_norm": 0.18254537880420685,
+      "learning_rate": 7.2003003986902474e-06,
+      "loss": 0.2779,
+      "step": 15790
+    },
+    {
+      "epoch": 0.5905086805149894,
+      "grad_norm": 0.6421458125114441,
+      "learning_rate": 7.19466315845482e-06,
+      "loss": 0.2935,
+      "step": 15795
+    },
+    {
+      "epoch": 0.5906956095053392,
+      "grad_norm": 0.296409010887146,
+      "learning_rate": 7.189026885699688e-06,
+      "loss": 0.2955,
+      "step": 15800
+    },
+    {
+      "epoch": 0.590882538495689,
+      "grad_norm": 0.444355845451355,
+      "learning_rate": 7.183391582368637e-06,
+      "loss": 0.2687,
+      "step": 15805
+    },
+    {
+      "epoch": 0.5910694674860387,
+      "grad_norm": 0.2945992052555084,
+      "learning_rate": 7.177757250405126e-06,
+      "loss": 0.2612,
+      "step": 15810
+    },
+    {
+      "epoch": 0.5912563964763885,
+      "grad_norm": 0.2884933650493622,
+      "learning_rate": 7.172123891752268e-06,
+      "loss": 0.2812,
+      "step": 15815
+    },
+    {
+      "epoch": 0.5914433254667383,
+      "grad_norm": 0.5446357727050781,
+      "learning_rate": 7.166491508352853e-06,
+      "loss": 0.2699,
+      "step": 15820
+    },
+    {
+      "epoch": 0.5916302544570882,
+      "grad_norm": 0.3308735489845276,
+      "learning_rate": 7.160860102149323e-06,
+      "loss": 0.2325,
+      "step": 15825
+    },
+    {
+      "epoch": 0.5918171834474379,
+      "grad_norm": 0.5215713977813721,
+      "learning_rate": 7.155229675083797e-06,
+      "loss": 0.269,
+      "step": 15830
+    },
+    {
+      "epoch": 0.5920041124377877,
+      "grad_norm": 0.39702996611595154,
+      "learning_rate": 7.1496002290980415e-06,
+      "loss": 0.3225,
+      "step": 15835
+    },
+    {
+      "epoch": 0.5921910414281375,
+      "grad_norm": 0.3645925521850586,
+      "learning_rate": 7.14397176613349e-06,
+      "loss": 0.2705,
+      "step": 15840
+    },
+    {
+      "epoch": 0.5923779704184873,
+      "grad_norm": 0.32077428698539734,
+      "learning_rate": 7.138344288131245e-06,
+      "loss": 0.3051,
+      "step": 15845
+    },
+    {
+      "epoch": 0.592564899408837,
+      "grad_norm": 0.5416418313980103,
+      "learning_rate": 7.132717797032056e-06,
+      "loss": 0.3141,
+      "step": 15850
+    },
+    {
+      "epoch": 0.5927518283991868,
+      "grad_norm": 0.5128157734870911,
+      "learning_rate": 7.127092294776343e-06,
+      "loss": 0.3008,
+      "step": 15855
+    },
+    {
+      "epoch": 0.5929387573895366,
+      "grad_norm": 0.820091962814331,
+      "learning_rate": 7.121467783304174e-06,
+      "loss": 0.3023,
+      "step": 15860
+    },
+    {
+      "epoch": 0.5931256863798864,
+      "grad_norm": 0.2677062153816223,
+      "learning_rate": 7.1158442645552896e-06,
+      "loss": 0.2956,
+      "step": 15865
+    },
+    {
+      "epoch": 0.5933126153702363,
+      "grad_norm": 0.3159048855304718,
+      "learning_rate": 7.110221740469074e-06,
+      "loss": 0.3155,
+      "step": 15870
+    },
+    {
+      "epoch": 0.593499544360586,
+      "grad_norm": 0.2539639472961426,
+      "learning_rate": 7.10460021298458e-06,
+      "loss": 0.3094,
+      "step": 15875
+    },
+    {
+      "epoch": 0.5936864733509358,
+      "grad_norm": 0.36676567792892456,
+      "learning_rate": 7.098979684040508e-06,
+      "loss": 0.2947,
+      "step": 15880
+    },
+    {
+      "epoch": 0.5938734023412856,
+      "grad_norm": 0.3667091131210327,
+      "learning_rate": 7.093360155575218e-06,
+      "loss": 0.2889,
+      "step": 15885
+    },
+    {
+      "epoch": 0.5940603313316354,
+      "grad_norm": 0.6631130576133728,
+      "learning_rate": 7.087741629526726e-06,
+      "loss": 0.2206,
+      "step": 15890
+    },
+    {
+      "epoch": 0.5942472603219852,
+      "grad_norm": 0.7274545431137085,
+      "learning_rate": 7.082124107832695e-06,
+      "loss": 0.3667,
+      "step": 15895
+    },
+    {
+      "epoch": 0.5944341893123349,
+      "grad_norm": 0.5482187867164612,
+      "learning_rate": 7.076507592430457e-06,
+      "loss": 0.3779,
+      "step": 15900
+    },
+    {
+      "epoch": 0.5946211183026847,
+      "grad_norm": 0.4760439991950989,
+      "learning_rate": 7.070892085256978e-06,
+      "loss": 0.2652,
+      "step": 15905
+    },
+    {
+      "epoch": 0.5948080472930346,
+      "grad_norm": 0.27402257919311523,
+      "learning_rate": 7.065277588248893e-06,
+      "loss": 0.2663,
+      "step": 15910
+    },
+    {
+      "epoch": 0.5949949762833844,
+      "grad_norm": 0.5011516213417053,
+      "learning_rate": 7.059664103342473e-06,
+      "loss": 0.364,
+      "step": 15915
+    },
+    {
+      "epoch": 0.5951819052737342,
+      "grad_norm": 0.3318321108818054,
+      "learning_rate": 7.0540516324736556e-06,
+      "loss": 0.3262,
+      "step": 15920
+    },
+    {
+      "epoch": 0.5953688342640839,
+      "grad_norm": 0.34024515748023987,
+      "learning_rate": 7.0484401775780175e-06,
+      "loss": 0.3343,
+      "step": 15925
+    },
+    {
+      "epoch": 0.5955557632544337,
+      "grad_norm": 0.2316112518310547,
+      "learning_rate": 7.0428297405907865e-06,
+      "loss": 0.3235,
+      "step": 15930
+    },
+    {
+      "epoch": 0.5957426922447835,
+      "grad_norm": 0.3854818046092987,
+      "learning_rate": 7.0372203234468474e-06,
+      "loss": 0.3301,
+      "step": 15935
+    },
+    {
+      "epoch": 0.5959296212351333,
+      "grad_norm": 0.34826406836509705,
+      "learning_rate": 7.031611928080721e-06,
+      "loss": 0.2321,
+      "step": 15940
+    },
+    {
+      "epoch": 0.596116550225483,
+      "grad_norm": 0.5255212783813477,
+      "learning_rate": 7.02600455642659e-06,
+      "loss": 0.2714,
+      "step": 15945
+    },
+    {
+      "epoch": 0.5963034792158329,
+      "grad_norm": 0.4153469204902649,
+      "learning_rate": 7.020398210418269e-06,
+      "loss": 0.3229,
+      "step": 15950
+    },
+    {
+      "epoch": 0.5964904082061827,
+      "grad_norm": 0.532195508480072,
+      "learning_rate": 7.014792891989232e-06,
+      "loss": 0.2495,
+      "step": 15955
+    },
+    {
+      "epoch": 0.5966773371965325,
+      "grad_norm": 0.12149854749441147,
+      "learning_rate": 7.009188603072586e-06,
+      "loss": 0.2962,
+      "step": 15960
+    },
+    {
+      "epoch": 0.5968642661868823,
+      "grad_norm": 0.5596807599067688,
+      "learning_rate": 7.003585345601095e-06,
+      "loss": 0.342,
+      "step": 15965
+    },
+    {
+      "epoch": 0.597051195177232,
+      "grad_norm": 0.4083070456981659,
+      "learning_rate": 6.9979831215071566e-06,
+      "loss": 0.2877,
+      "step": 15970
+    },
+    {
+      "epoch": 0.5972381241675818,
+      "grad_norm": 0.5761303901672363,
+      "learning_rate": 6.9923819327228235e-06,
+      "loss": 0.2234,
+      "step": 15975
+    },
+    {
+      "epoch": 0.5974250531579316,
+      "grad_norm": 0.42254355549812317,
+      "learning_rate": 6.986781781179786e-06,
+      "loss": 0.3016,
+      "step": 15980
+    },
+    {
+      "epoch": 0.5976119821482814,
+      "grad_norm": 0.43202266097068787,
+      "learning_rate": 6.981182668809365e-06,
+      "loss": 0.3697,
+      "step": 15985
+    },
+    {
+      "epoch": 0.5977989111386313,
+      "grad_norm": 0.30627140402793884,
+      "learning_rate": 6.975584597542549e-06,
+      "loss": 0.3126,
+      "step": 15990
+    },
+    {
+      "epoch": 0.597985840128981,
+      "grad_norm": 0.5201455950737,
+      "learning_rate": 6.9699875693099415e-06,
+      "loss": 0.2483,
+      "step": 15995
+    },
+    {
+      "epoch": 0.5981727691193308,
+      "grad_norm": 0.49727630615234375,
+      "learning_rate": 6.964391586041803e-06,
+      "loss": 0.3368,
+      "step": 16000
+    },
+    {
+      "epoch": 0.5983596981096806,
+      "grad_norm": 0.3812571167945862,
+      "learning_rate": 6.95879664966802e-06,
+      "loss": 0.3413,
+      "step": 16005
+    },
+    {
+      "epoch": 0.5985466271000304,
+      "grad_norm": 0.35332605242729187,
+      "learning_rate": 6.953202762118137e-06,
+      "loss": 0.2443,
+      "step": 16010
+    },
+    {
+      "epoch": 0.5987335560903801,
+      "grad_norm": 0.45708781480789185,
+      "learning_rate": 6.947609925321314e-06,
+      "loss": 0.2457,
+      "step": 16015
+    },
+    {
+      "epoch": 0.5989204850807299,
+      "grad_norm": 0.43644988536834717,
+      "learning_rate": 6.942018141206368e-06,
+      "loss": 0.2944,
+      "step": 16020
+    },
+    {
+      "epoch": 0.5991074140710797,
+      "grad_norm": 0.4386439323425293,
+      "learning_rate": 6.9364274117017446e-06,
+      "loss": 0.3314,
+      "step": 16025
+    },
+    {
+      "epoch": 0.5992943430614296,
+      "grad_norm": 0.48612180352211,
+      "learning_rate": 6.930837738735521e-06,
+      "loss": 0.2833,
+      "step": 16030
+    },
+    {
+      "epoch": 0.5994812720517794,
+      "grad_norm": 0.5489234924316406,
+      "learning_rate": 6.925249124235423e-06,
+      "loss": 0.3497,
+      "step": 16035
+    },
+    {
+      "epoch": 0.5996682010421291,
+      "grad_norm": 0.47888511419296265,
+      "learning_rate": 6.919661570128796e-06,
+      "loss": 0.3317,
+      "step": 16040
+    },
+    {
+      "epoch": 0.5998551300324789,
+      "grad_norm": 0.12318491190671921,
+      "learning_rate": 6.914075078342632e-06,
+      "loss": 0.2215,
+      "step": 16045
+    },
+    {
+      "epoch": 0.6000420590228287,
+      "grad_norm": 0.5285260081291199,
+      "learning_rate": 6.908489650803549e-06,
+      "loss": 0.2446,
+      "step": 16050
+    },
+    {
+      "epoch": 0.6002289880131785,
+      "grad_norm": 0.34625720977783203,
+      "learning_rate": 6.902905289437807e-06,
+      "loss": 0.2358,
+      "step": 16055
+    },
+    {
+      "epoch": 0.6004159170035283,
+      "grad_norm": 0.5598260760307312,
+      "learning_rate": 6.897321996171281e-06,
+      "loss": 0.2095,
+      "step": 16060
+    },
+    {
+      "epoch": 0.600602845993878,
+      "grad_norm": 0.35202714800834656,
+      "learning_rate": 6.891739772929499e-06,
+      "loss": 0.2681,
+      "step": 16065
+    },
+    {
+      "epoch": 0.6007897749842279,
+      "grad_norm": 0.39519229531288147,
+      "learning_rate": 6.886158621637608e-06,
+      "loss": 0.2615,
+      "step": 16070
+    },
+    {
+      "epoch": 0.6009767039745777,
+      "grad_norm": 0.6391472220420837,
+      "learning_rate": 6.880578544220382e-06,
+      "loss": 0.2642,
+      "step": 16075
+    },
+    {
+      "epoch": 0.6011636329649275,
+      "grad_norm": 0.3933602571487427,
+      "learning_rate": 6.874999542602237e-06,
+      "loss": 0.2837,
+      "step": 16080
+    },
+    {
+      "epoch": 0.6013505619552773,
+      "grad_norm": 0.6235082745552063,
+      "learning_rate": 6.8694216187072015e-06,
+      "loss": 0.3124,
+      "step": 16085
+    },
+    {
+      "epoch": 0.601537490945627,
+      "grad_norm": 0.46364444494247437,
+      "learning_rate": 6.863844774458954e-06,
+      "loss": 0.2186,
+      "step": 16090
+    },
+    {
+      "epoch": 0.6017244199359768,
+      "grad_norm": 0.37542879581451416,
+      "learning_rate": 6.8582690117807784e-06,
+      "loss": 0.28,
+      "step": 16095
+    },
+    {
+      "epoch": 0.6019113489263266,
+      "grad_norm": 0.4766859710216522,
+      "learning_rate": 6.852694332595601e-06,
+      "loss": 0.2415,
+      "step": 16100
+    },
+    {
+      "epoch": 0.6020982779166764,
+      "grad_norm": 0.384067177772522,
+      "learning_rate": 6.847120738825962e-06,
+      "loss": 0.2473,
+      "step": 16105
+    },
+    {
+      "epoch": 0.6022852069070261,
+      "grad_norm": 0.3013162314891815,
+      "learning_rate": 6.841548232394041e-06,
+      "loss": 0.2942,
+      "step": 16110
+    },
+    {
+      "epoch": 0.602472135897376,
+      "grad_norm": 0.4402746558189392,
+      "learning_rate": 6.835976815221637e-06,
+      "loss": 0.2965,
+      "step": 16115
+    },
+    {
+      "epoch": 0.6026590648877258,
+      "grad_norm": 0.42539361119270325,
+      "learning_rate": 6.830406489230162e-06,
+      "loss": 0.2081,
+      "step": 16120
+    },
+    {
+      "epoch": 0.6028459938780756,
+      "grad_norm": 0.47078409790992737,
+      "learning_rate": 6.824837256340674e-06,
+      "loss": 0.2934,
+      "step": 16125
+    },
+    {
+      "epoch": 0.6030329228684254,
+      "grad_norm": 0.4631364345550537,
+      "learning_rate": 6.819269118473833e-06,
+      "loss": 0.2897,
+      "step": 16130
+    },
+    {
+      "epoch": 0.6032198518587751,
+      "grad_norm": 0.48256585001945496,
+      "learning_rate": 6.813702077549935e-06,
+      "loss": 0.2225,
+      "step": 16135
+    },
+    {
+      "epoch": 0.6034067808491249,
+      "grad_norm": 0.5263893604278564,
+      "learning_rate": 6.808136135488892e-06,
+      "loss": 0.2288,
+      "step": 16140
+    },
+    {
+      "epoch": 0.6035937098394747,
+      "grad_norm": 0.3626660108566284,
+      "learning_rate": 6.802571294210239e-06,
+      "loss": 0.2377,
+      "step": 16145
+    },
+    {
+      "epoch": 0.6037806388298245,
+      "grad_norm": 1.2106562852859497,
+      "learning_rate": 6.797007555633124e-06,
+      "loss": 0.2988,
+      "step": 16150
+    },
+    {
+      "epoch": 0.6039675678201744,
+      "grad_norm": 0.2370201051235199,
+      "learning_rate": 6.791444921676327e-06,
+      "loss": 0.2449,
+      "step": 16155
+    },
+    {
+      "epoch": 0.6041544968105241,
+      "grad_norm": 0.5193468928337097,
+      "learning_rate": 6.785883394258241e-06,
+      "loss": 0.2805,
+      "step": 16160
+    },
+    {
+      "epoch": 0.6043414258008739,
+      "grad_norm": 0.5827836990356445,
+      "learning_rate": 6.780322975296877e-06,
+      "loss": 0.2585,
+      "step": 16165
+    },
+    {
+      "epoch": 0.6045283547912237,
+      "grad_norm": 0.22835935652256012,
+      "learning_rate": 6.7747636667098645e-06,
+      "loss": 0.2602,
+      "step": 16170
+    },
+    {
+      "epoch": 0.6047152837815735,
+      "grad_norm": 0.5145224928855896,
+      "learning_rate": 6.769205470414445e-06,
+      "loss": 0.278,
+      "step": 16175
+    },
+    {
+      "epoch": 0.6049022127719232,
+      "grad_norm": 0.26391491293907166,
+      "learning_rate": 6.763648388327488e-06,
+      "loss": 0.2507,
+      "step": 16180
+    },
+    {
+      "epoch": 0.605089141762273,
+      "grad_norm": 0.23646554350852966,
+      "learning_rate": 6.758092422365468e-06,
+      "loss": 0.2388,
+      "step": 16185
+    },
+    {
+      "epoch": 0.6052760707526228,
+      "grad_norm": 0.4633139669895172,
+      "learning_rate": 6.75253757444448e-06,
+      "loss": 0.3711,
+      "step": 16190
+    },
+    {
+      "epoch": 0.6054629997429727,
+      "grad_norm": 0.614550769329071,
+      "learning_rate": 6.746983846480226e-06,
+      "loss": 0.3509,
+      "step": 16195
+    },
+    {
+      "epoch": 0.6056499287333225,
+      "grad_norm": 1.0739442110061646,
+      "learning_rate": 6.7414312403880345e-06,
+      "loss": 0.351,
+      "step": 16200
+    },
+    {
+      "epoch": 0.6058368577236722,
+      "grad_norm": 0.317594051361084,
+      "learning_rate": 6.735879758082841e-06,
+      "loss": 0.341,
+      "step": 16205
+    },
+    {
+      "epoch": 0.606023786714022,
+      "grad_norm": 0.32532641291618347,
+      "learning_rate": 6.730329401479189e-06,
+      "loss": 0.2996,
+      "step": 16210
+    },
+    {
+      "epoch": 0.6062107157043718,
+      "grad_norm": 0.4760509431362152,
+      "learning_rate": 6.724780172491241e-06,
+      "loss": 0.2035,
+      "step": 16215
+    },
+    {
+      "epoch": 0.6063976446947216,
+      "grad_norm": 0.4633803963661194,
+      "learning_rate": 6.71923207303276e-06,
+      "loss": 0.3453,
+      "step": 16220
+    },
+    {
+      "epoch": 0.6065845736850713,
+      "grad_norm": 0.27814042568206787,
+      "learning_rate": 6.713685105017135e-06,
+      "loss": 0.3061,
+      "step": 16225
+    },
+    {
+      "epoch": 0.6067715026754211,
+      "grad_norm": 0.331685334444046,
+      "learning_rate": 6.708139270357348e-06,
+      "loss": 0.2412,
+      "step": 16230
+    },
+    {
+      "epoch": 0.606958431665771,
+      "grad_norm": 0.2586578130722046,
+      "learning_rate": 6.702594570966008e-06,
+      "loss": 0.2551,
+      "step": 16235
+    },
+    {
+      "epoch": 0.6071453606561208,
+      "grad_norm": 0.35025131702423096,
+      "learning_rate": 6.697051008755315e-06,
+      "loss": 0.2946,
+      "step": 16240
+    },
+    {
+      "epoch": 0.6073322896464706,
+      "grad_norm": 0.403495192527771,
+      "learning_rate": 6.691508585637085e-06,
+      "loss": 0.2499,
+      "step": 16245
+    },
+    {
+      "epoch": 0.6075192186368203,
+      "grad_norm": 0.6013374328613281,
+      "learning_rate": 6.6859673035227495e-06,
+      "loss": 0.2661,
+      "step": 16250
+    },
+    {
+      "epoch": 0.6077061476271701,
+      "grad_norm": 0.09546211361885071,
+      "learning_rate": 6.680427164323329e-06,
+      "loss": 0.2791,
+      "step": 16255
+    },
+    {
+      "epoch": 0.6078930766175199,
+      "grad_norm": 0.38992223143577576,
+      "learning_rate": 6.674888169949463e-06,
+      "loss": 0.3381,
+      "step": 16260
+    },
+    {
+      "epoch": 0.6080800056078697,
+      "grad_norm": 0.33082830905914307,
+      "learning_rate": 6.669350322311388e-06,
+      "loss": 0.3829,
+      "step": 16265
+    },
+    {
+      "epoch": 0.6082669345982195,
+      "grad_norm": 0.30394282937049866,
+      "learning_rate": 6.663813623318954e-06,
+      "loss": 0.287,
+      "step": 16270
+    },
+    {
+      "epoch": 0.6084538635885693,
+      "grad_norm": 0.27760547399520874,
+      "learning_rate": 6.658278074881605e-06,
+      "loss": 0.2794,
+      "step": 16275
+    },
+    {
+      "epoch": 0.6086407925789191,
+      "grad_norm": 0.7337043881416321,
+      "learning_rate": 6.652743678908399e-06,
+      "loss": 0.2566,
+      "step": 16280
+    },
+    {
+      "epoch": 0.6088277215692689,
+      "grad_norm": 0.36123204231262207,
+      "learning_rate": 6.647210437307985e-06,
+      "loss": 0.2533,
+      "step": 16285
+    },
+    {
+      "epoch": 0.6090146505596187,
+      "grad_norm": 0.4499099552631378,
+      "learning_rate": 6.641678351988619e-06,
+      "loss": 0.314,
+      "step": 16290
+    },
+    {
+      "epoch": 0.6092015795499685,
+      "grad_norm": 0.3133924901485443,
+      "learning_rate": 6.6361474248581655e-06,
+      "loss": 0.2611,
+      "step": 16295
+    },
+    {
+      "epoch": 0.6093885085403182,
+      "grad_norm": 0.39866819977760315,
+      "learning_rate": 6.630617657824078e-06,
+      "loss": 0.2835,
+      "step": 16300
+    },
+    {
+      "epoch": 0.609575437530668,
+      "grad_norm": 0.45732995867729187,
+      "learning_rate": 6.625089052793417e-06,
+      "loss": 0.3571,
+      "step": 16305
+    },
+    {
+      "epoch": 0.6097623665210178,
+      "grad_norm": 0.41225141286849976,
+      "learning_rate": 6.619561611672834e-06,
+      "loss": 0.3494,
+      "step": 16310
+    },
+    {
+      "epoch": 0.6099492955113677,
+      "grad_norm": 0.5021333694458008,
+      "learning_rate": 6.6140353363685914e-06,
+      "loss": 0.2647,
+      "step": 16315
+    },
+    {
+      "epoch": 0.6101362245017175,
+      "grad_norm": 0.2890709936618805,
+      "learning_rate": 6.6085102287865385e-06,
+      "loss": 0.278,
+      "step": 16320
+    },
+    {
+      "epoch": 0.6103231534920672,
+      "grad_norm": 0.5777745842933655,
+      "learning_rate": 6.602986290832134e-06,
+      "loss": 0.2815,
+      "step": 16325
+    },
+    {
+      "epoch": 0.610510082482417,
+      "grad_norm": 0.4748448133468628,
+      "learning_rate": 6.597463524410418e-06,
+      "loss": 0.2811,
+      "step": 16330
+    },
+    {
+      "epoch": 0.6106970114727668,
+      "grad_norm": 0.5496982336044312,
+      "learning_rate": 6.591941931426036e-06,
+      "loss": 0.2409,
+      "step": 16335
+    },
+    {
+      "epoch": 0.6108839404631166,
+      "grad_norm": 0.2635287046432495,
+      "learning_rate": 6.5864215137832325e-06,
+      "loss": 0.311,
+      "step": 16340
+    },
+    {
+      "epoch": 0.6110708694534663,
+      "grad_norm": 0.2311626374721527,
+      "learning_rate": 6.580902273385834e-06,
+      "loss": 0.2865,
+      "step": 16345
+    },
+    {
+      "epoch": 0.6112577984438161,
+      "grad_norm": 0.39265650510787964,
+      "learning_rate": 6.575384212137275e-06,
+      "loss": 0.2652,
+      "step": 16350
+    },
+    {
+      "epoch": 0.6114447274341659,
+      "grad_norm": 0.35634520649909973,
+      "learning_rate": 6.569867331940571e-06,
+      "loss": 0.2976,
+      "step": 16355
+    },
+    {
+      "epoch": 0.6116316564245158,
+      "grad_norm": 0.7832450866699219,
+      "learning_rate": 6.564351634698343e-06,
+      "loss": 0.2907,
+      "step": 16360
+    },
+    {
+      "epoch": 0.6118185854148656,
+      "grad_norm": 0.28447359800338745,
+      "learning_rate": 6.558837122312787e-06,
+      "loss": 0.2886,
+      "step": 16365
+    },
+    {
+      "epoch": 0.6120055144052153,
+      "grad_norm": 0.9879940748214722,
+      "learning_rate": 6.553323796685709e-06,
+      "loss": 0.3565,
+      "step": 16370
+    },
+    {
+      "epoch": 0.6121924433955651,
+      "grad_norm": 0.35768356919288635,
+      "learning_rate": 6.547811659718492e-06,
+      "loss": 0.3012,
+      "step": 16375
+    },
+    {
+      "epoch": 0.6123793723859149,
+      "grad_norm": 0.3880248963832855,
+      "learning_rate": 6.542300713312113e-06,
+      "loss": 0.372,
+      "step": 16380
+    },
+    {
+      "epoch": 0.6125663013762647,
+      "grad_norm": 0.1782693862915039,
+      "learning_rate": 6.536790959367149e-06,
+      "loss": 0.2869,
+      "step": 16385
+    },
+    {
+      "epoch": 0.6127532303666144,
+      "grad_norm": 0.4560245871543884,
+      "learning_rate": 6.5312823997837425e-06,
+      "loss": 0.2764,
+      "step": 16390
+    },
+    {
+      "epoch": 0.6129401593569642,
+      "grad_norm": 0.3002925217151642,
+      "learning_rate": 6.525775036461652e-06,
+      "loss": 0.2596,
+      "step": 16395
+    },
+    {
+      "epoch": 0.6131270883473141,
+      "grad_norm": 0.45390433073043823,
+      "learning_rate": 6.520268871300198e-06,
+      "loss": 0.2895,
+      "step": 16400
+    },
+    {
+      "epoch": 0.6133140173376639,
+      "grad_norm": 0.19361938536167145,
+      "learning_rate": 6.514763906198307e-06,
+      "loss": 0.3229,
+      "step": 16405
+    },
+    {
+      "epoch": 0.6135009463280137,
+      "grad_norm": 0.6821458339691162,
+      "learning_rate": 6.509260143054474e-06,
+      "loss": 0.2455,
+      "step": 16410
+    },
+    {
+      "epoch": 0.6136878753183634,
+      "grad_norm": 0.405543714761734,
+      "learning_rate": 6.503757583766802e-06,
+      "loss": 0.3357,
+      "step": 16415
+    },
+    {
+      "epoch": 0.6138748043087132,
+      "grad_norm": 0.6047316789627075,
+      "learning_rate": 6.4982562302329535e-06,
+      "loss": 0.3021,
+      "step": 16420
+    },
+    {
+      "epoch": 0.614061733299063,
+      "grad_norm": 0.49325546622276306,
+      "learning_rate": 6.492756084350196e-06,
+      "loss": 0.3521,
+      "step": 16425
+    },
+    {
+      "epoch": 0.6142486622894128,
+      "grad_norm": 0.8090613484382629,
+      "learning_rate": 6.4872571480153725e-06,
+      "loss": 0.3772,
+      "step": 16430
+    },
+    {
+      "epoch": 0.6144355912797625,
+      "grad_norm": 0.4884081780910492,
+      "learning_rate": 6.4817594231249e-06,
+      "loss": 0.2789,
+      "step": 16435
+    },
+    {
+      "epoch": 0.6146225202701124,
+      "grad_norm": 0.4141273498535156,
+      "learning_rate": 6.476262911574797e-06,
+      "loss": 0.3313,
+      "step": 16440
+    },
+    {
+      "epoch": 0.6148094492604622,
+      "grad_norm": 0.6283908486366272,
+      "learning_rate": 6.470767615260647e-06,
+      "loss": 0.2758,
+      "step": 16445
+    },
+    {
+      "epoch": 0.614996378250812,
+      "grad_norm": 0.3659379184246063,
+      "learning_rate": 6.465273536077623e-06,
+      "loss": 0.2827,
+      "step": 16450
+    },
+    {
+      "epoch": 0.6151833072411618,
+      "grad_norm": 0.4448320269584656,
+      "learning_rate": 6.459780675920468e-06,
+      "loss": 0.2774,
+      "step": 16455
+    },
+    {
+      "epoch": 0.6153702362315115,
+      "grad_norm": 0.43518292903900146,
+      "learning_rate": 6.454289036683523e-06,
+      "loss": 0.2542,
+      "step": 16460
+    },
+    {
+      "epoch": 0.6155571652218613,
+      "grad_norm": 0.26366111636161804,
+      "learning_rate": 6.448798620260688e-06,
+      "loss": 0.2958,
+      "step": 16465
+    },
+    {
+      "epoch": 0.6157440942122111,
+      "grad_norm": 0.5211677551269531,
+      "learning_rate": 6.443309428545457e-06,
+      "loss": 0.2952,
+      "step": 16470
+    },
+    {
+      "epoch": 0.6159310232025609,
+      "grad_norm": 0.2922613322734833,
+      "learning_rate": 6.4378214634308925e-06,
+      "loss": 0.194,
+      "step": 16475
+    },
+    {
+      "epoch": 0.6161179521929108,
+      "grad_norm": 0.3419044613838196,
+      "learning_rate": 6.4323347268096316e-06,
+      "loss": 0.2536,
+      "step": 16480
+    },
+    {
+      "epoch": 0.6163048811832605,
+      "grad_norm": 0.37760069966316223,
+      "learning_rate": 6.426849220573901e-06,
+      "loss": 0.3227,
+      "step": 16485
+    },
+    {
+      "epoch": 0.6164918101736103,
+      "grad_norm": 0.8805325627326965,
+      "learning_rate": 6.4213649466154894e-06,
+      "loss": 0.3393,
+      "step": 16490
+    },
+    {
+      "epoch": 0.6166787391639601,
+      "grad_norm": 0.47188910841941833,
+      "learning_rate": 6.415881906825767e-06,
+      "loss": 0.2425,
+      "step": 16495
+    },
+    {
+      "epoch": 0.6168656681543099,
+      "grad_norm": 0.47129133343696594,
+      "learning_rate": 6.4104001030956755e-06,
+      "loss": 0.3161,
+      "step": 16500
+    },
+    {
+      "epoch": 0.6170525971446597,
+      "grad_norm": 0.43512147665023804,
+      "learning_rate": 6.404919537315737e-06,
+      "loss": 0.2694,
+      "step": 16505
+    },
+    {
+      "epoch": 0.6172395261350094,
+      "grad_norm": 0.1717224419116974,
+      "learning_rate": 6.399440211376033e-06,
+      "loss": 0.3666,
+      "step": 16510
+    },
+    {
+      "epoch": 0.6174264551253592,
+      "grad_norm": 0.6517991423606873,
+      "learning_rate": 6.393962127166233e-06,
+      "loss": 0.3249,
+      "step": 16515
+    },
+    {
+      "epoch": 0.6176133841157091,
+      "grad_norm": 0.30633100867271423,
+      "learning_rate": 6.388485286575572e-06,
+      "loss": 0.2839,
+      "step": 16520
+    },
+    {
+      "epoch": 0.6178003131060589,
+      "grad_norm": 0.5757652521133423,
+      "learning_rate": 6.383009691492847e-06,
+      "loss": 0.2825,
+      "step": 16525
+    },
+    {
+      "epoch": 0.6179872420964087,
+      "grad_norm": 0.4330870509147644,
+      "learning_rate": 6.377535343806446e-06,
+      "loss": 0.2922,
+      "step": 16530
+    },
+    {
+      "epoch": 0.6181741710867584,
+      "grad_norm": 0.279230535030365,
+      "learning_rate": 6.372062245404302e-06,
+      "loss": 0.2907,
+      "step": 16535
+    },
+    {
+      "epoch": 0.6183611000771082,
+      "grad_norm": 0.6289684176445007,
+      "learning_rate": 6.366590398173942e-06,
+      "loss": 0.4383,
+      "step": 16540
+    },
+    {
+      "epoch": 0.618548029067458,
+      "grad_norm": 0.6121999621391296,
+      "learning_rate": 6.3611198040024405e-06,
+      "loss": 0.2341,
+      "step": 16545
+    },
+    {
+      "epoch": 0.6187349580578078,
+      "grad_norm": 0.421684592962265,
+      "learning_rate": 6.355650464776453e-06,
+      "loss": 0.2145,
+      "step": 16550
+    },
+    {
+      "epoch": 0.6189218870481575,
+      "grad_norm": 0.45362263917922974,
+      "learning_rate": 6.350182382382193e-06,
+      "loss": 0.3438,
+      "step": 16555
+    },
+    {
+      "epoch": 0.6191088160385074,
+      "grad_norm": 0.35668495297431946,
+      "learning_rate": 6.34471555870545e-06,
+      "loss": 0.2971,
+      "step": 16560
+    },
+    {
+      "epoch": 0.6192957450288572,
+      "grad_norm": 0.3048703372478485,
+      "learning_rate": 6.339249995631575e-06,
+      "loss": 0.2846,
+      "step": 16565
+    },
+    {
+      "epoch": 0.619482674019207,
+      "grad_norm": 0.42319127917289734,
+      "learning_rate": 6.33378569504548e-06,
+      "loss": 0.2441,
+      "step": 16570
+    },
+    {
+      "epoch": 0.6196696030095568,
+      "grad_norm": 0.40297672152519226,
+      "learning_rate": 6.328322658831652e-06,
+      "loss": 0.2304,
+      "step": 16575
+    },
+    {
+      "epoch": 0.6198565319999065,
+      "grad_norm": 0.32730260491371155,
+      "learning_rate": 6.322860888874129e-06,
+      "loss": 0.2582,
+      "step": 16580
+    },
+    {
+      "epoch": 0.6200434609902563,
+      "grad_norm": 0.31443408131599426,
+      "learning_rate": 6.3174003870565256e-06,
+      "loss": 0.3275,
+      "step": 16585
+    },
+    {
+      "epoch": 0.6202303899806061,
+      "grad_norm": 0.440044641494751,
+      "learning_rate": 6.311941155262007e-06,
+      "loss": 0.295,
+      "step": 16590
+    },
+    {
+      "epoch": 0.6204173189709559,
+      "grad_norm": 0.396515816450119,
+      "learning_rate": 6.306483195373309e-06,
+      "loss": 0.3439,
+      "step": 16595
+    },
+    {
+      "epoch": 0.6206042479613056,
+      "grad_norm": 0.43022897839546204,
+      "learning_rate": 6.301026509272721e-06,
+      "loss": 0.334,
+      "step": 16600
+    },
+    {
+      "epoch": 0.6207911769516555,
+      "grad_norm": 0.20897158980369568,
+      "learning_rate": 6.2955710988421e-06,
+      "loss": 0.2489,
+      "step": 16605
+    },
+    {
+      "epoch": 0.6209781059420053,
+      "grad_norm": 0.43990305066108704,
+      "learning_rate": 6.290116965962867e-06,
+      "loss": 0.2715,
+      "step": 16610
+    },
+    {
+      "epoch": 0.6211650349323551,
+      "grad_norm": 0.21257735788822174,
+      "learning_rate": 6.284664112515988e-06,
+      "loss": 0.2052,
+      "step": 16615
+    },
+    {
+      "epoch": 0.6213519639227049,
+      "grad_norm": 0.4820898175239563,
+      "learning_rate": 6.279212540382e-06,
+      "loss": 0.3537,
+      "step": 16620
+    },
+    {
+      "epoch": 0.6215388929130546,
+      "grad_norm": 0.2106875628232956,
+      "learning_rate": 6.273762251440991e-06,
+      "loss": 0.2318,
+      "step": 16625
+    },
+    {
+      "epoch": 0.6217258219034044,
+      "grad_norm": 0.3450741767883301,
+      "learning_rate": 6.268313247572614e-06,
+      "loss": 0.3046,
+      "step": 16630
+    },
+    {
+      "epoch": 0.6219127508937542,
+      "grad_norm": 0.2806203365325928,
+      "learning_rate": 6.262865530656069e-06,
+      "loss": 0.2983,
+      "step": 16635
+    },
+    {
+      "epoch": 0.622099679884104,
+      "grad_norm": 0.5870185494422913,
+      "learning_rate": 6.257419102570122e-06,
+      "loss": 0.2252,
+      "step": 16640
+    },
+    {
+      "epoch": 0.6222866088744539,
+      "grad_norm": 0.33746689558029175,
+      "learning_rate": 6.251973965193085e-06,
+      "loss": 0.2831,
+      "step": 16645
+    },
+    {
+      "epoch": 0.6224735378648036,
+      "grad_norm": 0.4964042901992798,
+      "learning_rate": 6.246530120402833e-06,
+      "loss": 0.3002,
+      "step": 16650
+    },
+    {
+      "epoch": 0.6226604668551534,
+      "grad_norm": 0.31772905588150024,
+      "learning_rate": 6.241087570076796e-06,
+      "loss": 0.3224,
+      "step": 16655
+    },
+    {
+      "epoch": 0.6228473958455032,
+      "grad_norm": 0.547088623046875,
+      "learning_rate": 6.235646316091945e-06,
+      "loss": 0.2861,
+      "step": 16660
+    },
+    {
+      "epoch": 0.623034324835853,
+      "grad_norm": 0.5915294885635376,
+      "learning_rate": 6.230206360324822e-06,
+      "loss": 0.3202,
+      "step": 16665
+    },
+    {
+      "epoch": 0.6232212538262027,
+      "grad_norm": 0.3158590793609619,
+      "learning_rate": 6.224767704651502e-06,
+      "loss": 0.3383,
+      "step": 16670
+    },
+    {
+      "epoch": 0.6234081828165525,
+      "grad_norm": 0.4911237061023712,
+      "learning_rate": 6.219330350947632e-06,
+      "loss": 0.247,
+      "step": 16675
+    },
+    {
+      "epoch": 0.6235951118069023,
+      "grad_norm": 0.253912091255188,
+      "learning_rate": 6.213894301088388e-06,
+      "loss": 0.2208,
+      "step": 16680
+    },
+    {
+      "epoch": 0.6237820407972522,
+      "grad_norm": 0.3339253067970276,
+      "learning_rate": 6.208459556948519e-06,
+      "loss": 0.2759,
+      "step": 16685
+    },
+    {
+      "epoch": 0.623968969787602,
+      "grad_norm": 0.6338766813278198,
+      "learning_rate": 6.2030261204023055e-06,
+      "loss": 0.2522,
+      "step": 16690
+    },
+    {
+      "epoch": 0.6241558987779517,
+      "grad_norm": 0.3243858814239502,
+      "learning_rate": 6.197593993323583e-06,
+      "loss": 0.3472,
+      "step": 16695
+    },
+    {
+      "epoch": 0.6243428277683015,
+      "grad_norm": 0.5515966415405273,
+      "learning_rate": 6.192163177585745e-06,
+      "loss": 0.2825,
+      "step": 16700
+    },
+    {
+      "epoch": 0.6245297567586513,
+      "grad_norm": 0.3765512704849243,
+      "learning_rate": 6.1867336750617155e-06,
+      "loss": 0.2992,
+      "step": 16705
+    },
+    {
+      "epoch": 0.6247166857490011,
+      "grad_norm": 0.3543432950973511,
+      "learning_rate": 6.181305487623981e-06,
+      "loss": 0.3033,
+      "step": 16710
+    },
+    {
+      "epoch": 0.6249036147393509,
+      "grad_norm": 0.29796430468559265,
+      "learning_rate": 6.175878617144559e-06,
+      "loss": 0.2713,
+      "step": 16715
+    },
+    {
+      "epoch": 0.6250905437297006,
+      "grad_norm": 0.38573959469795227,
+      "learning_rate": 6.170453065495032e-06,
+      "loss": 0.2466,
+      "step": 16720
+    },
+    {
+      "epoch": 0.6252774727200505,
+      "grad_norm": 0.44721537828445435,
+      "learning_rate": 6.165028834546507e-06,
+      "loss": 0.2684,
+      "step": 16725
+    },
+    {
+      "epoch": 0.6254644017104003,
+      "grad_norm": 0.290237694978714,
+      "learning_rate": 6.1596059261696564e-06,
+      "loss": 0.266,
+      "step": 16730
+    },
+    {
+      "epoch": 0.6256513307007501,
+      "grad_norm": 0.36702224612236023,
+      "learning_rate": 6.154184342234678e-06,
+      "loss": 0.2362,
+      "step": 16735
+    },
+    {
+      "epoch": 0.6258382596910999,
+      "grad_norm": 0.4877489507198334,
+      "learning_rate": 6.148764084611325e-06,
+      "loss": 0.2409,
+      "step": 16740
+    },
+    {
+      "epoch": 0.6260251886814496,
+      "grad_norm": 0.7535521388053894,
+      "learning_rate": 6.143345155168885e-06,
+      "loss": 0.2288,
+      "step": 16745
+    },
+    {
+      "epoch": 0.6262121176717994,
+      "grad_norm": 0.2947673201560974,
+      "learning_rate": 6.137927555776194e-06,
+      "loss": 0.2387,
+      "step": 16750
+    },
+    {
+      "epoch": 0.6263990466621492,
+      "grad_norm": 0.4088467061519623,
+      "learning_rate": 6.1325112883016306e-06,
+      "loss": 0.2728,
+      "step": 16755
+    },
+    {
+      "epoch": 0.626585975652499,
+      "grad_norm": 0.5079031586647034,
+      "learning_rate": 6.1270963546131005e-06,
+      "loss": 0.2597,
+      "step": 16760
+    },
+    {
+      "epoch": 0.6267729046428488,
+      "grad_norm": 0.31118345260620117,
+      "learning_rate": 6.121682756578069e-06,
+      "loss": 0.3091,
+      "step": 16765
+    },
+    {
+      "epoch": 0.6269598336331986,
+      "grad_norm": 0.7571595311164856,
+      "learning_rate": 6.116270496063523e-06,
+      "loss": 0.2882,
+      "step": 16770
+    },
+    {
+      "epoch": 0.6271467626235484,
+      "grad_norm": 0.3171136975288391,
+      "learning_rate": 6.110859574936006e-06,
+      "loss": 0.2565,
+      "step": 16775
+    },
+    {
+      "epoch": 0.6273336916138982,
+      "grad_norm": 0.448777973651886,
+      "learning_rate": 6.105449995061579e-06,
+      "loss": 0.2222,
+      "step": 16780
+    },
+    {
+      "epoch": 0.627520620604248,
+      "grad_norm": 0.44783681631088257,
+      "learning_rate": 6.1000417583058595e-06,
+      "loss": 0.3105,
+      "step": 16785
+    },
+    {
+      "epoch": 0.6277075495945977,
+      "grad_norm": 0.24276278913021088,
+      "learning_rate": 6.094634866533984e-06,
+      "loss": 0.2999,
+      "step": 16790
+    },
+    {
+      "epoch": 0.6278944785849475,
+      "grad_norm": 0.25277918577194214,
+      "learning_rate": 6.089229321610641e-06,
+      "loss": 0.2993,
+      "step": 16795
+    },
+    {
+      "epoch": 0.6280814075752973,
+      "grad_norm": 0.32562288641929626,
+      "learning_rate": 6.083825125400052e-06,
+      "loss": 0.2675,
+      "step": 16800
+    },
+    {
+      "epoch": 0.6282683365656472,
+      "grad_norm": 0.5652545094490051,
+      "learning_rate": 6.078422279765961e-06,
+      "loss": 0.2676,
+      "step": 16805
+    },
+    {
+      "epoch": 0.628455265555997,
+      "grad_norm": 0.5595022439956665,
+      "learning_rate": 6.073020786571662e-06,
+      "loss": 0.2698,
+      "step": 16810
+    },
+    {
+      "epoch": 0.6286421945463467,
+      "grad_norm": 0.18715853989124298,
+      "learning_rate": 6.067620647679966e-06,
+      "loss": 0.3081,
+      "step": 16815
+    },
+    {
+      "epoch": 0.6288291235366965,
+      "grad_norm": 0.4392125904560089,
+      "learning_rate": 6.062221864953237e-06,
+      "loss": 0.2487,
+      "step": 16820
+    },
+    {
+      "epoch": 0.6290160525270463,
+      "grad_norm": 0.542143702507019,
+      "learning_rate": 6.0568244402533525e-06,
+      "loss": 0.3002,
+      "step": 16825
+    },
+    {
+      "epoch": 0.6292029815173961,
+      "grad_norm": 0.3177799582481384,
+      "learning_rate": 6.051428375441735e-06,
+      "loss": 0.2498,
+      "step": 16830
+    },
+    {
+      "epoch": 0.6293899105077458,
+      "grad_norm": 0.731010377407074,
+      "learning_rate": 6.046033672379325e-06,
+      "loss": 0.2696,
+      "step": 16835
+    },
+    {
+      "epoch": 0.6295768394980956,
+      "grad_norm": 0.3050804138183594,
+      "learning_rate": 6.040640332926606e-06,
+      "loss": 0.2236,
+      "step": 16840
+    },
+    {
+      "epoch": 0.6297637684884454,
+      "grad_norm": 0.48626509308815,
+      "learning_rate": 6.035248358943591e-06,
+      "loss": 0.3319,
+      "step": 16845
+    },
+    {
+      "epoch": 0.6299506974787953,
+      "grad_norm": 0.4176667034626007,
+      "learning_rate": 6.0298577522898095e-06,
+      "loss": 0.2927,
+      "step": 16850
+    },
+    {
+      "epoch": 0.6301376264691451,
+      "grad_norm": 0.3971768319606781,
+      "learning_rate": 6.024468514824333e-06,
+      "loss": 0.2824,
+      "step": 16855
+    },
+    {
+      "epoch": 0.6303245554594948,
+      "grad_norm": 0.4421550929546356,
+      "learning_rate": 6.019080648405747e-06,
+      "loss": 0.2794,
+      "step": 16860
+    },
+    {
+      "epoch": 0.6305114844498446,
+      "grad_norm": 0.30311697721481323,
+      "learning_rate": 6.013694154892183e-06,
+      "loss": 0.2382,
+      "step": 16865
+    },
+    {
+      "epoch": 0.6306984134401944,
+      "grad_norm": 0.14912711083889008,
+      "learning_rate": 6.008309036141279e-06,
+      "loss": 0.2516,
+      "step": 16870
+    },
+    {
+      "epoch": 0.6308853424305442,
+      "grad_norm": 0.4212360978126526,
+      "learning_rate": 6.0029252940102154e-06,
+      "loss": 0.3549,
+      "step": 16875
+    },
+    {
+      "epoch": 0.631072271420894,
+      "grad_norm": 1.1165913343429565,
+      "learning_rate": 5.997542930355685e-06,
+      "loss": 0.3441,
+      "step": 16880
+    },
+    {
+      "epoch": 0.6312592004112437,
+      "grad_norm": 0.5724372267723083,
+      "learning_rate": 5.992161947033912e-06,
+      "loss": 0.273,
+      "step": 16885
+    },
+    {
+      "epoch": 0.6314461294015936,
+      "grad_norm": 0.39647606015205383,
+      "learning_rate": 5.9867823459006466e-06,
+      "loss": 0.2822,
+      "step": 16890
+    },
+    {
+      "epoch": 0.6316330583919434,
+      "grad_norm": 0.6420783996582031,
+      "learning_rate": 5.981404128811157e-06,
+      "loss": 0.353,
+      "step": 16895
+    },
+    {
+      "epoch": 0.6318199873822932,
+      "grad_norm": 0.21946968138217926,
+      "learning_rate": 5.976027297620237e-06,
+      "loss": 0.2933,
+      "step": 16900
+    },
+    {
+      "epoch": 0.632006916372643,
+      "grad_norm": 0.48058822751045227,
+      "learning_rate": 5.970651854182197e-06,
+      "loss": 0.2776,
+      "step": 16905
+    },
+    {
+      "epoch": 0.6321938453629927,
+      "grad_norm": 0.6671790480613708,
+      "learning_rate": 5.965277800350879e-06,
+      "loss": 0.2776,
+      "step": 16910
+    },
+    {
+      "epoch": 0.6323807743533425,
+      "grad_norm": 0.3052579164505005,
+      "learning_rate": 5.959905137979637e-06,
+      "loss": 0.3139,
+      "step": 16915
+    },
+    {
+      "epoch": 0.6325677033436923,
+      "grad_norm": 0.37527140974998474,
+      "learning_rate": 5.954533868921352e-06,
+      "loss": 0.2721,
+      "step": 16920
+    },
+    {
+      "epoch": 0.632754632334042,
+      "grad_norm": 0.3904482126235962,
+      "learning_rate": 5.949163995028418e-06,
+      "loss": 0.2949,
+      "step": 16925
+    },
+    {
+      "epoch": 0.6329415613243919,
+      "grad_norm": 0.3578369617462158,
+      "learning_rate": 5.943795518152747e-06,
+      "loss": 0.315,
+      "step": 16930
+    },
+    {
+      "epoch": 0.6331284903147417,
+      "grad_norm": 0.2731132507324219,
+      "learning_rate": 5.9384284401457824e-06,
+      "loss": 0.4051,
+      "step": 16935
+    },
+    {
+      "epoch": 0.6333154193050915,
+      "grad_norm": 0.3908912241458893,
+      "learning_rate": 5.933062762858467e-06,
+      "loss": 0.2259,
+      "step": 16940
+    },
+    {
+      "epoch": 0.6335023482954413,
+      "grad_norm": 0.6744188070297241,
+      "learning_rate": 5.927698488141273e-06,
+      "loss": 0.2906,
+      "step": 16945
+    },
+    {
+      "epoch": 0.633689277285791,
+      "grad_norm": 0.3423663377761841,
+      "learning_rate": 5.9223356178441835e-06,
+      "loss": 0.2811,
+      "step": 16950
+    },
+    {
+      "epoch": 0.6338762062761408,
+      "grad_norm": 0.4647408723831177,
+      "learning_rate": 5.9169741538167015e-06,
+      "loss": 0.2524,
+      "step": 16955
+    },
+    {
+      "epoch": 0.6340631352664906,
+      "grad_norm": 0.33333298563957214,
+      "learning_rate": 5.9116140979078364e-06,
+      "loss": 0.2098,
+      "step": 16960
+    },
+    {
+      "epoch": 0.6342500642568404,
+      "grad_norm": 0.41983723640441895,
+      "learning_rate": 5.906255451966127e-06,
+      "loss": 0.257,
+      "step": 16965
+    },
+    {
+      "epoch": 0.6344369932471903,
+      "grad_norm": 0.5163545608520508,
+      "learning_rate": 5.900898217839608e-06,
+      "loss": 0.3418,
+      "step": 16970
+    },
+    {
+      "epoch": 0.63462392223754,
+      "grad_norm": 0.5699335932731628,
+      "learning_rate": 5.895542397375837e-06,
+      "loss": 0.2937,
+      "step": 16975
+    },
+    {
+      "epoch": 0.6348108512278898,
+      "grad_norm": 0.3931114077568054,
+      "learning_rate": 5.8901879924218915e-06,
+      "loss": 0.3099,
+      "step": 16980
+    },
+    {
+      "epoch": 0.6349977802182396,
+      "grad_norm": 0.19795525074005127,
+      "learning_rate": 5.884835004824343e-06,
+      "loss": 0.3461,
+      "step": 16985
+    },
+    {
+      "epoch": 0.6351847092085894,
+      "grad_norm": 0.34459546208381653,
+      "learning_rate": 5.87948343642929e-06,
+      "loss": 0.2567,
+      "step": 16990
+    },
+    {
+      "epoch": 0.6353716381989392,
+      "grad_norm": 0.37620869278907776,
+      "learning_rate": 5.874133289082329e-06,
+      "loss": 0.2748,
+      "step": 16995
+    },
+    {
+      "epoch": 0.6355585671892889,
+      "grad_norm": 0.30327436327934265,
+      "learning_rate": 5.868784564628578e-06,
+      "loss": 0.2824,
+      "step": 17000
+    },
+    {
+      "epoch": 0.6357454961796387,
+      "grad_norm": 0.3879458010196686,
+      "learning_rate": 5.863437264912653e-06,
+      "loss": 0.3167,
+      "step": 17005
+    },
+    {
+      "epoch": 0.6359324251699886,
+      "grad_norm": 0.3175660967826843,
+      "learning_rate": 5.858091391778691e-06,
+      "loss": 0.3064,
+      "step": 17010
+    },
+    {
+      "epoch": 0.6361193541603384,
+      "grad_norm": 0.7663373351097107,
+      "learning_rate": 5.852746947070326e-06,
+      "loss": 0.3244,
+      "step": 17015
+    },
+    {
+      "epoch": 0.6363062831506882,
+      "grad_norm": 0.3658159375190735,
+      "learning_rate": 5.847403932630702e-06,
+      "loss": 0.2626,
+      "step": 17020
+    },
+    {
+      "epoch": 0.6364932121410379,
+      "grad_norm": 0.3354896008968353,
+      "learning_rate": 5.84206235030248e-06,
+      "loss": 0.2767,
+      "step": 17025
+    },
+    {
+      "epoch": 0.6366801411313877,
+      "grad_norm": 0.4783341586589813,
+      "learning_rate": 5.836722201927809e-06,
+      "loss": 0.383,
+      "step": 17030
+    },
+    {
+      "epoch": 0.6368670701217375,
+      "grad_norm": 0.5083820819854736,
+      "learning_rate": 5.831383489348361e-06,
+      "loss": 0.3022,
+      "step": 17035
+    },
+    {
+      "epoch": 0.6370539991120873,
+      "grad_norm": 0.34456634521484375,
+      "learning_rate": 5.826046214405298e-06,
+      "loss": 0.3142,
+      "step": 17040
+    },
+    {
+      "epoch": 0.637240928102437,
+      "grad_norm": 0.5159575343132019,
+      "learning_rate": 5.820710378939301e-06,
+      "loss": 0.2298,
+      "step": 17045
+    },
+    {
+      "epoch": 0.6374278570927869,
+      "grad_norm": 0.5502740740776062,
+      "learning_rate": 5.815375984790543e-06,
+      "loss": 0.4202,
+      "step": 17050
+    },
+    {
+      "epoch": 0.6376147860831367,
+      "grad_norm": 0.6210566759109497,
+      "learning_rate": 5.810043033798702e-06,
+      "loss": 0.3434,
+      "step": 17055
+    },
+    {
+      "epoch": 0.6378017150734865,
+      "grad_norm": 0.47056567668914795,
+      "learning_rate": 5.804711527802957e-06,
+      "loss": 0.2732,
+      "step": 17060
+    },
+    {
+      "epoch": 0.6379886440638363,
+      "grad_norm": 0.5329228043556213,
+      "learning_rate": 5.799381468641998e-06,
+      "loss": 0.2628,
+      "step": 17065
+    },
+    {
+      "epoch": 0.638175573054186,
+      "grad_norm": 0.5529435276985168,
+      "learning_rate": 5.794052858154009e-06,
+      "loss": 0.3167,
+      "step": 17070
+    },
+    {
+      "epoch": 0.6383625020445358,
+      "grad_norm": 0.27884042263031006,
+      "learning_rate": 5.788725698176672e-06,
+      "loss": 0.246,
+      "step": 17075
+    },
+    {
+      "epoch": 0.6385494310348856,
+      "grad_norm": 0.4293839931488037,
+      "learning_rate": 5.783399990547176e-06,
+      "loss": 0.2365,
+      "step": 17080
+    },
+    {
+      "epoch": 0.6387363600252354,
+      "grad_norm": 0.47030767798423767,
+      "learning_rate": 5.778075737102198e-06,
+      "loss": 0.3154,
+      "step": 17085
+    },
+    {
+      "epoch": 0.6389232890155851,
+      "grad_norm": 0.2606756091117859,
+      "learning_rate": 5.772752939677929e-06,
+      "loss": 0.2479,
+      "step": 17090
+    },
+    {
+      "epoch": 0.639110218005935,
+      "grad_norm": 0.527047872543335,
+      "learning_rate": 5.767431600110042e-06,
+      "loss": 0.3171,
+      "step": 17095
+    },
+    {
+      "epoch": 0.6392971469962848,
+      "grad_norm": 0.49253425002098083,
+      "learning_rate": 5.7621117202337205e-06,
+      "loss": 0.3053,
+      "step": 17100
+    },
+    {
+      "epoch": 0.6394840759866346,
+      "grad_norm": 0.5128486752510071,
+      "learning_rate": 5.7567933018836365e-06,
+      "loss": 0.3014,
+      "step": 17105
+    },
+    {
+      "epoch": 0.6396710049769844,
+      "grad_norm": 0.46093523502349854,
+      "learning_rate": 5.751476346893956e-06,
+      "loss": 0.3348,
+      "step": 17110
+    },
+    {
+      "epoch": 0.6398579339673341,
+      "grad_norm": 0.3886370360851288,
+      "learning_rate": 5.746160857098351e-06,
+      "loss": 0.2569,
+      "step": 17115
+    },
+    {
+      "epoch": 0.6400448629576839,
+      "grad_norm": 0.6568734049797058,
+      "learning_rate": 5.740846834329974e-06,
+      "loss": 0.2808,
+      "step": 17120
+    },
+    {
+      "epoch": 0.6402317919480337,
+      "grad_norm": 0.39880597591400146,
+      "learning_rate": 5.735534280421489e-06,
+      "loss": 0.3035,
+      "step": 17125
+    },
+    {
+      "epoch": 0.6404187209383835,
+      "grad_norm": 0.47509804368019104,
+      "learning_rate": 5.730223197205034e-06,
+      "loss": 0.2522,
+      "step": 17130
+    },
+    {
+      "epoch": 0.6406056499287334,
+      "grad_norm": 0.22290876507759094,
+      "learning_rate": 5.7249135865122575e-06,
+      "loss": 0.2568,
+      "step": 17135
+    },
+    {
+      "epoch": 0.6407925789190831,
+      "grad_norm": 0.36923328042030334,
+      "learning_rate": 5.719605450174283e-06,
+      "loss": 0.2765,
+      "step": 17140
+    },
+    {
+      "epoch": 0.6409795079094329,
+      "grad_norm": 0.43416711688041687,
+      "learning_rate": 5.7142987900217464e-06,
+      "loss": 0.3103,
+      "step": 17145
+    },
+    {
+      "epoch": 0.6411664368997827,
+      "grad_norm": 0.32463186979293823,
+      "learning_rate": 5.708993607884754e-06,
+      "loss": 0.2868,
+      "step": 17150
+    },
+    {
+      "epoch": 0.6413533658901325,
+      "grad_norm": 0.21319809556007385,
+      "learning_rate": 5.703689905592911e-06,
+      "loss": 0.244,
+      "step": 17155
+    },
+    {
+      "epoch": 0.6415402948804823,
+      "grad_norm": 0.39962127804756165,
+      "learning_rate": 5.698387684975317e-06,
+      "loss": 0.3021,
+      "step": 17160
+    },
+    {
+      "epoch": 0.641727223870832,
+      "grad_norm": 0.48100176453590393,
+      "learning_rate": 5.693086947860551e-06,
+      "loss": 0.3278,
+      "step": 17165
+    },
+    {
+      "epoch": 0.6419141528611818,
+      "grad_norm": 0.5302535891532898,
+      "learning_rate": 5.687787696076692e-06,
+      "loss": 0.2589,
+      "step": 17170
+    },
+    {
+      "epoch": 0.6421010818515317,
+      "grad_norm": 0.5527657866477966,
+      "learning_rate": 5.682489931451292e-06,
+      "loss": 0.3159,
+      "step": 17175
+    },
+    {
+      "epoch": 0.6422880108418815,
+      "grad_norm": 0.3536984920501709,
+      "learning_rate": 5.677193655811406e-06,
+      "loss": 0.2313,
+      "step": 17180
+    },
+    {
+      "epoch": 0.6424749398322313,
+      "grad_norm": 0.49602222442626953,
+      "learning_rate": 5.67189887098356e-06,
+      "loss": 0.3013,
+      "step": 17185
+    },
+    {
+      "epoch": 0.642661868822581,
+      "grad_norm": 0.5658937096595764,
+      "learning_rate": 5.666605578793782e-06,
+      "loss": 0.2457,
+      "step": 17190
+    },
+    {
+      "epoch": 0.6428487978129308,
+      "grad_norm": 0.10081786662340164,
+      "learning_rate": 5.661313781067572e-06,
+      "loss": 0.2528,
+      "step": 17195
+    },
+    {
+      "epoch": 0.6430357268032806,
+      "grad_norm": 0.2929933965206146,
+      "learning_rate": 5.656023479629915e-06,
+      "loss": 0.282,
+      "step": 17200
+    },
+    {
+      "epoch": 0.6432226557936304,
+      "grad_norm": 0.27145472168922424,
+      "learning_rate": 5.650734676305295e-06,
+      "loss": 0.2697,
+      "step": 17205
+    },
+    {
+      "epoch": 0.6434095847839801,
+      "grad_norm": 0.45547887682914734,
+      "learning_rate": 5.645447372917658e-06,
+      "loss": 0.2184,
+      "step": 17210
+    },
+    {
+      "epoch": 0.64359651377433,
+      "grad_norm": 0.3759501576423645,
+      "learning_rate": 5.640161571290452e-06,
+      "loss": 0.259,
+      "step": 17215
+    },
+    {
+      "epoch": 0.6437834427646798,
+      "grad_norm": 0.5527650117874146,
+      "learning_rate": 5.6348772732465925e-06,
+      "loss": 0.271,
+      "step": 17220
+    },
+    {
+      "epoch": 0.6439703717550296,
+      "grad_norm": 0.6469981670379639,
+      "learning_rate": 5.629594480608487e-06,
+      "loss": 0.2916,
+      "step": 17225
+    },
+    {
+      "epoch": 0.6441573007453794,
+      "grad_norm": 0.3704184591770172,
+      "learning_rate": 5.6243131951980145e-06,
+      "loss": 0.2614,
+      "step": 17230
+    },
+    {
+      "epoch": 0.6443442297357291,
+      "grad_norm": 0.3571639358997345,
+      "learning_rate": 5.619033418836545e-06,
+      "loss": 0.3213,
+      "step": 17235
+    },
+    {
+      "epoch": 0.6445311587260789,
+      "grad_norm": 0.40060603618621826,
+      "learning_rate": 5.613755153344918e-06,
+      "loss": 0.2876,
+      "step": 17240
+    },
+    {
+      "epoch": 0.6447180877164287,
+      "grad_norm": 0.38819968700408936,
+      "learning_rate": 5.608478400543455e-06,
+      "loss": 0.2737,
+      "step": 17245
+    },
+    {
+      "epoch": 0.6449050167067785,
+      "grad_norm": 0.38324692845344543,
+      "learning_rate": 5.60320316225196e-06,
+      "loss": 0.276,
+      "step": 17250
+    },
+    {
+      "epoch": 0.6450919456971284,
+      "grad_norm": 0.2735843360424042,
+      "learning_rate": 5.597929440289709e-06,
+      "loss": 0.3314,
+      "step": 17255
+    },
+    {
+      "epoch": 0.6452788746874781,
+      "grad_norm": 0.5796816349029541,
+      "learning_rate": 5.592657236475461e-06,
+      "loss": 0.3573,
+      "step": 17260
+    },
+    {
+      "epoch": 0.6454658036778279,
+      "grad_norm": 0.3068085312843323,
+      "learning_rate": 5.587386552627442e-06,
+      "loss": 0.3163,
+      "step": 17265
+    },
+    {
+      "epoch": 0.6456527326681777,
+      "grad_norm": 0.5472904443740845,
+      "learning_rate": 5.582117390563368e-06,
+      "loss": 0.3123,
+      "step": 17270
+    },
+    {
+      "epoch": 0.6458396616585275,
+      "grad_norm": 0.3581070005893707,
+      "learning_rate": 5.576849752100413e-06,
+      "loss": 0.2803,
+      "step": 17275
+    },
+    {
+      "epoch": 0.6460265906488772,
+      "grad_norm": 0.40146610140800476,
+      "learning_rate": 5.571583639055243e-06,
+      "loss": 0.3476,
+      "step": 17280
+    },
+    {
+      "epoch": 0.646213519639227,
+      "grad_norm": 0.43631863594055176,
+      "learning_rate": 5.566319053243983e-06,
+      "loss": 0.2792,
+      "step": 17285
+    },
+    {
+      "epoch": 0.6464004486295768,
+      "grad_norm": 0.3435373604297638,
+      "learning_rate": 5.561055996482243e-06,
+      "loss": 0.291,
+      "step": 17290
+    },
+    {
+      "epoch": 0.6465873776199267,
+      "grad_norm": 0.5556064248085022,
+      "learning_rate": 5.555794470585099e-06,
+      "loss": 0.2866,
+      "step": 17295
+    },
+    {
+      "epoch": 0.6467743066102765,
+      "grad_norm": 0.6091718077659607,
+      "learning_rate": 5.550534477367096e-06,
+      "loss": 0.3378,
+      "step": 17300
+    },
+    {
+      "epoch": 0.6469612356006262,
+      "grad_norm": 0.3173150420188904,
+      "learning_rate": 5.54527601864226e-06,
+      "loss": 0.2673,
+      "step": 17305
+    },
+    {
+      "epoch": 0.647148164590976,
+      "grad_norm": 0.629828929901123,
+      "learning_rate": 5.540019096224079e-06,
+      "loss": 0.2741,
+      "step": 17310
+    },
+    {
+      "epoch": 0.6473350935813258,
+      "grad_norm": 0.6558170914649963,
+      "learning_rate": 5.534763711925522e-06,
+      "loss": 0.3303,
+      "step": 17315
+    },
+    {
+      "epoch": 0.6475220225716756,
+      "grad_norm": 0.3189551532268524,
+      "learning_rate": 5.529509867559011e-06,
+      "loss": 0.2273,
+      "step": 17320
+    },
+    {
+      "epoch": 0.6477089515620253,
+      "grad_norm": 0.4152987003326416,
+      "learning_rate": 5.524257564936454e-06,
+      "loss": 0.2371,
+      "step": 17325
+    },
+    {
+      "epoch": 0.6478958805523751,
+      "grad_norm": 0.46988746523857117,
+      "learning_rate": 5.519006805869213e-06,
+      "loss": 0.2364,
+      "step": 17330
+    },
+    {
+      "epoch": 0.6480828095427249,
+      "grad_norm": 0.25283321738243103,
+      "learning_rate": 5.513757592168132e-06,
+      "loss": 0.2328,
+      "step": 17335
+    },
+    {
+      "epoch": 0.6482697385330748,
+      "grad_norm": 0.792521059513092,
+      "learning_rate": 5.508509925643511e-06,
+      "loss": 0.2674,
+      "step": 17340
+    },
+    {
+      "epoch": 0.6484566675234246,
+      "grad_norm": 0.7444917559623718,
+      "learning_rate": 5.503263808105114e-06,
+      "loss": 0.2861,
+      "step": 17345
+    },
+    {
+      "epoch": 0.6486435965137743,
+      "grad_norm": 0.4276942312717438,
+      "learning_rate": 5.4980192413621866e-06,
+      "loss": 0.2641,
+      "step": 17350
+    },
+    {
+      "epoch": 0.6488305255041241,
+      "grad_norm": 0.4908183217048645,
+      "learning_rate": 5.49277622722342e-06,
+      "loss": 0.2743,
+      "step": 17355
+    },
+    {
+      "epoch": 0.6490174544944739,
+      "grad_norm": 0.39298996329307556,
+      "learning_rate": 5.487534767496989e-06,
+      "loss": 0.2977,
+      "step": 17360
+    },
+    {
+      "epoch": 0.6492043834848237,
+      "grad_norm": 0.1759922206401825,
+      "learning_rate": 5.482294863990514e-06,
+      "loss": 0.3114,
+      "step": 17365
+    },
+    {
+      "epoch": 0.6493913124751735,
+      "grad_norm": 0.43271568417549133,
+      "learning_rate": 5.477056518511096e-06,
+      "loss": 0.2849,
+      "step": 17370
+    },
+    {
+      "epoch": 0.6495782414655232,
+      "grad_norm": 0.4080221951007843,
+      "learning_rate": 5.471819732865282e-06,
+      "loss": 0.2555,
+      "step": 17375
+    },
+    {
+      "epoch": 0.6497651704558731,
+      "grad_norm": 0.3136856257915497,
+      "learning_rate": 5.466584508859096e-06,
+      "loss": 0.1905,
+      "step": 17380
+    },
+    {
+      "epoch": 0.6499520994462229,
+      "grad_norm": 0.3478422164916992,
+      "learning_rate": 5.461350848298016e-06,
+      "loss": 0.356,
+      "step": 17385
+    },
+    {
+      "epoch": 0.6501390284365727,
+      "grad_norm": 0.3305181562900543,
+      "learning_rate": 5.456118752986975e-06,
+      "loss": 0.3352,
+      "step": 17390
+    },
+    {
+      "epoch": 0.6503259574269225,
+      "grad_norm": 0.35783740878105164,
+      "learning_rate": 5.4508882247303815e-06,
+      "loss": 0.2462,
+      "step": 17395
+    },
+    {
+      "epoch": 0.6505128864172722,
+      "grad_norm": 0.14547370374202728,
+      "learning_rate": 5.445659265332087e-06,
+      "loss": 0.2325,
+      "step": 17400
+    },
+    {
+      "epoch": 0.650699815407622,
+      "grad_norm": 0.26868870854377747,
+      "learning_rate": 5.440431876595418e-06,
+      "loss": 0.2746,
+      "step": 17405
+    },
+    {
+      "epoch": 0.6508867443979718,
+      "grad_norm": 0.4959259629249573,
+      "learning_rate": 5.435206060323142e-06,
+      "loss": 0.2315,
+      "step": 17410
+    },
+    {
+      "epoch": 0.6510736733883216,
+      "grad_norm": 0.5970064997673035,
+      "learning_rate": 5.4299818183175006e-06,
+      "loss": 0.3014,
+      "step": 17415
+    },
+    {
+      "epoch": 0.6512606023786714,
+      "grad_norm": 0.4646480679512024,
+      "learning_rate": 5.424759152380179e-06,
+      "loss": 0.3339,
+      "step": 17420
+    },
+    {
+      "epoch": 0.6514475313690212,
+      "grad_norm": 0.31364089250564575,
+      "learning_rate": 5.419538064312333e-06,
+      "loss": 0.3644,
+      "step": 17425
+    },
+    {
+      "epoch": 0.651634460359371,
+      "grad_norm": 0.4947362244129181,
+      "learning_rate": 5.414318555914563e-06,
+      "loss": 0.2448,
+      "step": 17430
+    },
+    {
+      "epoch": 0.6518213893497208,
+      "grad_norm": 0.3628036677837372,
+      "learning_rate": 5.409100628986921e-06,
+      "loss": 0.2198,
+      "step": 17435
+    },
+    {
+      "epoch": 0.6520083183400706,
+      "grad_norm": 0.38185828924179077,
+      "learning_rate": 5.4038842853289305e-06,
+      "loss": 0.3506,
+      "step": 17440
+    },
+    {
+      "epoch": 0.6521952473304203,
+      "grad_norm": 0.37392547726631165,
+      "learning_rate": 5.398669526739551e-06,
+      "loss": 0.2647,
+      "step": 17445
+    },
+    {
+      "epoch": 0.6523821763207701,
+      "grad_norm": 0.3992078900337219,
+      "learning_rate": 5.39345635501721e-06,
+      "loss": 0.2584,
+      "step": 17450
+    },
+    {
+      "epoch": 0.6525691053111199,
+      "grad_norm": 0.27611681818962097,
+      "learning_rate": 5.388244771959777e-06,
+      "loss": 0.2419,
+      "step": 17455
+    },
+    {
+      "epoch": 0.6527560343014698,
+      "grad_norm": 0.42422646284103394,
+      "learning_rate": 5.38303477936458e-06,
+      "loss": 0.2328,
+      "step": 17460
+    },
+    {
+      "epoch": 0.6529429632918196,
+      "grad_norm": 0.25942620635032654,
+      "learning_rate": 5.3778263790283905e-06,
+      "loss": 0.2184,
+      "step": 17465
+    },
+    {
+      "epoch": 0.6531298922821693,
+      "grad_norm": 0.4233878552913666,
+      "learning_rate": 5.372619572747442e-06,
+      "loss": 0.3018,
+      "step": 17470
+    },
+    {
+      "epoch": 0.6533168212725191,
+      "grad_norm": 0.6455023884773254,
+      "learning_rate": 5.3674143623174144e-06,
+      "loss": 0.269,
+      "step": 17475
+    },
+    {
+      "epoch": 0.6535037502628689,
+      "grad_norm": 0.635086178779602,
+      "learning_rate": 5.362210749533434e-06,
+      "loss": 0.2835,
+      "step": 17480
+    },
+    {
+      "epoch": 0.6536906792532187,
+      "grad_norm": 0.21044528484344482,
+      "learning_rate": 5.357008736190077e-06,
+      "loss": 0.3771,
+      "step": 17485
+    },
+    {
+      "epoch": 0.6538776082435684,
+      "grad_norm": 0.5174884796142578,
+      "learning_rate": 5.351808324081362e-06,
+      "loss": 0.2438,
+      "step": 17490
+    },
+    {
+      "epoch": 0.6540645372339182,
+      "grad_norm": 0.4793969988822937,
+      "learning_rate": 5.346609515000775e-06,
+      "loss": 0.2707,
+      "step": 17495
+    },
+    {
+      "epoch": 0.6542514662242681,
+      "grad_norm": 0.39202970266342163,
+      "learning_rate": 5.341412310741225e-06,
+      "loss": 0.2867,
+      "step": 17500
+    },
+    {
+      "epoch": 0.6544383952146179,
+      "grad_norm": 0.3729572296142578,
+      "learning_rate": 5.336216713095087e-06,
+      "loss": 0.3556,
+      "step": 17505
+    },
+    {
+      "epoch": 0.6546253242049677,
+      "grad_norm": 0.2661343514919281,
+      "learning_rate": 5.3310227238541665e-06,
+      "loss": 0.2963,
+      "step": 17510
+    },
+    {
+      "epoch": 0.6548122531953174,
+      "grad_norm": 0.6629586815834045,
+      "learning_rate": 5.325830344809726e-06,
+      "loss": 0.3187,
+      "step": 17515
+    },
+    {
+      "epoch": 0.6549991821856672,
+      "grad_norm": 0.5217400789260864,
+      "learning_rate": 5.320639577752471e-06,
+      "loss": 0.2794,
+      "step": 17520
+    },
+    {
+      "epoch": 0.655186111176017,
+      "grad_norm": 0.40809670090675354,
+      "learning_rate": 5.315450424472546e-06,
+      "loss": 0.3342,
+      "step": 17525
+    },
+    {
+      "epoch": 0.6553730401663668,
+      "grad_norm": 0.3134659230709076,
+      "learning_rate": 5.31026288675954e-06,
+      "loss": 0.2624,
+      "step": 17530
+    },
+    {
+      "epoch": 0.6555599691567165,
+      "grad_norm": 0.5745450258255005,
+      "learning_rate": 5.305076966402483e-06,
+      "loss": 0.3727,
+      "step": 17535
+    },
+    {
+      "epoch": 0.6557468981470664,
+      "grad_norm": 0.6128538846969604,
+      "learning_rate": 5.299892665189856e-06,
+      "loss": 0.2772,
+      "step": 17540
+    },
+    {
+      "epoch": 0.6559338271374162,
+      "grad_norm": 0.2917860746383667,
+      "learning_rate": 5.2947099849095695e-06,
+      "loss": 0.3028,
+      "step": 17545
+    },
+    {
+      "epoch": 0.656120756127766,
+      "grad_norm": 0.6575778722763062,
+      "learning_rate": 5.2895289273489915e-06,
+      "loss": 0.2669,
+      "step": 17550
+    },
+    {
+      "epoch": 0.6563076851181158,
+      "grad_norm": 0.4796570837497711,
+      "learning_rate": 5.2843494942949095e-06,
+      "loss": 0.3287,
+      "step": 17555
+    },
+    {
+      "epoch": 0.6564946141084655,
+      "grad_norm": 0.5242589712142944,
+      "learning_rate": 5.279171687533566e-06,
+      "loss": 0.2346,
+      "step": 17560
+    },
+    {
+      "epoch": 0.6566815430988153,
+      "grad_norm": 0.4839895963668823,
+      "learning_rate": 5.273995508850643e-06,
+      "loss": 0.2419,
+      "step": 17565
+    },
+    {
+      "epoch": 0.6568684720891651,
+      "grad_norm": 0.4953691065311432,
+      "learning_rate": 5.268820960031252e-06,
+      "loss": 0.3018,
+      "step": 17570
+    },
+    {
+      "epoch": 0.6570554010795149,
+      "grad_norm": 0.34803852438926697,
+      "learning_rate": 5.263648042859945e-06,
+      "loss": 0.2719,
+      "step": 17575
+    },
+    {
+      "epoch": 0.6572423300698647,
+      "grad_norm": 0.4037022888660431,
+      "learning_rate": 5.258476759120713e-06,
+      "loss": 0.251,
+      "step": 17580
+    },
+    {
+      "epoch": 0.6574292590602145,
+      "grad_norm": 0.38551491498947144,
+      "learning_rate": 5.253307110596988e-06,
+      "loss": 0.1929,
+      "step": 17585
+    },
+    {
+      "epoch": 0.6576161880505643,
+      "grad_norm": 0.33148205280303955,
+      "learning_rate": 5.248139099071625e-06,
+      "loss": 0.2561,
+      "step": 17590
+    },
+    {
+      "epoch": 0.6578031170409141,
+      "grad_norm": 0.42896026372909546,
+      "learning_rate": 5.242972726326934e-06,
+      "loss": 0.2475,
+      "step": 17595
+    },
+    {
+      "epoch": 0.6579900460312639,
+      "grad_norm": 0.5869537591934204,
+      "learning_rate": 5.23780799414464e-06,
+      "loss": 0.3102,
+      "step": 17600
+    },
+    {
+      "epoch": 0.6581769750216137,
+      "grad_norm": 0.42332693934440613,
+      "learning_rate": 5.232644904305914e-06,
+      "loss": 0.2316,
+      "step": 17605
+    },
+    {
+      "epoch": 0.6583639040119634,
+      "grad_norm": 0.20432531833648682,
+      "learning_rate": 5.227483458591364e-06,
+      "loss": 0.3037,
+      "step": 17610
+    },
+    {
+      "epoch": 0.6585508330023132,
+      "grad_norm": 0.5013585686683655,
+      "learning_rate": 5.222323658781018e-06,
+      "loss": 0.2422,
+      "step": 17615
+    },
+    {
+      "epoch": 0.658737761992663,
+      "grad_norm": 0.5252900719642639,
+      "learning_rate": 5.21716550665434e-06,
+      "loss": 0.2841,
+      "step": 17620
+    },
+    {
+      "epoch": 0.6589246909830129,
+      "grad_norm": 0.3324468433856964,
+      "learning_rate": 5.212009003990237e-06,
+      "loss": 0.2643,
+      "step": 17625
+    },
+    {
+      "epoch": 0.6591116199733627,
+      "grad_norm": 0.5473644137382507,
+      "learning_rate": 5.206854152567036e-06,
+      "loss": 0.2717,
+      "step": 17630
+    },
+    {
+      "epoch": 0.6592985489637124,
+      "grad_norm": 1.3611345291137695,
+      "learning_rate": 5.201700954162493e-06,
+      "loss": 0.2884,
+      "step": 17635
+    },
+    {
+      "epoch": 0.6594854779540622,
+      "grad_norm": 0.6788679957389832,
+      "learning_rate": 5.196549410553806e-06,
+      "loss": 0.2607,
+      "step": 17640
+    },
+    {
+      "epoch": 0.659672406944412,
+      "grad_norm": 0.5106036067008972,
+      "learning_rate": 5.191399523517586e-06,
+      "loss": 0.3072,
+      "step": 17645
+    },
+    {
+      "epoch": 0.6598593359347618,
+      "grad_norm": 0.8317034840583801,
+      "learning_rate": 5.1862512948298885e-06,
+      "loss": 0.3571,
+      "step": 17650
+    },
+    {
+      "epoch": 0.6600462649251115,
+      "grad_norm": 0.3770235776901245,
+      "learning_rate": 5.181104726266191e-06,
+      "loss": 0.3051,
+      "step": 17655
+    },
+    {
+      "epoch": 0.6602331939154613,
+      "grad_norm": 0.33272841572761536,
+      "learning_rate": 5.17595981960139e-06,
+      "loss": 0.3451,
+      "step": 17660
+    },
+    {
+      "epoch": 0.6604201229058112,
+      "grad_norm": 0.36768677830696106,
+      "learning_rate": 5.170816576609825e-06,
+      "loss": 0.259,
+      "step": 17665
+    },
+    {
+      "epoch": 0.660607051896161,
+      "grad_norm": 0.4106055498123169,
+      "learning_rate": 5.16567499906525e-06,
+      "loss": 0.342,
+      "step": 17670
+    },
+    {
+      "epoch": 0.6607939808865108,
+      "grad_norm": 0.44016748666763306,
+      "learning_rate": 5.1605350887408466e-06,
+      "loss": 0.3125,
+      "step": 17675
+    },
+    {
+      "epoch": 0.6609809098768605,
+      "grad_norm": 0.40315133333206177,
+      "learning_rate": 5.1553968474092185e-06,
+      "loss": 0.2638,
+      "step": 17680
+    },
+    {
+      "epoch": 0.6611678388672103,
+      "grad_norm": 0.22286388278007507,
+      "learning_rate": 5.150260276842407e-06,
+      "loss": 0.2822,
+      "step": 17685
+    },
+    {
+      "epoch": 0.6613547678575601,
+      "grad_norm": 0.3545617163181305,
+      "learning_rate": 5.1451253788118595e-06,
+      "loss": 0.3389,
+      "step": 17690
+    },
+    {
+      "epoch": 0.6615416968479099,
+      "grad_norm": 0.4508470594882965,
+      "learning_rate": 5.139992155088458e-06,
+      "loss": 0.28,
+      "step": 17695
+    },
+    {
+      "epoch": 0.6617286258382596,
+      "grad_norm": 0.3894427418708801,
+      "learning_rate": 5.13486060744251e-06,
+      "loss": 0.2735,
+      "step": 17700
+    },
+    {
+      "epoch": 0.6619155548286095,
+      "grad_norm": 0.5788761377334595,
+      "learning_rate": 5.12973073764373e-06,
+      "loss": 0.3752,
+      "step": 17705
+    },
+    {
+      "epoch": 0.6621024838189593,
+      "grad_norm": 0.2697499394416809,
+      "learning_rate": 5.124602547461273e-06,
+      "loss": 0.2663,
+      "step": 17710
+    },
+    {
+      "epoch": 0.6622894128093091,
+      "grad_norm": 0.41945934295654297,
+      "learning_rate": 5.119476038663699e-06,
+      "loss": 0.3097,
+      "step": 17715
+    },
+    {
+      "epoch": 0.6624763417996589,
+      "grad_norm": 0.20760861039161682,
+      "learning_rate": 5.1143512130189935e-06,
+      "loss": 0.2761,
+      "step": 17720
+    },
+    {
+      "epoch": 0.6626632707900086,
+      "grad_norm": 0.47304192185401917,
+      "learning_rate": 5.109228072294561e-06,
+      "loss": 0.277,
+      "step": 17725
+    },
+    {
+      "epoch": 0.6628501997803584,
+      "grad_norm": 0.3576745390892029,
+      "learning_rate": 5.1041066182572296e-06,
+      "loss": 0.2783,
+      "step": 17730
+    },
+    {
+      "epoch": 0.6630371287707082,
+      "grad_norm": 0.41994425654411316,
+      "learning_rate": 5.098986852673239e-06,
+      "loss": 0.2465,
+      "step": 17735
+    },
+    {
+      "epoch": 0.663224057761058,
+      "grad_norm": 0.3337131440639496,
+      "learning_rate": 5.093868777308251e-06,
+      "loss": 0.3563,
+      "step": 17740
+    },
+    {
+      "epoch": 0.6634109867514079,
+      "grad_norm": 0.6296584010124207,
+      "learning_rate": 5.088752393927345e-06,
+      "loss": 0.2916,
+      "step": 17745
+    },
+    {
+      "epoch": 0.6635979157417576,
+      "grad_norm": 0.1845739334821701,
+      "learning_rate": 5.08363770429501e-06,
+      "loss": 0.2693,
+      "step": 17750
+    },
+    {
+      "epoch": 0.6637848447321074,
+      "grad_norm": 0.3726886510848999,
+      "learning_rate": 5.0785247101751645e-06,
+      "loss": 0.2992,
+      "step": 17755
+    },
+    {
+      "epoch": 0.6639717737224572,
+      "grad_norm": 0.26406246423721313,
+      "learning_rate": 5.073413413331128e-06,
+      "loss": 0.2997,
+      "step": 17760
+    },
+    {
+      "epoch": 0.664158702712807,
+      "grad_norm": 0.421535849571228,
+      "learning_rate": 5.068303815525639e-06,
+      "loss": 0.2156,
+      "step": 17765
+    },
+    {
+      "epoch": 0.6643456317031567,
+      "grad_norm": 0.40606680512428284,
+      "learning_rate": 5.0631959185208514e-06,
+      "loss": 0.2638,
+      "step": 17770
+    },
+    {
+      "epoch": 0.6645325606935065,
+      "grad_norm": 0.31488245725631714,
+      "learning_rate": 5.0580897240783365e-06,
+      "loss": 0.2754,
+      "step": 17775
+    },
+    {
+      "epoch": 0.6647194896838563,
+      "grad_norm": 0.4173290729522705,
+      "learning_rate": 5.052985233959069e-06,
+      "loss": 0.2472,
+      "step": 17780
+    },
+    {
+      "epoch": 0.6649064186742062,
+      "grad_norm": 0.4380118250846863,
+      "learning_rate": 5.047882449923444e-06,
+      "loss": 0.2227,
+      "step": 17785
+    },
+    {
+      "epoch": 0.665093347664556,
+      "grad_norm": 0.22158585488796234,
+      "learning_rate": 5.04278137373127e-06,
+      "loss": 0.2522,
+      "step": 17790
+    },
+    {
+      "epoch": 0.6652802766549057,
+      "grad_norm": 0.5004199147224426,
+      "learning_rate": 5.037682007141754e-06,
+      "loss": 0.3162,
+      "step": 17795
+    },
+    {
+      "epoch": 0.6654672056452555,
+      "grad_norm": 0.5138402581214905,
+      "learning_rate": 5.03258435191353e-06,
+      "loss": 0.2277,
+      "step": 17800
+    },
+    {
+      "epoch": 0.6656541346356053,
+      "grad_norm": 0.569689929485321,
+      "learning_rate": 5.027488409804624e-06,
+      "loss": 0.2435,
+      "step": 17805
+    },
+    {
+      "epoch": 0.6658410636259551,
+      "grad_norm": 0.50981605052948,
+      "learning_rate": 5.022394182572487e-06,
+      "loss": 0.2654,
+      "step": 17810
+    },
+    {
+      "epoch": 0.6660279926163049,
+      "grad_norm": 0.39164555072784424,
+      "learning_rate": 5.017301671973973e-06,
+      "loss": 0.2715,
+      "step": 17815
+    },
+    {
+      "epoch": 0.6662149216066546,
+      "grad_norm": 0.31828561425209045,
+      "learning_rate": 5.012210879765339e-06,
+      "loss": 0.2559,
+      "step": 17820
+    },
+    {
+      "epoch": 0.6664018505970044,
+      "grad_norm": 0.6788938641548157,
+      "learning_rate": 5.0071218077022495e-06,
+      "loss": 0.3659,
+      "step": 17825
+    },
+    {
+      "epoch": 0.6665887795873543,
+      "grad_norm": 0.2640472948551178,
+      "learning_rate": 5.002034457539786e-06,
+      "loss": 0.2719,
+      "step": 17830
+    },
+    {
+      "epoch": 0.6667757085777041,
+      "grad_norm": 0.3954325318336487,
+      "learning_rate": 4.996948831032431e-06,
+      "loss": 0.2705,
+      "step": 17835
+    },
+    {
+      "epoch": 0.6669626375680539,
+      "grad_norm": 0.25173521041870117,
+      "learning_rate": 4.991864929934065e-06,
+      "loss": 0.2495,
+      "step": 17840
+    },
+    {
+      "epoch": 0.6671495665584036,
+      "grad_norm": 0.3650529384613037,
+      "learning_rate": 4.986782755997987e-06,
+      "loss": 0.3812,
+      "step": 17845
+    },
+    {
+      "epoch": 0.6673364955487534,
+      "grad_norm": 0.5799317359924316,
+      "learning_rate": 4.981702310976887e-06,
+      "loss": 0.2965,
+      "step": 17850
+    },
+    {
+      "epoch": 0.6675234245391032,
+      "grad_norm": 0.3628406822681427,
+      "learning_rate": 4.97662359662287e-06,
+      "loss": 0.3047,
+      "step": 17855
+    },
+    {
+      "epoch": 0.667710353529453,
+      "grad_norm": 0.8548519015312195,
+      "learning_rate": 4.971546614687437e-06,
+      "loss": 0.3368,
+      "step": 17860
+    },
+    {
+      "epoch": 0.6678972825198027,
+      "grad_norm": 0.4021095037460327,
+      "learning_rate": 4.966471366921493e-06,
+      "loss": 0.3546,
+      "step": 17865
+    },
+    {
+      "epoch": 0.6680842115101526,
+      "grad_norm": 0.23902134597301483,
+      "learning_rate": 4.961397855075343e-06,
+      "loss": 0.3073,
+      "step": 17870
+    },
+    {
+      "epoch": 0.6682711405005024,
+      "grad_norm": 0.24928809702396393,
+      "learning_rate": 4.956326080898697e-06,
+      "loss": 0.2987,
+      "step": 17875
+    },
+    {
+      "epoch": 0.6684580694908522,
+      "grad_norm": 0.4533480703830719,
+      "learning_rate": 4.951256046140671e-06,
+      "loss": 0.3074,
+      "step": 17880
+    },
+    {
+      "epoch": 0.668644998481202,
+      "grad_norm": 0.3241102695465088,
+      "learning_rate": 4.946187752549766e-06,
+      "loss": 0.2708,
+      "step": 17885
+    },
+    {
+      "epoch": 0.6688319274715517,
+      "grad_norm": 0.5674617290496826,
+      "learning_rate": 4.9411212018738984e-06,
+      "loss": 0.3197,
+      "step": 17890
+    },
+    {
+      "epoch": 0.6690188564619015,
+      "grad_norm": 0.41915619373321533,
+      "learning_rate": 4.936056395860369e-06,
+      "loss": 0.3046,
+      "step": 17895
+    },
+    {
+      "epoch": 0.6692057854522513,
+      "grad_norm": 0.2280120998620987,
+      "learning_rate": 4.930993336255892e-06,
+      "loss": 0.2369,
+      "step": 17900
+    },
+    {
+      "epoch": 0.6693927144426011,
+      "grad_norm": 0.4832659661769867,
+      "learning_rate": 4.925932024806569e-06,
+      "loss": 0.3249,
+      "step": 17905
+    },
+    {
+      "epoch": 0.669579643432951,
+      "grad_norm": 0.5970139503479004,
+      "learning_rate": 4.9208724632579e-06,
+      "loss": 0.3852,
+      "step": 17910
+    },
+    {
+      "epoch": 0.6697665724233007,
+      "grad_norm": 0.3375006914138794,
+      "learning_rate": 4.915814653354779e-06,
+      "loss": 0.2146,
+      "step": 17915
+    },
+    {
+      "epoch": 0.6699535014136505,
+      "grad_norm": 0.3070432245731354,
+      "learning_rate": 4.910758596841504e-06,
+      "loss": 0.2901,
+      "step": 17920
+    },
+    {
+      "epoch": 0.6701404304040003,
+      "grad_norm": 0.4842531681060791,
+      "learning_rate": 4.905704295461767e-06,
+      "loss": 0.1967,
+      "step": 17925
+    },
+    {
+      "epoch": 0.6703273593943501,
+      "grad_norm": 0.24802975356578827,
+      "learning_rate": 4.900651750958645e-06,
+      "loss": 0.3403,
+      "step": 17930
+    },
+    {
+      "epoch": 0.6705142883846998,
+      "grad_norm": 0.32612282037734985,
+      "learning_rate": 4.895600965074623e-06,
+      "loss": 0.2696,
+      "step": 17935
+    },
+    {
+      "epoch": 0.6707012173750496,
+      "grad_norm": 0.3664465844631195,
+      "learning_rate": 4.890551939551565e-06,
+      "loss": 0.2614,
+      "step": 17940
+    },
+    {
+      "epoch": 0.6708881463653994,
+      "grad_norm": 0.4860963523387909,
+      "learning_rate": 4.885504676130743e-06,
+      "loss": 0.2449,
+      "step": 17945
+    },
+    {
+      "epoch": 0.6710750753557493,
+      "grad_norm": 0.38064733147621155,
+      "learning_rate": 4.880459176552808e-06,
+      "loss": 0.285,
+      "step": 17950
+    },
+    {
+      "epoch": 0.6712620043460991,
+      "grad_norm": 0.38366106152534485,
+      "learning_rate": 4.87541544255781e-06,
+      "loss": 0.3247,
+      "step": 17955
+    },
+    {
+      "epoch": 0.6714489333364488,
+      "grad_norm": 0.7270559668540955,
+      "learning_rate": 4.8703734758851854e-06,
+      "loss": 0.3462,
+      "step": 17960
+    },
+    {
+      "epoch": 0.6716358623267986,
+      "grad_norm": 0.23612673580646515,
+      "learning_rate": 4.865333278273768e-06,
+      "loss": 0.2441,
+      "step": 17965
+    },
+    {
+      "epoch": 0.6718227913171484,
+      "grad_norm": 0.24252595007419586,
+      "learning_rate": 4.860294851461774e-06,
+      "loss": 0.2234,
+      "step": 17970
+    },
+    {
+      "epoch": 0.6720097203074982,
+      "grad_norm": 0.5922762751579285,
+      "learning_rate": 4.8552581971868154e-06,
+      "loss": 0.2714,
+      "step": 17975
+    },
+    {
+      "epoch": 0.672196649297848,
+      "grad_norm": 0.477463960647583,
+      "learning_rate": 4.850223317185891e-06,
+      "loss": 0.2555,
+      "step": 17980
+    },
+    {
+      "epoch": 0.6723835782881977,
+      "grad_norm": 0.4230765700340271,
+      "learning_rate": 4.845190213195382e-06,
+      "loss": 0.2681,
+      "step": 17985
+    },
+    {
+      "epoch": 0.6725705072785476,
+      "grad_norm": 0.7639047503471375,
+      "learning_rate": 4.840158886951069e-06,
+      "loss": 0.4085,
+      "step": 17990
+    },
+    {
+      "epoch": 0.6727574362688974,
+      "grad_norm": 0.27008453011512756,
+      "learning_rate": 4.835129340188101e-06,
+      "loss": 0.2701,
+      "step": 17995
+    },
+    {
+      "epoch": 0.6729443652592472,
+      "grad_norm": 0.48787766695022583,
+      "learning_rate": 4.8301015746410385e-06,
+      "loss": 0.3341,
+      "step": 18000
+    },
+    {
+      "epoch": 0.673131294249597,
+      "grad_norm": 0.4696618914604187,
+      "learning_rate": 4.825075592043805e-06,
+      "loss": 0.245,
+      "step": 18005
+    },
+    {
+      "epoch": 0.6733182232399467,
+      "grad_norm": 0.2987324893474579,
+      "learning_rate": 4.82005139412972e-06,
+      "loss": 0.3017,
+      "step": 18010
+    },
+    {
+      "epoch": 0.6735051522302965,
+      "grad_norm": 0.35396862030029297,
+      "learning_rate": 4.81502898263148e-06,
+      "loss": 0.2776,
+      "step": 18015
+    },
+    {
+      "epoch": 0.6736920812206463,
+      "grad_norm": 0.1255078911781311,
+      "learning_rate": 4.810008359281176e-06,
+      "loss": 0.2967,
+      "step": 18020
+    },
+    {
+      "epoch": 0.673879010210996,
+      "grad_norm": 0.4665721356868744,
+      "learning_rate": 4.804989525810282e-06,
+      "loss": 0.3912,
+      "step": 18025
+    },
+    {
+      "epoch": 0.6740659392013459,
+      "grad_norm": 0.4541762173175812,
+      "learning_rate": 4.79997248394964e-06,
+      "loss": 0.3989,
+      "step": 18030
+    },
+    {
+      "epoch": 0.6742528681916957,
+      "grad_norm": 0.17799994349479675,
+      "learning_rate": 4.794957235429491e-06,
+      "loss": 0.3017,
+      "step": 18035
+    },
+    {
+      "epoch": 0.6744397971820455,
+      "grad_norm": 0.5030390024185181,
+      "learning_rate": 4.789943781979447e-06,
+      "loss": 0.2906,
+      "step": 18040
+    },
+    {
+      "epoch": 0.6746267261723953,
+      "grad_norm": 0.3190663158893585,
+      "learning_rate": 4.784932125328507e-06,
+      "loss": 0.2697,
+      "step": 18045
+    },
+    {
+      "epoch": 0.674813655162745,
+      "grad_norm": 0.5195890069007874,
+      "learning_rate": 4.779922267205048e-06,
+      "loss": 0.2506,
+      "step": 18050
+    },
+    {
+      "epoch": 0.6750005841530948,
+      "grad_norm": 0.48416566848754883,
+      "learning_rate": 4.774914209336824e-06,
+      "loss": 0.367,
+      "step": 18055
+    },
+    {
+      "epoch": 0.6751875131434446,
+      "grad_norm": 0.38188669085502625,
+      "learning_rate": 4.769907953450968e-06,
+      "loss": 0.277,
+      "step": 18060
+    },
+    {
+      "epoch": 0.6753744421337944,
+      "grad_norm": 0.4269798696041107,
+      "learning_rate": 4.764903501273999e-06,
+      "loss": 0.3099,
+      "step": 18065
+    },
+    {
+      "epoch": 0.6755613711241442,
+      "grad_norm": 0.5362254977226257,
+      "learning_rate": 4.75990085453181e-06,
+      "loss": 0.3038,
+      "step": 18070
+    },
+    {
+      "epoch": 0.675748300114494,
+      "grad_norm": 0.3357028067111969,
+      "learning_rate": 4.754900014949665e-06,
+      "loss": 0.2606,
+      "step": 18075
+    },
+    {
+      "epoch": 0.6759352291048438,
+      "grad_norm": 0.43899163603782654,
+      "learning_rate": 4.749900984252218e-06,
+      "loss": 0.308,
+      "step": 18080
+    },
+    {
+      "epoch": 0.6761221580951936,
+      "grad_norm": 0.2953328788280487,
+      "learning_rate": 4.744903764163483e-06,
+      "loss": 0.2463,
+      "step": 18085
+    },
+    {
+      "epoch": 0.6763090870855434,
+      "grad_norm": 0.2157057821750641,
+      "learning_rate": 4.739908356406866e-06,
+      "loss": 0.3184,
+      "step": 18090
+    },
+    {
+      "epoch": 0.6764960160758932,
+      "grad_norm": 0.5682590007781982,
+      "learning_rate": 4.7349147627051365e-06,
+      "loss": 0.2243,
+      "step": 18095
+    },
+    {
+      "epoch": 0.6766829450662429,
+      "grad_norm": 0.3066447675228119,
+      "learning_rate": 4.729922984780441e-06,
+      "loss": 0.2636,
+      "step": 18100
+    },
+    {
+      "epoch": 0.6768698740565927,
+      "grad_norm": 0.35900914669036865,
+      "learning_rate": 4.724933024354298e-06,
+      "loss": 0.2162,
+      "step": 18105
+    },
+    {
+      "epoch": 0.6770568030469425,
+      "grad_norm": 0.40200889110565186,
+      "learning_rate": 4.719944883147605e-06,
+      "loss": 0.292,
+      "step": 18110
+    },
+    {
+      "epoch": 0.6772437320372924,
+      "grad_norm": 0.23843464255332947,
+      "learning_rate": 4.714958562880633e-06,
+      "loss": 0.2521,
+      "step": 18115
+    },
+    {
+      "epoch": 0.6774306610276422,
+      "grad_norm": 0.6935325860977173,
+      "learning_rate": 4.709974065273013e-06,
+      "loss": 0.2634,
+      "step": 18120
+    },
+    {
+      "epoch": 0.6776175900179919,
+      "grad_norm": 0.1865166425704956,
+      "learning_rate": 4.704991392043763e-06,
+      "loss": 0.3864,
+      "step": 18125
+    },
+    {
+      "epoch": 0.6778045190083417,
+      "grad_norm": 0.6192229986190796,
+      "learning_rate": 4.700010544911257e-06,
+      "loss": 0.2523,
+      "step": 18130
+    },
+    {
+      "epoch": 0.6779914479986915,
+      "grad_norm": 0.22409962117671967,
+      "learning_rate": 4.695031525593254e-06,
+      "loss": 0.2784,
+      "step": 18135
+    },
+    {
+      "epoch": 0.6781783769890413,
+      "grad_norm": 0.118705615401268,
+      "learning_rate": 4.690054335806872e-06,
+      "loss": 0.2382,
+      "step": 18140
+    },
+    {
+      "epoch": 0.678365305979391,
+      "grad_norm": 0.39932093024253845,
+      "learning_rate": 4.6850789772685955e-06,
+      "loss": 0.223,
+      "step": 18145
+    },
+    {
+      "epoch": 0.6785522349697408,
+      "grad_norm": 0.8999297022819519,
+      "learning_rate": 4.6801054516942924e-06,
+      "loss": 0.2789,
+      "step": 18150
+    },
+    {
+      "epoch": 0.6787391639600907,
+      "grad_norm": 0.44918587803840637,
+      "learning_rate": 4.675133760799181e-06,
+      "loss": 0.3355,
+      "step": 18155
+    },
+    {
+      "epoch": 0.6789260929504405,
+      "grad_norm": 0.4980359673500061,
+      "learning_rate": 4.6701639062978624e-06,
+      "loss": 0.2608,
+      "step": 18160
+    },
+    {
+      "epoch": 0.6791130219407903,
+      "grad_norm": 0.30972304940223694,
+      "learning_rate": 4.6651958899042895e-06,
+      "loss": 0.207,
+      "step": 18165
+    },
+    {
+      "epoch": 0.67929995093114,
+      "grad_norm": 0.34727078676223755,
+      "learning_rate": 4.660229713331797e-06,
+      "loss": 0.2291,
+      "step": 18170
+    },
+    {
+      "epoch": 0.6794868799214898,
+      "grad_norm": 0.2929920256137848,
+      "learning_rate": 4.655265378293068e-06,
+      "loss": 0.2834,
+      "step": 18175
+    },
+    {
+      "epoch": 0.6796738089118396,
+      "grad_norm": 0.37373247742652893,
+      "learning_rate": 4.650302886500168e-06,
+      "loss": 0.3505,
+      "step": 18180
+    },
+    {
+      "epoch": 0.6798607379021894,
+      "grad_norm": 0.3892793655395508,
+      "learning_rate": 4.645342239664511e-06,
+      "loss": 0.287,
+      "step": 18185
+    },
+    {
+      "epoch": 0.6800476668925391,
+      "grad_norm": 0.21411560475826263,
+      "learning_rate": 4.640383439496888e-06,
+      "loss": 0.1991,
+      "step": 18190
+    },
+    {
+      "epoch": 0.680234595882889,
+      "grad_norm": 0.30452221632003784,
+      "learning_rate": 4.635426487707445e-06,
+      "loss": 0.2585,
+      "step": 18195
+    },
+    {
+      "epoch": 0.6804215248732388,
+      "grad_norm": 0.3732444941997528,
+      "learning_rate": 4.630471386005688e-06,
+      "loss": 0.2668,
+      "step": 18200
+    },
+    {
+      "epoch": 0.6806084538635886,
+      "grad_norm": 0.4895915687084198,
+      "learning_rate": 4.625518136100498e-06,
+      "loss": 0.3098,
+      "step": 18205
+    },
+    {
+      "epoch": 0.6807953828539384,
+      "grad_norm": 0.25300148129463196,
+      "learning_rate": 4.6205667397001e-06,
+      "loss": 0.2581,
+      "step": 18210
+    },
+    {
+      "epoch": 0.6809823118442881,
+      "grad_norm": 0.38394615054130554,
+      "learning_rate": 4.615617198512097e-06,
+      "loss": 0.2979,
+      "step": 18215
+    },
+    {
+      "epoch": 0.6811692408346379,
+      "grad_norm": 0.3098186254501343,
+      "learning_rate": 4.6106695142434355e-06,
+      "loss": 0.2881,
+      "step": 18220
+    },
+    {
+      "epoch": 0.6813561698249877,
+      "grad_norm": 0.5445083379745483,
+      "learning_rate": 4.60572368860044e-06,
+      "loss": 0.2482,
+      "step": 18225
+    },
+    {
+      "epoch": 0.6815430988153375,
+      "grad_norm": 0.3289521634578705,
+      "learning_rate": 4.600779723288774e-06,
+      "loss": 0.3112,
+      "step": 18230
+    },
+    {
+      "epoch": 0.6817300278056874,
+      "grad_norm": 0.6174635291099548,
+      "learning_rate": 4.595837620013478e-06,
+      "loss": 0.2644,
+      "step": 18235
+    },
+    {
+      "epoch": 0.6819169567960371,
+      "grad_norm": 0.3422519266605377,
+      "learning_rate": 4.5908973804789385e-06,
+      "loss": 0.2874,
+      "step": 18240
+    },
+    {
+      "epoch": 0.6821038857863869,
+      "grad_norm": 0.42430946230888367,
+      "learning_rate": 4.585959006388898e-06,
+      "loss": 0.288,
+      "step": 18245
+    },
+    {
+      "epoch": 0.6822908147767367,
+      "grad_norm": 0.4555477499961853,
+      "learning_rate": 4.581022499446468e-06,
+      "loss": 0.3142,
+      "step": 18250
+    },
+    {
+      "epoch": 0.6824777437670865,
+      "grad_norm": 1.2911605834960938,
+      "learning_rate": 4.576087861354101e-06,
+      "loss": 0.54,
+      "step": 18255
+    },
+    {
+      "epoch": 0.6826646727574363,
+      "grad_norm": 0.3238535523414612,
+      "learning_rate": 4.571155093813619e-06,
+      "loss": 0.2987,
+      "step": 18260
+    },
+    {
+      "epoch": 0.682851601747786,
+      "grad_norm": 0.2774864137172699,
+      "learning_rate": 4.5662241985261865e-06,
+      "loss": 0.2595,
+      "step": 18265
+    },
+    {
+      "epoch": 0.6830385307381358,
+      "grad_norm": 0.3135109543800354,
+      "learning_rate": 4.5612951771923345e-06,
+      "loss": 0.2846,
+      "step": 18270
+    },
+    {
+      "epoch": 0.6832254597284857,
+      "grad_norm": 0.6082348823547363,
+      "learning_rate": 4.556368031511932e-06,
+      "loss": 0.2893,
+      "step": 18275
+    },
+    {
+      "epoch": 0.6834123887188355,
+      "grad_norm": 0.5355720520019531,
+      "learning_rate": 4.551442763184221e-06,
+      "loss": 0.3026,
+      "step": 18280
+    },
+    {
+      "epoch": 0.6835993177091853,
+      "grad_norm": 0.6126582026481628,
+      "learning_rate": 4.546519373907778e-06,
+      "loss": 0.2891,
+      "step": 18285
+    },
+    {
+      "epoch": 0.683786246699535,
+      "grad_norm": 0.47081178426742554,
+      "learning_rate": 4.541597865380539e-06,
+      "loss": 0.245,
+      "step": 18290
+    },
+    {
+      "epoch": 0.6839731756898848,
+      "grad_norm": 0.6696823835372925,
+      "learning_rate": 4.536678239299797e-06,
+      "loss": 0.3148,
+      "step": 18295
+    },
+    {
+      "epoch": 0.6841601046802346,
+      "grad_norm": 0.41798293590545654,
+      "learning_rate": 4.531760497362181e-06,
+      "loss": 0.2587,
+      "step": 18300
+    },
+    {
+      "epoch": 0.6843470336705844,
+      "grad_norm": 0.4290241003036499,
+      "learning_rate": 4.526844641263689e-06,
+      "loss": 0.2839,
+      "step": 18305
+    },
+    {
+      "epoch": 0.6845339626609341,
+      "grad_norm": 0.4475858509540558,
+      "learning_rate": 4.521930672699651e-06,
+      "loss": 0.3038,
+      "step": 18310
+    },
+    {
+      "epoch": 0.6847208916512839,
+      "grad_norm": 0.3587439954280853,
+      "learning_rate": 4.517018593364761e-06,
+      "loss": 0.2565,
+      "step": 18315
+    },
+    {
+      "epoch": 0.6849078206416338,
+      "grad_norm": 0.6752985715866089,
+      "learning_rate": 4.512108404953048e-06,
+      "loss": 0.2591,
+      "step": 18320
+    },
+    {
+      "epoch": 0.6850947496319836,
+      "grad_norm": 0.3796422779560089,
+      "learning_rate": 4.507200109157901e-06,
+      "loss": 0.2285,
+      "step": 18325
+    },
+    {
+      "epoch": 0.6852816786223334,
+      "grad_norm": 0.23058107495307922,
+      "learning_rate": 4.502293707672044e-06,
+      "loss": 0.2528,
+      "step": 18330
+    },
+    {
+      "epoch": 0.6854686076126831,
+      "grad_norm": 0.29818716645240784,
+      "learning_rate": 4.497389202187562e-06,
+      "loss": 0.265,
+      "step": 18335
+    },
+    {
+      "epoch": 0.6856555366030329,
+      "grad_norm": 0.48177286982536316,
+      "learning_rate": 4.492486594395875e-06,
+      "loss": 0.2451,
+      "step": 18340
+    },
+    {
+      "epoch": 0.6858424655933827,
+      "grad_norm": 0.4235864281654358,
+      "learning_rate": 4.487585885987747e-06,
+      "loss": 0.3306,
+      "step": 18345
+    },
+    {
+      "epoch": 0.6860293945837325,
+      "grad_norm": 0.5081339478492737,
+      "learning_rate": 4.4826870786533e-06,
+      "loss": 0.3331,
+      "step": 18350
+    },
+    {
+      "epoch": 0.6862163235740822,
+      "grad_norm": 0.5094728469848633,
+      "learning_rate": 4.477790174081984e-06,
+      "loss": 0.2585,
+      "step": 18355
+    },
+    {
+      "epoch": 0.6864032525644321,
+      "grad_norm": 0.29224979877471924,
+      "learning_rate": 4.47289517396261e-06,
+      "loss": 0.2684,
+      "step": 18360
+    },
+    {
+      "epoch": 0.6865901815547819,
+      "grad_norm": 0.2801983058452606,
+      "learning_rate": 4.468002079983315e-06,
+      "loss": 0.3254,
+      "step": 18365
+    },
+    {
+      "epoch": 0.6867771105451317,
+      "grad_norm": 0.688702404499054,
+      "learning_rate": 4.463110893831596e-06,
+      "loss": 0.2631,
+      "step": 18370
+    },
+    {
+      "epoch": 0.6869640395354815,
+      "grad_norm": 0.43564876914024353,
+      "learning_rate": 4.458221617194273e-06,
+      "loss": 0.2727,
+      "step": 18375
+    },
+    {
+      "epoch": 0.6871509685258312,
+      "grad_norm": 0.28386905789375305,
+      "learning_rate": 4.453334251757526e-06,
+      "loss": 0.3505,
+      "step": 18380
+    },
+    {
+      "epoch": 0.687337897516181,
+      "grad_norm": 0.3805743157863617,
+      "learning_rate": 4.448448799206863e-06,
+      "loss": 0.2635,
+      "step": 18385
+    },
+    {
+      "epoch": 0.6875248265065308,
+      "grad_norm": 0.2629093825817108,
+      "learning_rate": 4.443565261227134e-06,
+      "loss": 0.2978,
+      "step": 18390
+    },
+    {
+      "epoch": 0.6877117554968806,
+      "grad_norm": 0.26355159282684326,
+      "learning_rate": 4.438683639502538e-06,
+      "loss": 0.3312,
+      "step": 18395
+    },
+    {
+      "epoch": 0.6878986844872305,
+      "grad_norm": 0.42087700963020325,
+      "learning_rate": 4.4338039357165985e-06,
+      "loss": 0.2332,
+      "step": 18400
+    },
+    {
+      "epoch": 0.6880856134775802,
+      "grad_norm": 0.42999762296676636,
+      "learning_rate": 4.428926151552194e-06,
+      "loss": 0.3205,
+      "step": 18405
+    },
+    {
+      "epoch": 0.68827254246793,
+      "grad_norm": 0.38532260060310364,
+      "learning_rate": 4.424050288691525e-06,
+      "loss": 0.2726,
+      "step": 18410
+    },
+    {
+      "epoch": 0.6884594714582798,
+      "grad_norm": 0.40023544430732727,
+      "learning_rate": 4.419176348816144e-06,
+      "loss": 0.2823,
+      "step": 18415
+    },
+    {
+      "epoch": 0.6886464004486296,
+      "grad_norm": 0.35504770278930664,
+      "learning_rate": 4.414304333606926e-06,
+      "loss": 0.2744,
+      "step": 18420
+    },
+    {
+      "epoch": 0.6888333294389793,
+      "grad_norm": 0.5017563700675964,
+      "learning_rate": 4.409434244744095e-06,
+      "loss": 0.3578,
+      "step": 18425
+    },
+    {
+      "epoch": 0.6890202584293291,
+      "grad_norm": 0.4949704706668854,
+      "learning_rate": 4.4045660839072045e-06,
+      "loss": 0.2341,
+      "step": 18430
+    },
+    {
+      "epoch": 0.6892071874196789,
+      "grad_norm": 0.40918684005737305,
+      "learning_rate": 4.399699852775138e-06,
+      "loss": 0.2464,
+      "step": 18435
+    },
+    {
+      "epoch": 0.6893941164100288,
+      "grad_norm": 0.7892428040504456,
+      "learning_rate": 4.394835553026128e-06,
+      "loss": 0.4378,
+      "step": 18440
+    },
+    {
+      "epoch": 0.6895810454003786,
+      "grad_norm": 0.2694770395755768,
+      "learning_rate": 4.3899731863377225e-06,
+      "loss": 0.2667,
+      "step": 18445
+    },
+    {
+      "epoch": 0.6897679743907283,
+      "grad_norm": 0.35795870423316956,
+      "learning_rate": 4.385112754386821e-06,
+      "loss": 0.2793,
+      "step": 18450
+    },
+    {
+      "epoch": 0.6899549033810781,
+      "grad_norm": 0.3480895161628723,
+      "learning_rate": 4.380254258849641e-06,
+      "loss": 0.2119,
+      "step": 18455
+    },
+    {
+      "epoch": 0.6901418323714279,
+      "grad_norm": 0.5051251649856567,
+      "learning_rate": 4.375397701401745e-06,
+      "loss": 0.2969,
+      "step": 18460
+    },
+    {
+      "epoch": 0.6903287613617777,
+      "grad_norm": 0.3847442865371704,
+      "learning_rate": 4.370543083718012e-06,
+      "loss": 0.2343,
+      "step": 18465
+    },
+    {
+      "epoch": 0.6905156903521275,
+      "grad_norm": 0.3477119207382202,
+      "learning_rate": 4.365690407472668e-06,
+      "loss": 0.2356,
+      "step": 18470
+    },
+    {
+      "epoch": 0.6907026193424772,
+      "grad_norm": 0.6154903769493103,
+      "learning_rate": 4.36083967433926e-06,
+      "loss": 0.2936,
+      "step": 18475
+    },
+    {
+      "epoch": 0.6908895483328271,
+      "grad_norm": 0.5352329611778259,
+      "learning_rate": 4.355990885990663e-06,
+      "loss": 0.2828,
+      "step": 18480
+    },
+    {
+      "epoch": 0.6910764773231769,
+      "grad_norm": 0.39464572072029114,
+      "learning_rate": 4.351144044099091e-06,
+      "loss": 0.2015,
+      "step": 18485
+    },
+    {
+      "epoch": 0.6912634063135267,
+      "grad_norm": 0.483270525932312,
+      "learning_rate": 4.346299150336074e-06,
+      "loss": 0.2555,
+      "step": 18490
+    },
+    {
+      "epoch": 0.6914503353038765,
+      "grad_norm": 0.39474979043006897,
+      "learning_rate": 4.341456206372485e-06,
+      "loss": 0.3986,
+      "step": 18495
+    },
+    {
+      "epoch": 0.6916372642942262,
+      "grad_norm": 0.5631116628646851,
+      "learning_rate": 4.336615213878509e-06,
+      "loss": 0.2867,
+      "step": 18500
+    },
+    {
+      "epoch": 0.691824193284576,
+      "grad_norm": 0.623786985874176,
+      "learning_rate": 4.331776174523673e-06,
+      "loss": 0.3206,
+      "step": 18505
+    },
+    {
+      "epoch": 0.6920111222749258,
+      "grad_norm": 0.2712309658527374,
+      "learning_rate": 4.326939089976815e-06,
+      "loss": 0.3039,
+      "step": 18510
+    },
+    {
+      "epoch": 0.6921980512652756,
+      "grad_norm": 0.34029409289360046,
+      "learning_rate": 4.322103961906113e-06,
+      "loss": 0.2136,
+      "step": 18515
+    },
+    {
+      "epoch": 0.6923849802556254,
+      "grad_norm": 0.52415931224823,
+      "learning_rate": 4.317270791979063e-06,
+      "loss": 0.1956,
+      "step": 18520
+    },
+    {
+      "epoch": 0.6925719092459752,
+      "grad_norm": 0.38135063648223877,
+      "learning_rate": 4.312439581862488e-06,
+      "loss": 0.2654,
+      "step": 18525
+    },
+    {
+      "epoch": 0.692758838236325,
+      "grad_norm": 0.8677012920379639,
+      "learning_rate": 4.307610333222532e-06,
+      "loss": 0.28,
+      "step": 18530
+    },
+    {
+      "epoch": 0.6929457672266748,
+      "grad_norm": 0.19962137937545776,
+      "learning_rate": 4.30278304772466e-06,
+      "loss": 0.2984,
+      "step": 18535
+    },
+    {
+      "epoch": 0.6931326962170246,
+      "grad_norm": 0.4274756610393524,
+      "learning_rate": 4.297957727033673e-06,
+      "loss": 0.2288,
+      "step": 18540
+    },
+    {
+      "epoch": 0.6933196252073743,
+      "grad_norm": 0.3446785509586334,
+      "learning_rate": 4.293134372813678e-06,
+      "loss": 0.2482,
+      "step": 18545
+    },
+    {
+      "epoch": 0.6935065541977241,
+      "grad_norm": 0.30156296491622925,
+      "learning_rate": 4.288312986728119e-06,
+      "loss": 0.254,
+      "step": 18550
+    },
+    {
+      "epoch": 0.6936934831880739,
+      "grad_norm": 0.20253904163837433,
+      "learning_rate": 4.283493570439746e-06,
+      "loss": 0.2544,
+      "step": 18555
+    },
+    {
+      "epoch": 0.6938804121784237,
+      "grad_norm": 0.2638651728630066,
+      "learning_rate": 4.278676125610644e-06,
+      "loss": 0.2614,
+      "step": 18560
+    },
+    {
+      "epoch": 0.6940673411687736,
+      "grad_norm": 0.2870006859302521,
+      "learning_rate": 4.2738606539022105e-06,
+      "loss": 0.3594,
+      "step": 18565
+    },
+    {
+      "epoch": 0.6942542701591233,
+      "grad_norm": 0.4544857442378998,
+      "learning_rate": 4.269047156975166e-06,
+      "loss": 0.3418,
+      "step": 18570
+    },
+    {
+      "epoch": 0.6944411991494731,
+      "grad_norm": 0.21792225539684296,
+      "learning_rate": 4.264235636489542e-06,
+      "loss": 0.3146,
+      "step": 18575
+    },
+    {
+      "epoch": 0.6946281281398229,
+      "grad_norm": 0.5876854658126831,
+      "learning_rate": 4.2594260941046935e-06,
+      "loss": 0.2753,
+      "step": 18580
+    },
+    {
+      "epoch": 0.6948150571301727,
+      "grad_norm": 0.4777786135673523,
+      "learning_rate": 4.254618531479301e-06,
+      "loss": 0.3524,
+      "step": 18585
+    },
+    {
+      "epoch": 0.6950019861205224,
+      "grad_norm": 0.3921513259410858,
+      "learning_rate": 4.249812950271347e-06,
+      "loss": 0.3673,
+      "step": 18590
+    },
+    {
+      "epoch": 0.6951889151108722,
+      "grad_norm": 0.284984290599823,
+      "learning_rate": 4.245009352138146e-06,
+      "loss": 0.3071,
+      "step": 18595
+    },
+    {
+      "epoch": 0.695375844101222,
+      "grad_norm": 0.4456053078174591,
+      "learning_rate": 4.240207738736315e-06,
+      "loss": 0.3151,
+      "step": 18600
+    },
+    {
+      "epoch": 0.6955627730915719,
+      "grad_norm": 0.5414977669715881,
+      "learning_rate": 4.235408111721796e-06,
+      "loss": 0.249,
+      "step": 18605
+    },
+    {
+      "epoch": 0.6957497020819217,
+      "grad_norm": 0.7804057598114014,
+      "learning_rate": 4.230610472749847e-06,
+      "loss": 0.3312,
+      "step": 18610
+    },
+    {
+      "epoch": 0.6959366310722714,
+      "grad_norm": 0.4957026541233063,
+      "learning_rate": 4.225814823475031e-06,
+      "loss": 0.222,
+      "step": 18615
+    },
+    {
+      "epoch": 0.6961235600626212,
+      "grad_norm": 0.2853618860244751,
+      "learning_rate": 4.221021165551232e-06,
+      "loss": 0.2572,
+      "step": 18620
+    },
+    {
+      "epoch": 0.696310489052971,
+      "grad_norm": 0.4198627769947052,
+      "learning_rate": 4.21622950063164e-06,
+      "loss": 0.2863,
+      "step": 18625
+    },
+    {
+      "epoch": 0.6964974180433208,
+      "grad_norm": 0.25387126207351685,
+      "learning_rate": 4.211439830368771e-06,
+      "loss": 0.2627,
+      "step": 18630
+    },
+    {
+      "epoch": 0.6966843470336705,
+      "grad_norm": 0.3733402192592621,
+      "learning_rate": 4.206652156414437e-06,
+      "loss": 0.3436,
+      "step": 18635
+    },
+    {
+      "epoch": 0.6968712760240203,
+      "grad_norm": 0.24343183636665344,
+      "learning_rate": 4.2018664804197784e-06,
+      "loss": 0.2873,
+      "step": 18640
+    },
+    {
+      "epoch": 0.6970582050143702,
+      "grad_norm": 0.2914699614048004,
+      "learning_rate": 4.19708280403523e-06,
+      "loss": 0.2815,
+      "step": 18645
+    },
+    {
+      "epoch": 0.69724513400472,
+      "grad_norm": 0.5060830116271973,
+      "learning_rate": 4.192301128910546e-06,
+      "loss": 0.2341,
+      "step": 18650
+    },
+    {
+      "epoch": 0.6974320629950698,
+      "grad_norm": 0.6719004511833191,
+      "learning_rate": 4.187521456694797e-06,
+      "loss": 0.2972,
+      "step": 18655
+    },
+    {
+      "epoch": 0.6976189919854195,
+      "grad_norm": 0.49988633394241333,
+      "learning_rate": 4.182743789036346e-06,
+      "loss": 0.3975,
+      "step": 18660
+    },
+    {
+      "epoch": 0.6978059209757693,
+      "grad_norm": 0.3273427188396454,
+      "learning_rate": 4.1779681275828795e-06,
+      "loss": 0.278,
+      "step": 18665
+    },
+    {
+      "epoch": 0.6979928499661191,
+      "grad_norm": 1.3453400135040283,
+      "learning_rate": 4.173194473981379e-06,
+      "loss": 0.3058,
+      "step": 18670
+    },
+    {
+      "epoch": 0.6981797789564689,
+      "grad_norm": 0.30567625164985657,
+      "learning_rate": 4.168422829878148e-06,
+      "loss": 0.3541,
+      "step": 18675
+    },
+    {
+      "epoch": 0.6983667079468187,
+      "grad_norm": 0.45551925897598267,
+      "learning_rate": 4.163653196918784e-06,
+      "loss": 0.334,
+      "step": 18680
+    },
+    {
+      "epoch": 0.6985536369371685,
+      "grad_norm": 0.33953145146369934,
+      "learning_rate": 4.158885576748205e-06,
+      "loss": 0.3156,
+      "step": 18685
+    },
+    {
+      "epoch": 0.6987405659275183,
+      "grad_norm": 0.5441725254058838,
+      "learning_rate": 4.154119971010616e-06,
+      "loss": 0.2464,
+      "step": 18690
+    },
+    {
+      "epoch": 0.6989274949178681,
+      "grad_norm": 0.25098875164985657,
+      "learning_rate": 4.149356381349544e-06,
+      "loss": 0.2644,
+      "step": 18695
+    },
+    {
+      "epoch": 0.6991144239082179,
+      "grad_norm": 0.35483863949775696,
+      "learning_rate": 4.144594809407818e-06,
+      "loss": 0.2527,
+      "step": 18700
+    },
+    {
+      "epoch": 0.6993013528985677,
+      "grad_norm": 0.6076955795288086,
+      "learning_rate": 4.139835256827559e-06,
+      "loss": 0.2728,
+      "step": 18705
+    },
+    {
+      "epoch": 0.6994882818889174,
+      "grad_norm": 0.26839420199394226,
+      "learning_rate": 4.135077725250209e-06,
+      "loss": 0.2568,
+      "step": 18710
+    },
+    {
+      "epoch": 0.6996752108792672,
+      "grad_norm": 0.3890112042427063,
+      "learning_rate": 4.130322216316502e-06,
+      "loss": 0.2509,
+      "step": 18715
+    },
+    {
+      "epoch": 0.699862139869617,
+      "grad_norm": 0.4457304775714874,
+      "learning_rate": 4.125568731666473e-06,
+      "loss": 0.2774,
+      "step": 18720
+    },
+    {
+      "epoch": 0.7000490688599669,
+      "grad_norm": 0.3597055971622467,
+      "learning_rate": 4.120817272939462e-06,
+      "loss": 0.2028,
+      "step": 18725
+    },
+    {
+      "epoch": 0.7002359978503166,
+      "grad_norm": 0.2370787411928177,
+      "learning_rate": 4.116067841774116e-06,
+      "loss": 0.3078,
+      "step": 18730
+    },
+    {
+      "epoch": 0.7004229268406664,
+      "grad_norm": 0.7779338359832764,
+      "learning_rate": 4.111320439808373e-06,
+      "loss": 0.3016,
+      "step": 18735
+    },
+    {
+      "epoch": 0.7006098558310162,
+      "grad_norm": 0.3370744585990906,
+      "learning_rate": 4.106575068679477e-06,
+      "loss": 0.274,
+      "step": 18740
+    },
+    {
+      "epoch": 0.700796784821366,
+      "grad_norm": 0.46344175934791565,
+      "learning_rate": 4.101831730023978e-06,
+      "loss": 0.3209,
+      "step": 18745
+    },
+    {
+      "epoch": 0.7009837138117158,
+      "grad_norm": 0.3550253212451935,
+      "learning_rate": 4.097090425477706e-06,
+      "loss": 0.2878,
+      "step": 18750
+    },
+    {
+      "epoch": 0.7011706428020655,
+      "grad_norm": 0.5622664093971252,
+      "learning_rate": 4.092351156675809e-06,
+      "loss": 0.2869,
+      "step": 18755
+    },
+    {
+      "epoch": 0.7013575717924153,
+      "grad_norm": 0.27434638142585754,
+      "learning_rate": 4.087613925252723e-06,
+      "loss": 0.3175,
+      "step": 18760
+    },
+    {
+      "epoch": 0.7015445007827652,
+      "grad_norm": 0.5039125084877014,
+      "learning_rate": 4.082878732842185e-06,
+      "loss": 0.279,
+      "step": 18765
+    },
+    {
+      "epoch": 0.701731429773115,
+      "grad_norm": 0.37588292360305786,
+      "learning_rate": 4.07814558107722e-06,
+      "loss": 0.2938,
+      "step": 18770
+    },
+    {
+      "epoch": 0.7019183587634648,
+      "grad_norm": 0.36863166093826294,
+      "learning_rate": 4.073414471590165e-06,
+      "loss": 0.3311,
+      "step": 18775
+    },
+    {
+      "epoch": 0.7021052877538145,
+      "grad_norm": 0.3147136867046356,
+      "learning_rate": 4.068685406012637e-06,
+      "loss": 0.3818,
+      "step": 18780
+    },
+    {
+      "epoch": 0.7022922167441643,
+      "grad_norm": 0.3484380841255188,
+      "learning_rate": 4.06395838597556e-06,
+      "loss": 0.3061,
+      "step": 18785
+    },
+    {
+      "epoch": 0.7024791457345141,
+      "grad_norm": 0.34823089838027954,
+      "learning_rate": 4.059233413109148e-06,
+      "loss": 0.2871,
+      "step": 18790
+    },
+    {
+      "epoch": 0.7026660747248639,
+      "grad_norm": 0.21362809836864471,
+      "learning_rate": 4.054510489042906e-06,
+      "loss": 0.2635,
+      "step": 18795
+    },
+    {
+      "epoch": 0.7028530037152136,
+      "grad_norm": 0.4324122369289398,
+      "learning_rate": 4.049789615405638e-06,
+      "loss": 0.2149,
+      "step": 18800
+    },
+    {
+      "epoch": 0.7030399327055634,
+      "grad_norm": 0.6976597905158997,
+      "learning_rate": 4.0450707938254385e-06,
+      "loss": 0.2078,
+      "step": 18805
+    },
+    {
+      "epoch": 0.7032268616959133,
+      "grad_norm": 0.7980237603187561,
+      "learning_rate": 4.0403540259296905e-06,
+      "loss": 0.3072,
+      "step": 18810
+    },
+    {
+      "epoch": 0.7034137906862631,
+      "grad_norm": 0.3287063241004944,
+      "learning_rate": 4.03563931334507e-06,
+      "loss": 0.2836,
+      "step": 18815
+    },
+    {
+      "epoch": 0.7036007196766129,
+      "grad_norm": 0.6408100128173828,
+      "learning_rate": 4.030926657697554e-06,
+      "loss": 0.2664,
+      "step": 18820
+    },
+    {
+      "epoch": 0.7037876486669626,
+      "grad_norm": 0.44458410143852234,
+      "learning_rate": 4.0262160606123946e-06,
+      "loss": 0.3106,
+      "step": 18825
+    },
+    {
+      "epoch": 0.7039745776573124,
+      "grad_norm": 0.19885852932929993,
+      "learning_rate": 4.021507523714145e-06,
+      "loss": 0.2218,
+      "step": 18830
+    },
+    {
+      "epoch": 0.7041615066476622,
+      "grad_norm": 0.46404361724853516,
+      "learning_rate": 4.016801048626648e-06,
+      "loss": 0.251,
+      "step": 18835
+    },
+    {
+      "epoch": 0.704348435638012,
+      "grad_norm": 0.6827042102813721,
+      "learning_rate": 4.012096636973027e-06,
+      "loss": 0.2417,
+      "step": 18840
+    },
+    {
+      "epoch": 0.7045353646283617,
+      "grad_norm": 0.3932143747806549,
+      "learning_rate": 4.007394290375703e-06,
+      "loss": 0.3334,
+      "step": 18845
+    },
+    {
+      "epoch": 0.7047222936187116,
+      "grad_norm": 0.5734818577766418,
+      "learning_rate": 4.002694010456379e-06,
+      "loss": 0.3273,
+      "step": 18850
+    },
+    {
+      "epoch": 0.7049092226090614,
+      "grad_norm": 0.6905113458633423,
+      "learning_rate": 3.997995798836046e-06,
+      "loss": 0.2476,
+      "step": 18855
+    },
+    {
+      "epoch": 0.7050961515994112,
+      "grad_norm": 0.5125166177749634,
+      "learning_rate": 3.993299657134979e-06,
+      "loss": 0.2125,
+      "step": 18860
+    },
+    {
+      "epoch": 0.705283080589761,
+      "grad_norm": 0.22294095158576965,
+      "learning_rate": 3.98860558697275e-06,
+      "loss": 0.2655,
+      "step": 18865
+    },
+    {
+      "epoch": 0.7054700095801107,
+      "grad_norm": 0.5617648959159851,
+      "learning_rate": 3.983913589968202e-06,
+      "loss": 0.3351,
+      "step": 18870
+    },
+    {
+      "epoch": 0.7056569385704605,
+      "grad_norm": 0.2911111116409302,
+      "learning_rate": 3.979223667739475e-06,
+      "loss": 0.2118,
+      "step": 18875
+    },
+    {
+      "epoch": 0.7058438675608103,
+      "grad_norm": 0.49967023730278015,
+      "learning_rate": 3.974535821903992e-06,
+      "loss": 0.2808,
+      "step": 18880
+    },
+    {
+      "epoch": 0.7060307965511601,
+      "grad_norm": 0.20392078161239624,
+      "learning_rate": 3.969850054078448e-06,
+      "loss": 0.2092,
+      "step": 18885
+    },
+    {
+      "epoch": 0.70621772554151,
+      "grad_norm": 0.31886929273605347,
+      "learning_rate": 3.965166365878839e-06,
+      "loss": 0.2398,
+      "step": 18890
+    },
+    {
+      "epoch": 0.7064046545318597,
+      "grad_norm": 0.4940277934074402,
+      "learning_rate": 3.960484758920426e-06,
+      "loss": 0.2671,
+      "step": 18895
+    },
+    {
+      "epoch": 0.7065915835222095,
+      "grad_norm": 0.4553227424621582,
+      "learning_rate": 3.95580523481777e-06,
+      "loss": 0.3112,
+      "step": 18900
+    },
+    {
+      "epoch": 0.7067785125125593,
+      "grad_norm": 0.4576134979724884,
+      "learning_rate": 3.9511277951847e-06,
+      "loss": 0.302,
+      "step": 18905
+    },
+    {
+      "epoch": 0.7069654415029091,
+      "grad_norm": 0.4832530617713928,
+      "learning_rate": 3.946452441634332e-06,
+      "loss": 0.2758,
+      "step": 18910
+    },
+    {
+      "epoch": 0.7071523704932589,
+      "grad_norm": 0.4215802550315857,
+      "learning_rate": 3.9417791757790565e-06,
+      "loss": 0.2889,
+      "step": 18915
+    },
+    {
+      "epoch": 0.7073392994836086,
+      "grad_norm": 0.6536740660667419,
+      "learning_rate": 3.937107999230554e-06,
+      "loss": 0.3214,
+      "step": 18920
+    },
+    {
+      "epoch": 0.7075262284739584,
+      "grad_norm": 0.4992019832134247,
+      "learning_rate": 3.932438913599781e-06,
+      "loss": 0.2528,
+      "step": 18925
+    },
+    {
+      "epoch": 0.7077131574643083,
+      "grad_norm": 0.21796809136867523,
+      "learning_rate": 3.927771920496967e-06,
+      "loss": 0.3074,
+      "step": 18930
+    },
+    {
+      "epoch": 0.7079000864546581,
+      "grad_norm": 0.25290024280548096,
+      "learning_rate": 3.923107021531629e-06,
+      "loss": 0.3147,
+      "step": 18935
+    },
+    {
+      "epoch": 0.7080870154450079,
+      "grad_norm": 0.2794661521911621,
+      "learning_rate": 3.918444218312551e-06,
+      "loss": 0.304,
+      "step": 18940
+    },
+    {
+      "epoch": 0.7082739444353576,
+      "grad_norm": 0.9677742719650269,
+      "learning_rate": 3.913783512447806e-06,
+      "loss": 0.3212,
+      "step": 18945
+    },
+    {
+      "epoch": 0.7084608734257074,
+      "grad_norm": 0.3141147792339325,
+      "learning_rate": 3.909124905544737e-06,
+      "loss": 0.2581,
+      "step": 18950
+    },
+    {
+      "epoch": 0.7086478024160572,
+      "grad_norm": 0.8014910817146301,
+      "learning_rate": 3.9044683992099616e-06,
+      "loss": 0.3311,
+      "step": 18955
+    },
+    {
+      "epoch": 0.708834731406407,
+      "grad_norm": 0.2427234798669815,
+      "learning_rate": 3.899813995049373e-06,
+      "loss": 0.285,
+      "step": 18960
+    },
+    {
+      "epoch": 0.7090216603967567,
+      "grad_norm": 0.49129942059516907,
+      "learning_rate": 3.895161694668144e-06,
+      "loss": 0.2669,
+      "step": 18965
+    },
+    {
+      "epoch": 0.7092085893871066,
+      "grad_norm": 0.351224422454834,
+      "learning_rate": 3.890511499670726e-06,
+      "loss": 0.2503,
+      "step": 18970
+    },
+    {
+      "epoch": 0.7093955183774564,
+      "grad_norm": 0.4046824872493744,
+      "learning_rate": 3.885863411660829e-06,
+      "loss": 0.2601,
+      "step": 18975
+    },
+    {
+      "epoch": 0.7095824473678062,
+      "grad_norm": 0.49013751745224,
+      "learning_rate": 3.881217432241451e-06,
+      "loss": 0.2791,
+      "step": 18980
+    },
+    {
+      "epoch": 0.709769376358156,
+      "grad_norm": 0.3391354978084564,
+      "learning_rate": 3.876573563014854e-06,
+      "loss": 0.2573,
+      "step": 18985
+    },
+    {
+      "epoch": 0.7099563053485057,
+      "grad_norm": 0.28667691349983215,
+      "learning_rate": 3.8719318055825785e-06,
+      "loss": 0.2763,
+      "step": 18990
+    },
+    {
+      "epoch": 0.7101432343388555,
+      "grad_norm": 0.2968943417072296,
+      "learning_rate": 3.8672921615454325e-06,
+      "loss": 0.2952,
+      "step": 18995
+    },
+    {
+      "epoch": 0.7103301633292053,
+      "grad_norm": 0.2967950403690338,
+      "learning_rate": 3.862654632503495e-06,
+      "loss": 0.3106,
+      "step": 19000
+    },
+    {
+      "epoch": 0.7105170923195551,
+      "grad_norm": 0.21600590646266937,
+      "learning_rate": 3.858019220056115e-06,
+      "loss": 0.1917,
+      "step": 19005
+    },
+    {
+      "epoch": 0.710704021309905,
+      "grad_norm": 0.5353944897651672,
+      "learning_rate": 3.853385925801916e-06,
+      "loss": 0.2285,
+      "step": 19010
+    },
+    {
+      "epoch": 0.7108909503002547,
+      "grad_norm": 0.44608166813850403,
+      "learning_rate": 3.848754751338792e-06,
+      "loss": 0.2501,
+      "step": 19015
+    },
+    {
+      "epoch": 0.7110778792906045,
+      "grad_norm": 0.38213208317756653,
+      "learning_rate": 3.844125698263896e-06,
+      "loss": 0.2357,
+      "step": 19020
+    },
+    {
+      "epoch": 0.7112648082809543,
+      "grad_norm": 0.2469312697649002,
+      "learning_rate": 3.8394987681736596e-06,
+      "loss": 0.2981,
+      "step": 19025
+    },
+    {
+      "epoch": 0.7114517372713041,
+      "grad_norm": 0.30986660718917847,
+      "learning_rate": 3.834873962663775e-06,
+      "loss": 0.3125,
+      "step": 19030
+    },
+    {
+      "epoch": 0.7116386662616538,
+      "grad_norm": 0.5508708357810974,
+      "learning_rate": 3.830251283329211e-06,
+      "loss": 0.2756,
+      "step": 19035
+    },
+    {
+      "epoch": 0.7118255952520036,
+      "grad_norm": 0.44314640760421753,
+      "learning_rate": 3.825630731764195e-06,
+      "loss": 0.3161,
+      "step": 19040
+    },
+    {
+      "epoch": 0.7120125242423534,
+      "grad_norm": 0.3538496494293213,
+      "learning_rate": 3.8210123095622164e-06,
+      "loss": 0.2465,
+      "step": 19045
+    },
+    {
+      "epoch": 0.7121994532327032,
+      "grad_norm": 0.4966811537742615,
+      "learning_rate": 3.816396018316047e-06,
+      "loss": 0.2047,
+      "step": 19050
+    },
+    {
+      "epoch": 0.7123863822230531,
+      "grad_norm": 0.4113473892211914,
+      "learning_rate": 3.8117818596177035e-06,
+      "loss": 0.232,
+      "step": 19055
+    },
+    {
+      "epoch": 0.7125733112134028,
+      "grad_norm": 0.42282649874687195,
+      "learning_rate": 3.807169835058485e-06,
+      "loss": 0.3454,
+      "step": 19060
+    },
+    {
+      "epoch": 0.7127602402037526,
+      "grad_norm": 0.495076060295105,
+      "learning_rate": 3.8025599462289407e-06,
+      "loss": 0.2909,
+      "step": 19065
+    },
+    {
+      "epoch": 0.7129471691941024,
+      "grad_norm": 0.4929969608783722,
+      "learning_rate": 3.7979521947188946e-06,
+      "loss": 0.3251,
+      "step": 19070
+    },
+    {
+      "epoch": 0.7131340981844522,
+      "grad_norm": 0.2015666663646698,
+      "learning_rate": 3.793346582117422e-06,
+      "loss": 0.2641,
+      "step": 19075
+    },
+    {
+      "epoch": 0.713321027174802,
+      "grad_norm": 0.42519763112068176,
+      "learning_rate": 3.7887431100128746e-06,
+      "loss": 0.3006,
+      "step": 19080
+    },
+    {
+      "epoch": 0.7135079561651517,
+      "grad_norm": 0.543765664100647,
+      "learning_rate": 3.784141779992849e-06,
+      "loss": 0.319,
+      "step": 19085
+    },
+    {
+      "epoch": 0.7136948851555015,
+      "grad_norm": 0.39010390639305115,
+      "learning_rate": 3.77954259364422e-06,
+      "loss": 0.2191,
+      "step": 19090
+    },
+    {
+      "epoch": 0.7138818141458514,
+      "grad_norm": 0.6654005646705627,
+      "learning_rate": 3.7749455525531122e-06,
+      "loss": 0.2967,
+      "step": 19095
+    },
+    {
+      "epoch": 0.7140687431362012,
+      "grad_norm": 0.3600732982158661,
+      "learning_rate": 3.7703506583049097e-06,
+      "loss": 0.333,
+      "step": 19100
+    },
+    {
+      "epoch": 0.714255672126551,
+      "grad_norm": 0.42062684893608093,
+      "learning_rate": 3.765757912484266e-06,
+      "loss": 0.3078,
+      "step": 19105
+    },
+    {
+      "epoch": 0.7144426011169007,
+      "grad_norm": 0.324552983045578,
+      "learning_rate": 3.7611673166750816e-06,
+      "loss": 0.2595,
+      "step": 19110
+    },
+    {
+      "epoch": 0.7146295301072505,
+      "grad_norm": 0.522148072719574,
+      "learning_rate": 3.7565788724605278e-06,
+      "loss": 0.2575,
+      "step": 19115
+    },
+    {
+      "epoch": 0.7148164590976003,
+      "grad_norm": 0.2913423776626587,
+      "learning_rate": 3.751992581423021e-06,
+      "loss": 0.2701,
+      "step": 19120
+    },
+    {
+      "epoch": 0.71500338808795,
+      "grad_norm": 0.5544187426567078,
+      "learning_rate": 3.7474084451442484e-06,
+      "loss": 0.2332,
+      "step": 19125
+    },
+    {
+      "epoch": 0.7151903170782998,
+      "grad_norm": 0.7291717529296875,
+      "learning_rate": 3.7428264652051393e-06,
+      "loss": 0.3321,
+      "step": 19130
+    },
+    {
+      "epoch": 0.7153772460686497,
+      "grad_norm": 0.5494731664657593,
+      "learning_rate": 3.7382466431858966e-06,
+      "loss": 0.318,
+      "step": 19135
+    },
+    {
+      "epoch": 0.7155641750589995,
+      "grad_norm": 0.3049234449863434,
+      "learning_rate": 3.733668980665963e-06,
+      "loss": 0.2599,
+      "step": 19140
+    },
+    {
+      "epoch": 0.7157511040493493,
+      "grad_norm": 0.3012910783290863,
+      "learning_rate": 3.729093479224043e-06,
+      "loss": 0.2457,
+      "step": 19145
+    },
+    {
+      "epoch": 0.715938033039699,
+      "grad_norm": 0.23188547790050507,
+      "learning_rate": 3.7245201404381006e-06,
+      "loss": 0.2909,
+      "step": 19150
+    },
+    {
+      "epoch": 0.7161249620300488,
+      "grad_norm": 0.1563846319913864,
+      "learning_rate": 3.7199489658853428e-06,
+      "loss": 0.3587,
+      "step": 19155
+    },
+    {
+      "epoch": 0.7163118910203986,
+      "grad_norm": 0.34965288639068604,
+      "learning_rate": 3.7153799571422444e-06,
+      "loss": 0.2779,
+      "step": 19160
+    },
+    {
+      "epoch": 0.7164988200107484,
+      "grad_norm": 0.3783499300479889,
+      "learning_rate": 3.710813115784517e-06,
+      "loss": 0.3366,
+      "step": 19165
+    },
+    {
+      "epoch": 0.7166857490010982,
+      "grad_norm": 0.5413819551467896,
+      "learning_rate": 3.7062484433871426e-06,
+      "loss": 0.3233,
+      "step": 19170
+    },
+    {
+      "epoch": 0.716872677991448,
+      "grad_norm": 0.43029576539993286,
+      "learning_rate": 3.7016859415243377e-06,
+      "loss": 0.2896,
+      "step": 19175
+    },
+    {
+      "epoch": 0.7170596069817978,
+      "grad_norm": 0.23014762997627258,
+      "learning_rate": 3.6971256117695853e-06,
+      "loss": 0.34,
+      "step": 19180
+    },
+    {
+      "epoch": 0.7172465359721476,
+      "grad_norm": 0.5080450773239136,
+      "learning_rate": 3.692567455695609e-06,
+      "loss": 0.2345,
+      "step": 19185
+    },
+    {
+      "epoch": 0.7174334649624974,
+      "grad_norm": 0.36543214321136475,
+      "learning_rate": 3.6880114748743832e-06,
+      "loss": 0.3117,
+      "step": 19190
+    },
+    {
+      "epoch": 0.7176203939528472,
+      "grad_norm": 0.38558727502822876,
+      "learning_rate": 3.683457670877142e-06,
+      "loss": 0.2428,
+      "step": 19195
+    },
+    {
+      "epoch": 0.7178073229431969,
+      "grad_norm": 0.16458238661289215,
+      "learning_rate": 3.678906045274355e-06,
+      "loss": 0.2913,
+      "step": 19200
+    },
+    {
+      "epoch": 0.7179942519335467,
+      "grad_norm": 0.4184258282184601,
+      "learning_rate": 3.6743565996357534e-06,
+      "loss": 0.3226,
+      "step": 19205
+    },
+    {
+      "epoch": 0.7181811809238965,
+      "grad_norm": 0.33344846963882446,
+      "learning_rate": 3.6698093355303054e-06,
+      "loss": 0.3476,
+      "step": 19210
+    },
+    {
+      "epoch": 0.7183681099142464,
+      "grad_norm": 0.9731647372245789,
+      "learning_rate": 3.6652642545262374e-06,
+      "loss": 0.4277,
+      "step": 19215
+    },
+    {
+      "epoch": 0.7185550389045962,
+      "grad_norm": 0.38592931628227234,
+      "learning_rate": 3.6607213581910116e-06,
+      "loss": 0.3507,
+      "step": 19220
+    },
+    {
+      "epoch": 0.7187419678949459,
+      "grad_norm": 0.5030714273452759,
+      "learning_rate": 3.65618064809135e-06,
+      "loss": 0.2249,
+      "step": 19225
+    },
+    {
+      "epoch": 0.7189288968852957,
+      "grad_norm": 0.3900645673274994,
+      "learning_rate": 3.6516421257932054e-06,
+      "loss": 0.304,
+      "step": 19230
+    },
+    {
+      "epoch": 0.7191158258756455,
+      "grad_norm": 0.4885767698287964,
+      "learning_rate": 3.6471057928617913e-06,
+      "loss": 0.2138,
+      "step": 19235
+    },
+    {
+      "epoch": 0.7193027548659953,
+      "grad_norm": 0.33923253417015076,
+      "learning_rate": 3.6425716508615574e-06,
+      "loss": 0.2623,
+      "step": 19240
+    },
+    {
+      "epoch": 0.719489683856345,
+      "grad_norm": 0.3621087670326233,
+      "learning_rate": 3.638039701356193e-06,
+      "loss": 0.2484,
+      "step": 19245
+    },
+    {
+      "epoch": 0.7196766128466948,
+      "grad_norm": 0.18470481038093567,
+      "learning_rate": 3.6335099459086453e-06,
+      "loss": 0.3044,
+      "step": 19250
+    },
+    {
+      "epoch": 0.7198635418370447,
+      "grad_norm": 0.3943520188331604,
+      "learning_rate": 3.6289823860810926e-06,
+      "loss": 0.2953,
+      "step": 19255
+    },
+    {
+      "epoch": 0.7200504708273945,
+      "grad_norm": 0.47943511605262756,
+      "learning_rate": 3.624457023434964e-06,
+      "loss": 0.3055,
+      "step": 19260
+    },
+    {
+      "epoch": 0.7202373998177443,
+      "grad_norm": 0.5454005002975464,
+      "learning_rate": 3.619933859530923e-06,
+      "loss": 0.2572,
+      "step": 19265
+    },
+    {
+      "epoch": 0.720424328808094,
+      "grad_norm": 0.5797293186187744,
+      "learning_rate": 3.6154128959288847e-06,
+      "loss": 0.311,
+      "step": 19270
+    },
+    {
+      "epoch": 0.7206112577984438,
+      "grad_norm": 0.2344737946987152,
+      "learning_rate": 3.610894134187993e-06,
+      "loss": 0.2628,
+      "step": 19275
+    },
+    {
+      "epoch": 0.7207981867887936,
+      "grad_norm": 0.279786616563797,
+      "learning_rate": 3.6063775758666485e-06,
+      "loss": 0.2749,
+      "step": 19280
+    },
+    {
+      "epoch": 0.7209851157791434,
+      "grad_norm": 0.28806987404823303,
+      "learning_rate": 3.601863222522477e-06,
+      "loss": 0.4239,
+      "step": 19285
+    },
+    {
+      "epoch": 0.7211720447694931,
+      "grad_norm": 0.3944340646266937,
+      "learning_rate": 3.5973510757123464e-06,
+      "loss": 0.3092,
+      "step": 19290
+    },
+    {
+      "epoch": 0.7213589737598429,
+      "grad_norm": 0.7157394886016846,
+      "learning_rate": 3.5928411369923743e-06,
+      "loss": 0.3422,
+      "step": 19295
+    },
+    {
+      "epoch": 0.7215459027501928,
+      "grad_norm": 0.6664456129074097,
+      "learning_rate": 3.5883334079179023e-06,
+      "loss": 0.2451,
+      "step": 19300
+    },
+    {
+      "epoch": 0.7217328317405426,
+      "grad_norm": 0.32842952013015747,
+      "learning_rate": 3.583827890043524e-06,
+      "loss": 0.2177,
+      "step": 19305
+    },
+    {
+      "epoch": 0.7219197607308924,
+      "grad_norm": 0.3629484474658966,
+      "learning_rate": 3.5793245849230563e-06,
+      "loss": 0.2519,
+      "step": 19310
+    },
+    {
+      "epoch": 0.7221066897212421,
+      "grad_norm": 0.506600022315979,
+      "learning_rate": 3.574823494109567e-06,
+      "loss": 0.3817,
+      "step": 19315
+    },
+    {
+      "epoch": 0.7222936187115919,
+      "grad_norm": 0.38654983043670654,
+      "learning_rate": 3.5703246191553463e-06,
+      "loss": 0.2811,
+      "step": 19320
+    },
+    {
+      "epoch": 0.7224805477019417,
+      "grad_norm": 0.488619863986969,
+      "learning_rate": 3.565827961611935e-06,
+      "loss": 0.2265,
+      "step": 19325
+    },
+    {
+      "epoch": 0.7226674766922915,
+      "grad_norm": 0.38500773906707764,
+      "learning_rate": 3.561333523030097e-06,
+      "loss": 0.2155,
+      "step": 19330
+    },
+    {
+      "epoch": 0.7228544056826413,
+      "grad_norm": 0.3992687463760376,
+      "learning_rate": 3.5568413049598326e-06,
+      "loss": 0.2698,
+      "step": 19335
+    },
+    {
+      "epoch": 0.7230413346729911,
+      "grad_norm": 0.5223041772842407,
+      "learning_rate": 3.552351308950386e-06,
+      "loss": 0.3776,
+      "step": 19340
+    },
+    {
+      "epoch": 0.7232282636633409,
+      "grad_norm": 0.32374194264411926,
+      "learning_rate": 3.54786353655022e-06,
+      "loss": 0.2399,
+      "step": 19345
+    },
+    {
+      "epoch": 0.7234151926536907,
+      "grad_norm": 0.3728601038455963,
+      "learning_rate": 3.5433779893070477e-06,
+      "loss": 0.2624,
+      "step": 19350
+    },
+    {
+      "epoch": 0.7236021216440405,
+      "grad_norm": 0.29935961961746216,
+      "learning_rate": 3.538894668767797e-06,
+      "loss": 0.2591,
+      "step": 19355
+    },
+    {
+      "epoch": 0.7237890506343903,
+      "grad_norm": 0.3965166211128235,
+      "learning_rate": 3.534413576478645e-06,
+      "loss": 0.2904,
+      "step": 19360
+    },
+    {
+      "epoch": 0.72397597962474,
+      "grad_norm": 0.32862594723701477,
+      "learning_rate": 3.5299347139849836e-06,
+      "loss": 0.3974,
+      "step": 19365
+    },
+    {
+      "epoch": 0.7241629086150898,
+      "grad_norm": 0.3905617594718933,
+      "learning_rate": 3.5254580828314524e-06,
+      "loss": 0.2661,
+      "step": 19370
+    },
+    {
+      "epoch": 0.7243498376054396,
+      "grad_norm": 0.7117414474487305,
+      "learning_rate": 3.5209836845619093e-06,
+      "loss": 0.2346,
+      "step": 19375
+    },
+    {
+      "epoch": 0.7245367665957895,
+      "grad_norm": 0.2594413757324219,
+      "learning_rate": 3.5165115207194435e-06,
+      "loss": 0.2891,
+      "step": 19380
+    },
+    {
+      "epoch": 0.7247236955861392,
+      "grad_norm": 0.6064930558204651,
+      "learning_rate": 3.5120415928463813e-06,
+      "loss": 0.2849,
+      "step": 19385
+    },
+    {
+      "epoch": 0.724910624576489,
+      "grad_norm": 0.34241193532943726,
+      "learning_rate": 3.507573902484267e-06,
+      "loss": 0.2779,
+      "step": 19390
+    },
+    {
+      "epoch": 0.7250975535668388,
+      "grad_norm": 0.4620073735713959,
+      "learning_rate": 3.5031084511738855e-06,
+      "loss": 0.3194,
+      "step": 19395
+    },
+    {
+      "epoch": 0.7252844825571886,
+      "grad_norm": 0.36498600244522095,
+      "learning_rate": 3.4986452404552362e-06,
+      "loss": 0.3193,
+      "step": 19400
+    },
+    {
+      "epoch": 0.7254714115475384,
+      "grad_norm": 0.24594563245773315,
+      "learning_rate": 3.49418427186756e-06,
+      "loss": 0.3006,
+      "step": 19405
+    },
+    {
+      "epoch": 0.7256583405378881,
+      "grad_norm": 0.6980525255203247,
+      "learning_rate": 3.4897255469493096e-06,
+      "loss": 0.2412,
+      "step": 19410
+    },
+    {
+      "epoch": 0.7258452695282379,
+      "grad_norm": 0.5322422385215759,
+      "learning_rate": 3.4852690672381785e-06,
+      "loss": 0.2604,
+      "step": 19415
+    },
+    {
+      "epoch": 0.7260321985185878,
+      "grad_norm": 0.27667078375816345,
+      "learning_rate": 3.480814834271072e-06,
+      "loss": 0.306,
+      "step": 19420
+    },
+    {
+      "epoch": 0.7262191275089376,
+      "grad_norm": 0.38868507742881775,
+      "learning_rate": 3.476362849584134e-06,
+      "loss": 0.2896,
+      "step": 19425
+    },
+    {
+      "epoch": 0.7264060564992874,
+      "grad_norm": 0.2022697925567627,
+      "learning_rate": 3.4719131147127237e-06,
+      "loss": 0.3543,
+      "step": 19430
+    },
+    {
+      "epoch": 0.7265929854896371,
+      "grad_norm": 0.41005682945251465,
+      "learning_rate": 3.4674656311914233e-06,
+      "loss": 0.3022,
+      "step": 19435
+    },
+    {
+      "epoch": 0.7267799144799869,
+      "grad_norm": 0.33465057611465454,
+      "learning_rate": 3.463020400554049e-06,
+      "loss": 0.3551,
+      "step": 19440
+    },
+    {
+      "epoch": 0.7269668434703367,
+      "grad_norm": 0.3077739477157593,
+      "learning_rate": 3.4585774243336277e-06,
+      "loss": 0.2626,
+      "step": 19445
+    },
+    {
+      "epoch": 0.7271537724606865,
+      "grad_norm": 1.222378134727478,
+      "learning_rate": 3.45413670406242e-06,
+      "loss": 0.3006,
+      "step": 19450
+    },
+    {
+      "epoch": 0.7273407014510362,
+      "grad_norm": 0.28431418538093567,
+      "learning_rate": 3.449698241271897e-06,
+      "loss": 0.2822,
+      "step": 19455
+    },
+    {
+      "epoch": 0.7275276304413861,
+      "grad_norm": 0.5643324851989746,
+      "learning_rate": 3.445262037492765e-06,
+      "loss": 0.306,
+      "step": 19460
+    },
+    {
+      "epoch": 0.7277145594317359,
+      "grad_norm": 0.39042869210243225,
+      "learning_rate": 3.4408280942549343e-06,
+      "loss": 0.2474,
+      "step": 19465
+    },
+    {
+      "epoch": 0.7279014884220857,
+      "grad_norm": 0.5540686249732971,
+      "learning_rate": 3.436396413087555e-06,
+      "loss": 0.2843,
+      "step": 19470
+    },
+    {
+      "epoch": 0.7280884174124355,
+      "grad_norm": 0.3987734317779541,
+      "learning_rate": 3.4319669955189806e-06,
+      "loss": 0.34,
+      "step": 19475
+    },
+    {
+      "epoch": 0.7282753464027852,
+      "grad_norm": 0.4319436252117157,
+      "learning_rate": 3.427539843076788e-06,
+      "loss": 0.3197,
+      "step": 19480
+    },
+    {
+      "epoch": 0.728462275393135,
+      "grad_norm": 0.3150979280471802,
+      "learning_rate": 3.423114957287783e-06,
+      "loss": 0.3086,
+      "step": 19485
+    },
+    {
+      "epoch": 0.7286492043834848,
+      "grad_norm": 0.34863778948783875,
+      "learning_rate": 3.4186923396779735e-06,
+      "loss": 0.2332,
+      "step": 19490
+    },
+    {
+      "epoch": 0.7288361333738346,
+      "grad_norm": 0.41178667545318604,
+      "learning_rate": 3.414271991772602e-06,
+      "loss": 0.2602,
+      "step": 19495
+    },
+    {
+      "epoch": 0.7290230623641845,
+      "grad_norm": 0.5623154044151306,
+      "learning_rate": 3.4098539150961107e-06,
+      "loss": 0.2037,
+      "step": 19500
+    },
+    {
+      "epoch": 0.7292099913545342,
+      "grad_norm": 0.26398512721061707,
+      "learning_rate": 3.4054381111721767e-06,
+      "loss": 0.2504,
+      "step": 19505
+    },
+    {
+      "epoch": 0.729396920344884,
+      "grad_norm": 0.4237319231033325,
+      "learning_rate": 3.4010245815236775e-06,
+      "loss": 0.2686,
+      "step": 19510
+    },
+    {
+      "epoch": 0.7295838493352338,
+      "grad_norm": 0.3419848084449768,
+      "learning_rate": 3.3966133276727178e-06,
+      "loss": 0.238,
+      "step": 19515
+    },
+    {
+      "epoch": 0.7297707783255836,
+      "grad_norm": 0.34581613540649414,
+      "learning_rate": 3.392204351140611e-06,
+      "loss": 0.2861,
+      "step": 19520
+    },
+    {
+      "epoch": 0.7299577073159333,
+      "grad_norm": 0.27548453211784363,
+      "learning_rate": 3.3877976534478816e-06,
+      "loss": 0.3436,
+      "step": 19525
+    },
+    {
+      "epoch": 0.7301446363062831,
+      "grad_norm": 0.5232365727424622,
+      "learning_rate": 3.383393236114283e-06,
+      "loss": 0.2877,
+      "step": 19530
+    },
+    {
+      "epoch": 0.7303315652966329,
+      "grad_norm": 0.36319199204444885,
+      "learning_rate": 3.378991100658764e-06,
+      "loss": 0.4038,
+      "step": 19535
+    },
+    {
+      "epoch": 0.7305184942869827,
+      "grad_norm": 0.3827814757823944,
+      "learning_rate": 3.3745912485995e-06,
+      "loss": 0.2816,
+      "step": 19540
+    },
+    {
+      "epoch": 0.7307054232773326,
+      "grad_norm": 0.7302720546722412,
+      "learning_rate": 3.370193681453872e-06,
+      "loss": 0.3484,
+      "step": 19545
+    },
+    {
+      "epoch": 0.7308923522676823,
+      "grad_norm": 0.3170529901981354,
+      "learning_rate": 3.3657984007384757e-06,
+      "loss": 0.2089,
+      "step": 19550
+    },
+    {
+      "epoch": 0.7310792812580321,
+      "grad_norm": 0.6026202440261841,
+      "learning_rate": 3.361405407969115e-06,
+      "loss": 0.3252,
+      "step": 19555
+    },
+    {
+      "epoch": 0.7312662102483819,
+      "grad_norm": 0.6313705444335938,
+      "learning_rate": 3.3570147046608125e-06,
+      "loss": 0.2608,
+      "step": 19560
+    },
+    {
+      "epoch": 0.7314531392387317,
+      "grad_norm": 0.6118645071983337,
+      "learning_rate": 3.352626292327793e-06,
+      "loss": 0.303,
+      "step": 19565
+    },
+    {
+      "epoch": 0.7316400682290815,
+      "grad_norm": 0.9984606504440308,
+      "learning_rate": 3.348240172483491e-06,
+      "loss": 0.269,
+      "step": 19570
+    },
+    {
+      "epoch": 0.7318269972194312,
+      "grad_norm": 0.4851844906806946,
+      "learning_rate": 3.3438563466405595e-06,
+      "loss": 0.2853,
+      "step": 19575
+    },
+    {
+      "epoch": 0.732013926209781,
+      "grad_norm": 0.5005273222923279,
+      "learning_rate": 3.339474816310847e-06,
+      "loss": 0.3988,
+      "step": 19580
+    },
+    {
+      "epoch": 0.7322008552001309,
+      "grad_norm": 0.31317126750946045,
+      "learning_rate": 3.3350955830054267e-06,
+      "loss": 0.3109,
+      "step": 19585
+    },
+    {
+      "epoch": 0.7323877841904807,
+      "grad_norm": 0.33924540877342224,
+      "learning_rate": 3.330718648234562e-06,
+      "loss": 0.3024,
+      "step": 19590
+    },
+    {
+      "epoch": 0.7325747131808305,
+      "grad_norm": 0.5410972833633423,
+      "learning_rate": 3.326344013507741e-06,
+      "loss": 0.3328,
+      "step": 19595
+    },
+    {
+      "epoch": 0.7327616421711802,
+      "grad_norm": 0.4295196235179901,
+      "learning_rate": 3.321971680333641e-06,
+      "loss": 0.2226,
+      "step": 19600
+    },
+    {
+      "epoch": 0.73294857116153,
+      "grad_norm": 0.3176816999912262,
+      "learning_rate": 3.317601650220159e-06,
+      "loss": 0.2551,
+      "step": 19605
+    },
+    {
+      "epoch": 0.7331355001518798,
+      "grad_norm": 0.5756362676620483,
+      "learning_rate": 3.313233924674396e-06,
+      "loss": 0.2447,
+      "step": 19610
+    },
+    {
+      "epoch": 0.7333224291422296,
+      "grad_norm": 0.6090577840805054,
+      "learning_rate": 3.3088685052026524e-06,
+      "loss": 0.3359,
+      "step": 19615
+    },
+    {
+      "epoch": 0.7335093581325793,
+      "grad_norm": 0.27702534198760986,
+      "learning_rate": 3.3045053933104366e-06,
+      "loss": 0.2796,
+      "step": 19620
+    },
+    {
+      "epoch": 0.7336962871229292,
+      "grad_norm": 0.3996005654335022,
+      "learning_rate": 3.3001445905024567e-06,
+      "loss": 0.2924,
+      "step": 19625
+    },
+    {
+      "epoch": 0.733883216113279,
+      "grad_norm": 0.4653981029987335,
+      "learning_rate": 3.2957860982826363e-06,
+      "loss": 0.242,
+      "step": 19630
+    },
+    {
+      "epoch": 0.7340701451036288,
+      "grad_norm": 0.43299591541290283,
+      "learning_rate": 3.2914299181540866e-06,
+      "loss": 0.3355,
+      "step": 19635
+    },
+    {
+      "epoch": 0.7342570740939786,
+      "grad_norm": 0.35684916377067566,
+      "learning_rate": 3.287076051619137e-06,
+      "loss": 0.3164,
+      "step": 19640
+    },
+    {
+      "epoch": 0.7344440030843283,
+      "grad_norm": 0.5310947895050049,
+      "learning_rate": 3.282724500179304e-06,
+      "loss": 0.3019,
+      "step": 19645
+    },
+    {
+      "epoch": 0.7346309320746781,
+      "grad_norm": 0.38951942324638367,
+      "learning_rate": 3.2783752653353164e-06,
+      "loss": 0.2736,
+      "step": 19650
+    },
+    {
+      "epoch": 0.7348178610650279,
+      "grad_norm": 0.19010038673877716,
+      "learning_rate": 3.2740283485871038e-06,
+      "loss": 0.2463,
+      "step": 19655
+    },
+    {
+      "epoch": 0.7350047900553777,
+      "grad_norm": 0.3501550853252411,
+      "learning_rate": 3.269683751433791e-06,
+      "loss": 0.2534,
+      "step": 19660
+    },
+    {
+      "epoch": 0.7351917190457276,
+      "grad_norm": 0.40544456243515015,
+      "learning_rate": 3.2653414753737047e-06,
+      "loss": 0.2525,
+      "step": 19665
+    },
+    {
+      "epoch": 0.7353786480360773,
+      "grad_norm": 1.1853599548339844,
+      "learning_rate": 3.261001521904368e-06,
+      "loss": 0.2623,
+      "step": 19670
+    },
+    {
+      "epoch": 0.7355655770264271,
+      "grad_norm": 0.24595674872398376,
+      "learning_rate": 3.2566638925225113e-06,
+      "loss": 0.2254,
+      "step": 19675
+    },
+    {
+      "epoch": 0.7357525060167769,
+      "grad_norm": 0.2667876183986664,
+      "learning_rate": 3.2523285887240553e-06,
+      "loss": 0.3439,
+      "step": 19680
+    },
+    {
+      "epoch": 0.7359394350071267,
+      "grad_norm": 0.44477343559265137,
+      "learning_rate": 3.2479956120041266e-06,
+      "loss": 0.3082,
+      "step": 19685
+    },
+    {
+      "epoch": 0.7361263639974764,
+      "grad_norm": 0.455649197101593,
+      "learning_rate": 3.243664963857038e-06,
+      "loss": 0.2301,
+      "step": 19690
+    },
+    {
+      "epoch": 0.7363132929878262,
+      "grad_norm": 0.4365465044975281,
+      "learning_rate": 3.239336645776311e-06,
+      "loss": 0.2563,
+      "step": 19695
+    },
+    {
+      "epoch": 0.736500221978176,
+      "grad_norm": 0.3654913902282715,
+      "learning_rate": 3.23501065925466e-06,
+      "loss": 0.2998,
+      "step": 19700
+    },
+    {
+      "epoch": 0.7366871509685259,
+      "grad_norm": 0.46953657269477844,
+      "learning_rate": 3.230687005783992e-06,
+      "loss": 0.2587,
+      "step": 19705
+    },
+    {
+      "epoch": 0.7368740799588757,
+      "grad_norm": 0.5593171119689941,
+      "learning_rate": 3.2263656868554092e-06,
+      "loss": 0.2911,
+      "step": 19710
+    },
+    {
+      "epoch": 0.7370610089492254,
+      "grad_norm": 0.4246465861797333,
+      "learning_rate": 3.2220467039592097e-06,
+      "loss": 0.2599,
+      "step": 19715
+    },
+    {
+      "epoch": 0.7372479379395752,
+      "grad_norm": 0.38245370984077454,
+      "learning_rate": 3.2177300585848916e-06,
+      "loss": 0.1948,
+      "step": 19720
+    },
+    {
+      "epoch": 0.737434866929925,
+      "grad_norm": 0.525850772857666,
+      "learning_rate": 3.213415752221136e-06,
+      "loss": 0.261,
+      "step": 19725
+    },
+    {
+      "epoch": 0.7376217959202748,
+      "grad_norm": 0.3949277997016907,
+      "learning_rate": 3.2091037863558316e-06,
+      "loss": 0.2761,
+      "step": 19730
+    },
+    {
+      "epoch": 0.7378087249106245,
+      "grad_norm": 0.47665491700172424,
+      "learning_rate": 3.2047941624760435e-06,
+      "loss": 0.2357,
+      "step": 19735
+    },
+    {
+      "epoch": 0.7379956539009743,
+      "grad_norm": 0.3770470917224884,
+      "learning_rate": 3.2004868820680423e-06,
+      "loss": 0.2551,
+      "step": 19740
+    },
+    {
+      "epoch": 0.7381825828913242,
+      "grad_norm": 0.24684104323387146,
+      "learning_rate": 3.196181946617287e-06,
+      "loss": 0.2372,
+      "step": 19745
+    },
+    {
+      "epoch": 0.738369511881674,
+      "grad_norm": 0.5357893705368042,
+      "learning_rate": 3.191879357608425e-06,
+      "loss": 0.3861,
+      "step": 19750
+    },
+    {
+      "epoch": 0.7385564408720238,
+      "grad_norm": 0.4339257776737213,
+      "learning_rate": 3.187579116525291e-06,
+      "loss": 0.2941,
+      "step": 19755
+    },
+    {
+      "epoch": 0.7387433698623735,
+      "grad_norm": 0.4764021933078766,
+      "learning_rate": 3.1832812248509235e-06,
+      "loss": 0.2143,
+      "step": 19760
+    },
+    {
+      "epoch": 0.7389302988527233,
+      "grad_norm": 0.433449923992157,
+      "learning_rate": 3.178985684067537e-06,
+      "loss": 0.2832,
+      "step": 19765
+    },
+    {
+      "epoch": 0.7391172278430731,
+      "grad_norm": 0.5205506086349487,
+      "learning_rate": 3.1746924956565385e-06,
+      "loss": 0.2622,
+      "step": 19770
+    },
+    {
+      "epoch": 0.7393041568334229,
+      "grad_norm": 0.37828561663627625,
+      "learning_rate": 3.1704016610985313e-06,
+      "loss": 0.2705,
+      "step": 19775
+    },
+    {
+      "epoch": 0.7394910858237727,
+      "grad_norm": 0.24176892638206482,
+      "learning_rate": 3.166113181873296e-06,
+      "loss": 0.3125,
+      "step": 19780
+    },
+    {
+      "epoch": 0.7396780148141224,
+      "grad_norm": 0.40736111998558044,
+      "learning_rate": 3.1618270594598076e-06,
+      "loss": 0.3194,
+      "step": 19785
+    },
+    {
+      "epoch": 0.7398649438044723,
+      "grad_norm": 0.47427597641944885,
+      "learning_rate": 3.1575432953362317e-06,
+      "loss": 0.3029,
+      "step": 19790
+    },
+    {
+      "epoch": 0.7400518727948221,
+      "grad_norm": 0.32264068722724915,
+      "learning_rate": 3.1532618909799095e-06,
+      "loss": 0.3093,
+      "step": 19795
+    },
+    {
+      "epoch": 0.7402388017851719,
+      "grad_norm": 0.5120447874069214,
+      "learning_rate": 3.14898284786738e-06,
+      "loss": 0.2977,
+      "step": 19800
+    },
+    {
+      "epoch": 0.7404257307755217,
+      "grad_norm": 0.3379383981227875,
+      "learning_rate": 3.1447061674743594e-06,
+      "loss": 0.2976,
+      "step": 19805
+    },
+    {
+      "epoch": 0.7406126597658714,
+      "grad_norm": 0.6211456060409546,
+      "learning_rate": 3.1404318512757525e-06,
+      "loss": 0.2624,
+      "step": 19810
+    },
+    {
+      "epoch": 0.7407995887562212,
+      "grad_norm": 0.306180864572525,
+      "learning_rate": 3.1361599007456456e-06,
+      "loss": 0.2349,
+      "step": 19815
+    },
+    {
+      "epoch": 0.740986517746571,
+      "grad_norm": 0.5241755247116089,
+      "learning_rate": 3.131890317357319e-06,
+      "loss": 0.3073,
+      "step": 19820
+    },
+    {
+      "epoch": 0.7411734467369208,
+      "grad_norm": 0.4787485599517822,
+      "learning_rate": 3.1276231025832217e-06,
+      "loss": 0.2572,
+      "step": 19825
+    },
+    {
+      "epoch": 0.7413603757272706,
+      "grad_norm": 0.27676764130592346,
+      "learning_rate": 3.123358257894997e-06,
+      "loss": 0.3357,
+      "step": 19830
+    },
+    {
+      "epoch": 0.7415473047176204,
+      "grad_norm": 0.4201495945453644,
+      "learning_rate": 3.119095784763473e-06,
+      "loss": 0.2738,
+      "step": 19835
+    },
+    {
+      "epoch": 0.7417342337079702,
+      "grad_norm": 0.27705347537994385,
+      "learning_rate": 3.114835684658647e-06,
+      "loss": 0.2443,
+      "step": 19840
+    },
+    {
+      "epoch": 0.74192116269832,
+      "grad_norm": 0.4122698903083801,
+      "learning_rate": 3.1105779590497108e-06,
+      "loss": 0.3632,
+      "step": 19845
+    },
+    {
+      "epoch": 0.7421080916886698,
+      "grad_norm": 0.39222484827041626,
+      "learning_rate": 3.1063226094050304e-06,
+      "loss": 0.2928,
+      "step": 19850
+    },
+    {
+      "epoch": 0.7422950206790195,
+      "grad_norm": 0.5306823253631592,
+      "learning_rate": 3.102069637192152e-06,
+      "loss": 0.2495,
+      "step": 19855
+    },
+    {
+      "epoch": 0.7424819496693693,
+      "grad_norm": 0.3978886306285858,
+      "learning_rate": 3.0978190438778022e-06,
+      "loss": 0.2175,
+      "step": 19860
+    },
+    {
+      "epoch": 0.7426688786597191,
+      "grad_norm": 0.34500226378440857,
+      "learning_rate": 3.0935708309278956e-06,
+      "loss": 0.2046,
+      "step": 19865
+    },
+    {
+      "epoch": 0.742855807650069,
+      "grad_norm": 0.3210207521915436,
+      "learning_rate": 3.0893249998075116e-06,
+      "loss": 0.2618,
+      "step": 19870
+    },
+    {
+      "epoch": 0.7430427366404188,
+      "grad_norm": 0.7616083025932312,
+      "learning_rate": 3.0850815519809184e-06,
+      "loss": 0.3427,
+      "step": 19875
+    },
+    {
+      "epoch": 0.7432296656307685,
+      "grad_norm": 0.44957032799720764,
+      "learning_rate": 3.080840488911565e-06,
+      "loss": 0.2465,
+      "step": 19880
+    },
+    {
+      "epoch": 0.7434165946211183,
+      "grad_norm": 0.3584950566291809,
+      "learning_rate": 3.0766018120620643e-06,
+      "loss": 0.2426,
+      "step": 19885
+    },
+    {
+      "epoch": 0.7436035236114681,
+      "grad_norm": 0.6310529112815857,
+      "learning_rate": 3.072365522894221e-06,
+      "loss": 0.2662,
+      "step": 19890
+    },
+    {
+      "epoch": 0.7437904526018179,
+      "grad_norm": 0.575434684753418,
+      "learning_rate": 3.068131622869007e-06,
+      "loss": 0.2728,
+      "step": 19895
+    },
+    {
+      "epoch": 0.7439773815921676,
+      "grad_norm": 0.051615819334983826,
+      "learning_rate": 3.063900113446574e-06,
+      "loss": 0.204,
+      "step": 19900
+    },
+    {
+      "epoch": 0.7441643105825174,
+      "grad_norm": 0.575394332408905,
+      "learning_rate": 3.0596709960862436e-06,
+      "loss": 0.2808,
+      "step": 19905
+    },
+    {
+      "epoch": 0.7443512395728673,
+      "grad_norm": 0.6706823110580444,
+      "learning_rate": 3.055444272246524e-06,
+      "loss": 0.3299,
+      "step": 19910
+    },
+    {
+      "epoch": 0.7445381685632171,
+      "grad_norm": 0.2847936451435089,
+      "learning_rate": 3.0512199433850855e-06,
+      "loss": 0.2401,
+      "step": 19915
+    },
+    {
+      "epoch": 0.7447250975535669,
+      "grad_norm": 0.4267846941947937,
+      "learning_rate": 3.0469980109587803e-06,
+      "loss": 0.3545,
+      "step": 19920
+    },
+    {
+      "epoch": 0.7449120265439166,
+      "grad_norm": 0.3833447992801666,
+      "learning_rate": 3.042778476423637e-06,
+      "loss": 0.3209,
+      "step": 19925
+    },
+    {
+      "epoch": 0.7450989555342664,
+      "grad_norm": 0.41514524817466736,
+      "learning_rate": 3.0385613412348423e-06,
+      "loss": 0.222,
+      "step": 19930
+    },
+    {
+      "epoch": 0.7452858845246162,
+      "grad_norm": 0.22775611281394958,
+      "learning_rate": 3.0343466068467752e-06,
+      "loss": 0.2635,
+      "step": 19935
+    },
+    {
+      "epoch": 0.745472813514966,
+      "grad_norm": 0.4837653636932373,
+      "learning_rate": 3.030134274712968e-06,
+      "loss": 0.294,
+      "step": 19940
+    },
+    {
+      "epoch": 0.7456597425053157,
+      "grad_norm": 0.41959109902381897,
+      "learning_rate": 3.0259243462861423e-06,
+      "loss": 0.3014,
+      "step": 19945
+    },
+    {
+      "epoch": 0.7458466714956656,
+      "grad_norm": 0.3355468511581421,
+      "learning_rate": 3.021716823018176e-06,
+      "loss": 0.2535,
+      "step": 19950
+    },
+    {
+      "epoch": 0.7460336004860154,
+      "grad_norm": 0.31884121894836426,
+      "learning_rate": 3.0175117063601235e-06,
+      "loss": 0.2149,
+      "step": 19955
+    },
+    {
+      "epoch": 0.7462205294763652,
+      "grad_norm": 1.0035449266433716,
+      "learning_rate": 3.0133089977622076e-06,
+      "loss": 0.3098,
+      "step": 19960
+    },
+    {
+      "epoch": 0.746407458466715,
+      "grad_norm": 0.34479397535324097,
+      "learning_rate": 3.009108698673825e-06,
+      "loss": 0.2726,
+      "step": 19965
+    },
+    {
+      "epoch": 0.7465943874570647,
+      "grad_norm": 0.619194746017456,
+      "learning_rate": 3.00491081054354e-06,
+      "loss": 0.3021,
+      "step": 19970
+    },
+    {
+      "epoch": 0.7467813164474145,
+      "grad_norm": 0.4144986867904663,
+      "learning_rate": 3.0007153348190786e-06,
+      "loss": 0.2219,
+      "step": 19975
+    },
+    {
+      "epoch": 0.7469682454377643,
+      "grad_norm": 0.580058753490448,
+      "learning_rate": 2.9965222729473474e-06,
+      "loss": 0.3113,
+      "step": 19980
+    },
+    {
+      "epoch": 0.7471551744281141,
+      "grad_norm": 0.1870926320552826,
+      "learning_rate": 2.992331626374405e-06,
+      "loss": 0.2202,
+      "step": 19985
+    },
+    {
+      "epoch": 0.747342103418464,
+      "grad_norm": 0.48588237166404724,
+      "learning_rate": 2.988143396545493e-06,
+      "loss": 0.2335,
+      "step": 19990
+    },
+    {
+      "epoch": 0.7475290324088137,
+      "grad_norm": 0.36052894592285156,
+      "learning_rate": 2.9839575849050094e-06,
+      "loss": 0.2429,
+      "step": 19995
+    },
+    {
+      "epoch": 0.7477159613991635,
+      "grad_norm": 0.42960643768310547,
+      "learning_rate": 2.9797741928965185e-06,
+      "loss": 0.2448,
+      "step": 20000
+    },
+    {
+      "epoch": 0.7479028903895133,
+      "grad_norm": 0.455585241317749,
+      "learning_rate": 2.9755932219627514e-06,
+      "loss": 0.262,
+      "step": 20005
+    },
+    {
+      "epoch": 0.7480898193798631,
+      "grad_norm": 0.22561267018318176,
+      "learning_rate": 2.9714146735456063e-06,
+      "loss": 0.289,
+      "step": 20010
+    },
+    {
+      "epoch": 0.7482767483702129,
+      "grad_norm": 0.3798793852329254,
+      "learning_rate": 2.96723854908615e-06,
+      "loss": 0.3006,
+      "step": 20015
+    },
+    {
+      "epoch": 0.7484636773605626,
+      "grad_norm": 0.45209431648254395,
+      "learning_rate": 2.9630648500245993e-06,
+      "loss": 0.3088,
+      "step": 20020
+    },
+    {
+      "epoch": 0.7486506063509124,
+      "grad_norm": 0.5453251004219055,
+      "learning_rate": 2.9588935778003526e-06,
+      "loss": 0.4175,
+      "step": 20025
+    },
+    {
+      "epoch": 0.7488375353412622,
+      "grad_norm": 0.5267879366874695,
+      "learning_rate": 2.9547247338519547e-06,
+      "loss": 0.2464,
+      "step": 20030
+    },
+    {
+      "epoch": 0.7490244643316121,
+      "grad_norm": 0.3326359987258911,
+      "learning_rate": 2.950558319617126e-06,
+      "loss": 0.2875,
+      "step": 20035
+    },
+    {
+      "epoch": 0.7492113933219618,
+      "grad_norm": 0.3526333272457123,
+      "learning_rate": 2.9463943365327406e-06,
+      "loss": 0.2775,
+      "step": 20040
+    },
+    {
+      "epoch": 0.7493983223123116,
+      "grad_norm": 0.2863275110721588,
+      "learning_rate": 2.9422327860348377e-06,
+      "loss": 0.2816,
+      "step": 20045
+    },
+    {
+      "epoch": 0.7495852513026614,
+      "grad_norm": 0.5883467793464661,
+      "learning_rate": 2.938073669558613e-06,
+      "loss": 0.2752,
+      "step": 20050
+    },
+    {
+      "epoch": 0.7497721802930112,
+      "grad_norm": 0.5096171498298645,
+      "learning_rate": 2.93391698853843e-06,
+      "loss": 0.348,
+      "step": 20055
+    },
+    {
+      "epoch": 0.749959109283361,
+      "grad_norm": 0.2166975438594818,
+      "learning_rate": 2.9297627444078115e-06,
+      "loss": 0.2923,
+      "step": 20060
+    },
+    {
+      "epoch": 0.7501460382737107,
+      "grad_norm": 0.3335493803024292,
+      "learning_rate": 2.9256109385994326e-06,
+      "loss": 0.2554,
+      "step": 20065
+    },
+    {
+      "epoch": 0.7503329672640605,
+      "grad_norm": 1.0838913917541504,
+      "learning_rate": 2.9214615725451354e-06,
+      "loss": 0.3143,
+      "step": 20070
+    },
+    {
+      "epoch": 0.7505198962544104,
+      "grad_norm": 0.5258083939552307,
+      "learning_rate": 2.917314647675914e-06,
+      "loss": 0.2903,
+      "step": 20075
+    },
+    {
+      "epoch": 0.7507068252447602,
+      "grad_norm": 0.4601660370826721,
+      "learning_rate": 2.913170165421929e-06,
+      "loss": 0.2453,
+      "step": 20080
+    },
+    {
+      "epoch": 0.75089375423511,
+      "grad_norm": 0.19206954538822174,
+      "learning_rate": 2.909028127212491e-06,
+      "loss": 0.2646,
+      "step": 20085
+    },
+    {
+      "epoch": 0.7510806832254597,
+      "grad_norm": 0.3423701524734497,
+      "learning_rate": 2.904888534476069e-06,
+      "loss": 0.2862,
+      "step": 20090
+    },
+    {
+      "epoch": 0.7512676122158095,
+      "grad_norm": 0.5520522594451904,
+      "learning_rate": 2.9007513886402884e-06,
+      "loss": 0.3379,
+      "step": 20095
+    },
+    {
+      "epoch": 0.7514545412061593,
+      "grad_norm": 0.2825580835342407,
+      "learning_rate": 2.896616691131934e-06,
+      "loss": 0.3071,
+      "step": 20100
+    },
+    {
+      "epoch": 0.7516414701965091,
+      "grad_norm": 0.4348238706588745,
+      "learning_rate": 2.892484443376948e-06,
+      "loss": 0.2731,
+      "step": 20105
+    },
+    {
+      "epoch": 0.7518283991868588,
+      "grad_norm": 0.4713153541088104,
+      "learning_rate": 2.8883546468004196e-06,
+      "loss": 0.2838,
+      "step": 20110
+    },
+    {
+      "epoch": 0.7520153281772087,
+      "grad_norm": 1.168811559677124,
+      "learning_rate": 2.884227302826601e-06,
+      "loss": 0.2472,
+      "step": 20115
+    },
+    {
+      "epoch": 0.7522022571675585,
+      "grad_norm": 0.5519205331802368,
+      "learning_rate": 2.8801024128788903e-06,
+      "loss": 0.3046,
+      "step": 20120
+    },
+    {
+      "epoch": 0.7523891861579083,
+      "grad_norm": 0.5526771545410156,
+      "learning_rate": 2.8759799783798503e-06,
+      "loss": 0.2885,
+      "step": 20125
+    },
+    {
+      "epoch": 0.7525761151482581,
+      "grad_norm": 0.4945368468761444,
+      "learning_rate": 2.871860000751182e-06,
+      "loss": 0.3089,
+      "step": 20130
+    },
+    {
+      "epoch": 0.7527630441386078,
+      "grad_norm": 0.5120470523834229,
+      "learning_rate": 2.8677424814137565e-06,
+      "loss": 0.3819,
+      "step": 20135
+    },
+    {
+      "epoch": 0.7529499731289576,
+      "grad_norm": 0.3834647536277771,
+      "learning_rate": 2.8636274217875846e-06,
+      "loss": 0.3076,
+      "step": 20140
+    },
+    {
+      "epoch": 0.7531369021193074,
+      "grad_norm": 0.7137815356254578,
+      "learning_rate": 2.859514823291829e-06,
+      "loss": 0.2814,
+      "step": 20145
+    },
+    {
+      "epoch": 0.7533238311096572,
+      "grad_norm": 0.5744923949241638,
+      "learning_rate": 2.8554046873448127e-06,
+      "loss": 0.3303,
+      "step": 20150
+    },
+    {
+      "epoch": 0.7535107601000071,
+      "grad_norm": 0.7045077085494995,
+      "learning_rate": 2.8512970153639976e-06,
+      "loss": 0.227,
+      "step": 20155
+    },
+    {
+      "epoch": 0.7536976890903568,
+      "grad_norm": 0.6118825078010559,
+      "learning_rate": 2.8471918087660087e-06,
+      "loss": 0.3046,
+      "step": 20160
+    },
+    {
+      "epoch": 0.7538846180807066,
+      "grad_norm": 0.6602465510368347,
+      "learning_rate": 2.843089068966609e-06,
+      "loss": 0.2856,
+      "step": 20165
+    },
+    {
+      "epoch": 0.7540715470710564,
+      "grad_norm": 0.5076795220375061,
+      "learning_rate": 2.8389887973807207e-06,
+      "loss": 0.2302,
+      "step": 20170
+    },
+    {
+      "epoch": 0.7542584760614062,
+      "grad_norm": 0.5142863988876343,
+      "learning_rate": 2.8348909954224037e-06,
+      "loss": 0.2519,
+      "step": 20175
+    },
+    {
+      "epoch": 0.754445405051756,
+      "grad_norm": 0.3825984001159668,
+      "learning_rate": 2.8307956645048795e-06,
+      "loss": 0.2828,
+      "step": 20180
+    },
+    {
+      "epoch": 0.7546323340421057,
+      "grad_norm": 0.5186363458633423,
+      "learning_rate": 2.8267028060405066e-06,
+      "loss": 0.3063,
+      "step": 20185
+    },
+    {
+      "epoch": 0.7548192630324555,
+      "grad_norm": 0.43224039673805237,
+      "learning_rate": 2.8226124214407912e-06,
+      "loss": 0.2481,
+      "step": 20190
+    },
+    {
+      "epoch": 0.7550061920228054,
+      "grad_norm": 0.3157874047756195,
+      "learning_rate": 2.8185245121163986e-06,
+      "loss": 0.3186,
+      "step": 20195
+    },
+    {
+      "epoch": 0.7551931210131552,
+      "grad_norm": 0.2855377197265625,
+      "learning_rate": 2.8144390794771215e-06,
+      "loss": 0.2652,
+      "step": 20200
+    },
+    {
+      "epoch": 0.7553800500035049,
+      "grad_norm": 0.29042690992355347,
+      "learning_rate": 2.810356124931918e-06,
+      "loss": 0.3023,
+      "step": 20205
+    },
+    {
+      "epoch": 0.7555669789938547,
+      "grad_norm": 0.38155996799468994,
+      "learning_rate": 2.806275649888873e-06,
+      "loss": 0.2603,
+      "step": 20210
+    },
+    {
+      "epoch": 0.7557539079842045,
+      "grad_norm": 0.3918231427669525,
+      "learning_rate": 2.8021976557552346e-06,
+      "loss": 0.2806,
+      "step": 20215
+    },
+    {
+      "epoch": 0.7559408369745543,
+      "grad_norm": 0.3024686574935913,
+      "learning_rate": 2.7981221439373774e-06,
+      "loss": 0.3187,
+      "step": 20220
+    },
+    {
+      "epoch": 0.756127765964904,
+      "grad_norm": 0.5795729160308838,
+      "learning_rate": 2.7940491158408367e-06,
+      "loss": 0.2417,
+      "step": 20225
+    },
+    {
+      "epoch": 0.7563146949552538,
+      "grad_norm": 1.9002882242202759,
+      "learning_rate": 2.7899785728702787e-06,
+      "loss": 0.2685,
+      "step": 20230
+    },
+    {
+      "epoch": 0.7565016239456037,
+      "grad_norm": 0.6263231039047241,
+      "learning_rate": 2.785910516429515e-06,
+      "loss": 0.3513,
+      "step": 20235
+    },
+    {
+      "epoch": 0.7566885529359535,
+      "grad_norm": 0.36941060423851013,
+      "learning_rate": 2.781844947921508e-06,
+      "loss": 0.4064,
+      "step": 20240
+    },
+    {
+      "epoch": 0.7568754819263033,
+      "grad_norm": 0.30503350496292114,
+      "learning_rate": 2.7777818687483483e-06,
+      "loss": 0.2724,
+      "step": 20245
+    },
+    {
+      "epoch": 0.757062410916653,
+      "grad_norm": 0.27014967799186707,
+      "learning_rate": 2.7737212803112824e-06,
+      "loss": 0.2964,
+      "step": 20250
+    },
+    {
+      "epoch": 0.7572493399070028,
+      "grad_norm": 0.432959645986557,
+      "learning_rate": 2.7696631840106847e-06,
+      "loss": 0.271,
+      "step": 20255
+    },
+    {
+      "epoch": 0.7574362688973526,
+      "grad_norm": 0.19095739722251892,
+      "learning_rate": 2.7656075812460835e-06,
+      "loss": 0.3006,
+      "step": 20260
+    },
+    {
+      "epoch": 0.7576231978877024,
+      "grad_norm": 0.3333872854709625,
+      "learning_rate": 2.7615544734161315e-06,
+      "loss": 0.2004,
+      "step": 20265
+    },
+    {
+      "epoch": 0.7578101268780522,
+      "grad_norm": 0.37564340233802795,
+      "learning_rate": 2.757503861918638e-06,
+      "loss": 0.3621,
+      "step": 20270
+    },
+    {
+      "epoch": 0.7579970558684019,
+      "grad_norm": 0.3397802412509918,
+      "learning_rate": 2.7534557481505385e-06,
+      "loss": 0.249,
+      "step": 20275
+    },
+    {
+      "epoch": 0.7581839848587518,
+      "grad_norm": 0.5983813405036926,
+      "learning_rate": 2.7494101335079094e-06,
+      "loss": 0.2332,
+      "step": 20280
+    },
+    {
+      "epoch": 0.7583709138491016,
+      "grad_norm": 0.38677194714546204,
+      "learning_rate": 2.7453670193859716e-06,
+      "loss": 0.3318,
+      "step": 20285
+    },
+    {
+      "epoch": 0.7585578428394514,
+      "grad_norm": 0.45902055501937866,
+      "learning_rate": 2.7413264071790747e-06,
+      "loss": 0.2212,
+      "step": 20290
+    },
+    {
+      "epoch": 0.7587447718298012,
+      "grad_norm": 0.6074469685554504,
+      "learning_rate": 2.737288298280715e-06,
+      "loss": 0.2984,
+      "step": 20295
+    },
+    {
+      "epoch": 0.7589317008201509,
+      "grad_norm": 0.4030331075191498,
+      "learning_rate": 2.7332526940835156e-06,
+      "loss": 0.3323,
+      "step": 20300
+    },
+    {
+      "epoch": 0.7591186298105007,
+      "grad_norm": 0.3118947744369507,
+      "learning_rate": 2.729219595979247e-06,
+      "loss": 0.263,
+      "step": 20305
+    },
+    {
+      "epoch": 0.7593055588008505,
+      "grad_norm": 0.2766875624656677,
+      "learning_rate": 2.7251890053588015e-06,
+      "loss": 0.2779,
+      "step": 20310
+    },
+    {
+      "epoch": 0.7594924877912003,
+      "grad_norm": 0.4155539572238922,
+      "learning_rate": 2.7211609236122216e-06,
+      "loss": 0.2856,
+      "step": 20315
+    },
+    {
+      "epoch": 0.7596794167815502,
+      "grad_norm": 0.37500113248825073,
+      "learning_rate": 2.717135352128671e-06,
+      "loss": 0.2353,
+      "step": 20320
+    },
+    {
+      "epoch": 0.7598663457718999,
+      "grad_norm": 0.60191810131073,
+      "learning_rate": 2.7131122922964603e-06,
+      "loss": 0.3352,
+      "step": 20325
+    },
+    {
+      "epoch": 0.7600532747622497,
+      "grad_norm": 0.2800582945346832,
+      "learning_rate": 2.709091745503024e-06,
+      "loss": 0.3133,
+      "step": 20330
+    },
+    {
+      "epoch": 0.7602402037525995,
+      "grad_norm": 0.4130517840385437,
+      "learning_rate": 2.7050737131349315e-06,
+      "loss": 0.2603,
+      "step": 20335
+    },
+    {
+      "epoch": 0.7604271327429493,
+      "grad_norm": 0.5701847076416016,
+      "learning_rate": 2.7010581965778914e-06,
+      "loss": 0.2705,
+      "step": 20340
+    },
+    {
+      "epoch": 0.760614061733299,
+      "grad_norm": 0.47026559710502625,
+      "learning_rate": 2.6970451972167355e-06,
+      "loss": 0.2877,
+      "step": 20345
+    },
+    {
+      "epoch": 0.7608009907236488,
+      "grad_norm": 0.5374239683151245,
+      "learning_rate": 2.6930347164354376e-06,
+      "loss": 0.2791,
+      "step": 20350
+    },
+    {
+      "epoch": 0.7609879197139986,
+      "grad_norm": 0.29200440645217896,
+      "learning_rate": 2.6890267556170925e-06,
+      "loss": 0.2677,
+      "step": 20355
+    },
+    {
+      "epoch": 0.7611748487043485,
+      "grad_norm": 0.3506094515323639,
+      "learning_rate": 2.6850213161439363e-06,
+      "loss": 0.2583,
+      "step": 20360
+    },
+    {
+      "epoch": 0.7613617776946983,
+      "grad_norm": 0.652692973613739,
+      "learning_rate": 2.6810183993973247e-06,
+      "loss": 0.259,
+      "step": 20365
+    },
+    {
+      "epoch": 0.761548706685048,
+      "grad_norm": 0.6381142139434814,
+      "learning_rate": 2.6770180067577547e-06,
+      "loss": 0.3052,
+      "step": 20370
+    },
+    {
+      "epoch": 0.7617356356753978,
+      "grad_norm": 0.3937336206436157,
+      "learning_rate": 2.6730201396048437e-06,
+      "loss": 0.2877,
+      "step": 20375
+    },
+    {
+      "epoch": 0.7619225646657476,
+      "grad_norm": 0.44311031699180603,
+      "learning_rate": 2.6690247993173393e-06,
+      "loss": 0.374,
+      "step": 20380
+    },
+    {
+      "epoch": 0.7621094936560974,
+      "grad_norm": 2.144091844558716,
+      "learning_rate": 2.6650319872731258e-06,
+      "loss": 0.4321,
+      "step": 20385
+    },
+    {
+      "epoch": 0.7622964226464471,
+      "grad_norm": 0.479422390460968,
+      "learning_rate": 2.661041704849203e-06,
+      "loss": 0.274,
+      "step": 20390
+    },
+    {
+      "epoch": 0.7624833516367969,
+      "grad_norm": 0.392315536737442,
+      "learning_rate": 2.657053953421712e-06,
+      "loss": 0.2914,
+      "step": 20395
+    },
+    {
+      "epoch": 0.7626702806271468,
+      "grad_norm": 0.34140515327453613,
+      "learning_rate": 2.6530687343659067e-06,
+      "loss": 0.2824,
+      "step": 20400
+    },
+    {
+      "epoch": 0.7628572096174966,
+      "grad_norm": 0.2744537889957428,
+      "learning_rate": 2.649086049056182e-06,
+      "loss": 0.2552,
+      "step": 20405
+    },
+    {
+      "epoch": 0.7630441386078464,
+      "grad_norm": 0.23247292637825012,
+      "learning_rate": 2.645105898866046e-06,
+      "loss": 0.2724,
+      "step": 20410
+    },
+    {
+      "epoch": 0.7632310675981961,
+      "grad_norm": 0.2504008412361145,
+      "learning_rate": 2.641128285168144e-06,
+      "loss": 0.2436,
+      "step": 20415
+    },
+    {
+      "epoch": 0.7634179965885459,
+      "grad_norm": 0.3857523500919342,
+      "learning_rate": 2.637153209334239e-06,
+      "loss": 0.294,
+      "step": 20420
+    },
+    {
+      "epoch": 0.7636049255788957,
+      "grad_norm": 0.42099565267562866,
+      "learning_rate": 2.633180672735215e-06,
+      "loss": 0.3373,
+      "step": 20425
+    },
+    {
+      "epoch": 0.7637918545692455,
+      "grad_norm": 0.2764213979244232,
+      "learning_rate": 2.6292106767410953e-06,
+      "loss": 0.2514,
+      "step": 20430
+    },
+    {
+      "epoch": 0.7639787835595953,
+      "grad_norm": 0.8510448932647705,
+      "learning_rate": 2.62524322272101e-06,
+      "loss": 0.2924,
+      "step": 20435
+    },
+    {
+      "epoch": 0.7641657125499451,
+      "grad_norm": 0.271930456161499,
+      "learning_rate": 2.621278312043226e-06,
+      "loss": 0.2673,
+      "step": 20440
+    },
+    {
+      "epoch": 0.7643526415402949,
+      "grad_norm": 0.4646458327770233,
+      "learning_rate": 2.617315946075123e-06,
+      "loss": 0.2376,
+      "step": 20445
+    },
+    {
+      "epoch": 0.7645395705306447,
+      "grad_norm": 0.5632447600364685,
+      "learning_rate": 2.613356126183212e-06,
+      "loss": 0.2435,
+      "step": 20450
+    },
+    {
+      "epoch": 0.7647264995209945,
+      "grad_norm": 0.29168501496315,
+      "learning_rate": 2.6093988537331163e-06,
+      "loss": 0.3008,
+      "step": 20455
+    },
+    {
+      "epoch": 0.7649134285113443,
+      "grad_norm": 0.24056875705718994,
+      "learning_rate": 2.6054441300895905e-06,
+      "loss": 0.2312,
+      "step": 20460
+    },
+    {
+      "epoch": 0.765100357501694,
+      "grad_norm": 0.3789188265800476,
+      "learning_rate": 2.601491956616504e-06,
+      "loss": 0.2843,
+      "step": 20465
+    },
+    {
+      "epoch": 0.7652872864920438,
+      "grad_norm": 0.35783255100250244,
+      "learning_rate": 2.597542334676846e-06,
+      "loss": 0.2711,
+      "step": 20470
+    },
+    {
+      "epoch": 0.7654742154823936,
+      "grad_norm": 0.6306200623512268,
+      "learning_rate": 2.59359526563273e-06,
+      "loss": 0.3031,
+      "step": 20475
+    },
+    {
+      "epoch": 0.7656611444727435,
+      "grad_norm": 0.30697113275527954,
+      "learning_rate": 2.589650750845385e-06,
+      "loss": 0.2589,
+      "step": 20480
+    },
+    {
+      "epoch": 0.7658480734630932,
+      "grad_norm": 0.41806843876838684,
+      "learning_rate": 2.5857087916751656e-06,
+      "loss": 0.2346,
+      "step": 20485
+    },
+    {
+      "epoch": 0.766035002453443,
+      "grad_norm": 0.4137856960296631,
+      "learning_rate": 2.5817693894815342e-06,
+      "loss": 0.2488,
+      "step": 20490
+    },
+    {
+      "epoch": 0.7662219314437928,
+      "grad_norm": 1.0388176441192627,
+      "learning_rate": 2.5778325456230845e-06,
+      "loss": 0.2959,
+      "step": 20495
+    },
+    {
+      "epoch": 0.7664088604341426,
+      "grad_norm": 0.3841733932495117,
+      "learning_rate": 2.5738982614575147e-06,
+      "loss": 0.3137,
+      "step": 20500
+    },
+    {
+      "epoch": 0.7665957894244924,
+      "grad_norm": 0.3577060103416443,
+      "learning_rate": 2.569966538341654e-06,
+      "loss": 0.3331,
+      "step": 20505
+    },
+    {
+      "epoch": 0.7667827184148421,
+      "grad_norm": 0.23667654395103455,
+      "learning_rate": 2.5660373776314318e-06,
+      "loss": 0.3054,
+      "step": 20510
+    },
+    {
+      "epoch": 0.7669696474051919,
+      "grad_norm": 0.5192760825157166,
+      "learning_rate": 2.5621107806819125e-06,
+      "loss": 0.2532,
+      "step": 20515
+    },
+    {
+      "epoch": 0.7671565763955417,
+      "grad_norm": 0.2387147843837738,
+      "learning_rate": 2.558186748847262e-06,
+      "loss": 0.2256,
+      "step": 20520
+    },
+    {
+      "epoch": 0.7673435053858916,
+      "grad_norm": 0.5738237500190735,
+      "learning_rate": 2.5542652834807634e-06,
+      "loss": 0.2913,
+      "step": 20525
+    },
+    {
+      "epoch": 0.7675304343762414,
+      "grad_norm": 0.5508410334587097,
+      "learning_rate": 2.5503463859348245e-06,
+      "loss": 0.2927,
+      "step": 20530
+    },
+    {
+      "epoch": 0.7677173633665911,
+      "grad_norm": 0.5864548087120056,
+      "learning_rate": 2.5464300575609547e-06,
+      "loss": 0.355,
+      "step": 20535
+    },
+    {
+      "epoch": 0.7679042923569409,
+      "grad_norm": 0.5955525636672974,
+      "learning_rate": 2.5425162997097896e-06,
+      "loss": 0.263,
+      "step": 20540
+    },
+    {
+      "epoch": 0.7680912213472907,
+      "grad_norm": 0.3412562906742096,
+      "learning_rate": 2.538605113731065e-06,
+      "loss": 0.3519,
+      "step": 20545
+    },
+    {
+      "epoch": 0.7682781503376405,
+      "grad_norm": 0.44794994592666626,
+      "learning_rate": 2.5346965009736445e-06,
+      "loss": 0.3314,
+      "step": 20550
+    },
+    {
+      "epoch": 0.7684650793279902,
+      "grad_norm": 0.5643365979194641,
+      "learning_rate": 2.5307904627854895e-06,
+      "loss": 0.2622,
+      "step": 20555
+    },
+    {
+      "epoch": 0.76865200831834,
+      "grad_norm": 0.3417246639728546,
+      "learning_rate": 2.526887000513687e-06,
+      "loss": 0.3421,
+      "step": 20560
+    },
+    {
+      "epoch": 0.7688389373086899,
+      "grad_norm": 0.1929822713136673,
+      "learning_rate": 2.5229861155044254e-06,
+      "loss": 0.267,
+      "step": 20565
+    },
+    {
+      "epoch": 0.7690258662990397,
+      "grad_norm": 0.2237076312303543,
+      "learning_rate": 2.5190878091030067e-06,
+      "loss": 0.2641,
+      "step": 20570
+    },
+    {
+      "epoch": 0.7692127952893895,
+      "grad_norm": 0.43035122752189636,
+      "learning_rate": 2.5151920826538514e-06,
+      "loss": 0.2557,
+      "step": 20575
+    },
+    {
+      "epoch": 0.7693997242797392,
+      "grad_norm": 0.3139476180076599,
+      "learning_rate": 2.511298937500476e-06,
+      "loss": 0.2566,
+      "step": 20580
+    },
+    {
+      "epoch": 0.769586653270089,
+      "grad_norm": 0.2134442925453186,
+      "learning_rate": 2.5074083749855216e-06,
+      "loss": 0.2199,
+      "step": 20585
+    },
+    {
+      "epoch": 0.7697735822604388,
+      "grad_norm": 0.6720663905143738,
+      "learning_rate": 2.503520396450725e-06,
+      "loss": 0.2664,
+      "step": 20590
+    },
+    {
+      "epoch": 0.7699605112507886,
+      "grad_norm": 0.7962212562561035,
+      "learning_rate": 2.4996350032369467e-06,
+      "loss": 0.3161,
+      "step": 20595
+    },
+    {
+      "epoch": 0.7701474402411383,
+      "grad_norm": 0.24116773903369904,
+      "learning_rate": 2.4957521966841393e-06,
+      "loss": 0.3047,
+      "step": 20600
+    },
+    {
+      "epoch": 0.7703343692314882,
+      "grad_norm": 0.21904657781124115,
+      "learning_rate": 2.4918719781313782e-06,
+      "loss": 0.3049,
+      "step": 20605
+    },
+    {
+      "epoch": 0.770521298221838,
+      "grad_norm": 0.40606293082237244,
+      "learning_rate": 2.487994348916837e-06,
+      "loss": 0.3494,
+      "step": 20610
+    },
+    {
+      "epoch": 0.7707082272121878,
+      "grad_norm": 0.586861789226532,
+      "learning_rate": 2.484119310377796e-06,
+      "loss": 0.2998,
+      "step": 20615
+    },
+    {
+      "epoch": 0.7708951562025376,
+      "grad_norm": 0.5949821472167969,
+      "learning_rate": 2.4802468638506505e-06,
+      "loss": 0.3086,
+      "step": 20620
+    },
+    {
+      "epoch": 0.7710820851928873,
+      "grad_norm": 0.3286888599395752,
+      "learning_rate": 2.4763770106708907e-06,
+      "loss": 0.2415,
+      "step": 20625
+    },
+    {
+      "epoch": 0.7712690141832371,
+      "grad_norm": 0.1509125977754593,
+      "learning_rate": 2.4725097521731232e-06,
+      "loss": 0.2471,
+      "step": 20630
+    },
+    {
+      "epoch": 0.7714559431735869,
+      "grad_norm": 0.5915478467941284,
+      "learning_rate": 2.4686450896910497e-06,
+      "loss": 0.2614,
+      "step": 20635
+    },
+    {
+      "epoch": 0.7716428721639367,
+      "grad_norm": 0.725960910320282,
+      "learning_rate": 2.4647830245574865e-06,
+      "loss": 0.3098,
+      "step": 20640
+    },
+    {
+      "epoch": 0.7718298011542866,
+      "grad_norm": 0.4178033769130707,
+      "learning_rate": 2.4609235581043457e-06,
+      "loss": 0.2995,
+      "step": 20645
+    },
+    {
+      "epoch": 0.7720167301446363,
+      "grad_norm": 0.7236891984939575,
+      "learning_rate": 2.4570666916626484e-06,
+      "loss": 0.2878,
+      "step": 20650
+    },
+    {
+      "epoch": 0.7722036591349861,
+      "grad_norm": 0.33673229813575745,
+      "learning_rate": 2.4532124265625155e-06,
+      "loss": 0.3246,
+      "step": 20655
+    },
+    {
+      "epoch": 0.7723905881253359,
+      "grad_norm": 0.4866393804550171,
+      "learning_rate": 2.4493607641331762e-06,
+      "loss": 0.2776,
+      "step": 20660
+    },
+    {
+      "epoch": 0.7725775171156857,
+      "grad_norm": 0.7614867091178894,
+      "learning_rate": 2.4455117057029566e-06,
+      "loss": 0.3473,
+      "step": 20665
+    },
+    {
+      "epoch": 0.7727644461060355,
+      "grad_norm": 0.46578964591026306,
+      "learning_rate": 2.441665252599282e-06,
+      "loss": 0.2915,
+      "step": 20670
+    },
+    {
+      "epoch": 0.7729513750963852,
+      "grad_norm": 0.5103464126586914,
+      "learning_rate": 2.4378214061486925e-06,
+      "loss": 0.246,
+      "step": 20675
+    },
+    {
+      "epoch": 0.773138304086735,
+      "grad_norm": 0.5652568936347961,
+      "learning_rate": 2.433980167676813e-06,
+      "loss": 0.3631,
+      "step": 20680
+    },
+    {
+      "epoch": 0.7733252330770849,
+      "grad_norm": 0.46092647314071655,
+      "learning_rate": 2.4301415385083828e-06,
+      "loss": 0.3188,
+      "step": 20685
+    },
+    {
+      "epoch": 0.7735121620674347,
+      "grad_norm": 0.5116592645645142,
+      "learning_rate": 2.426305519967228e-06,
+      "loss": 0.2404,
+      "step": 20690
+    },
+    {
+      "epoch": 0.7736990910577844,
+      "grad_norm": 0.6798794269561768,
+      "learning_rate": 2.4224721133762864e-06,
+      "loss": 0.2467,
+      "step": 20695
+    },
+    {
+      "epoch": 0.7738860200481342,
+      "grad_norm": 0.4135216772556305,
+      "learning_rate": 2.418641320057592e-06,
+      "loss": 0.2992,
+      "step": 20700
+    },
+    {
+      "epoch": 0.774072949038484,
+      "grad_norm": 0.5038582682609558,
+      "learning_rate": 2.414813141332274e-06,
+      "loss": 0.2296,
+      "step": 20705
+    },
+    {
+      "epoch": 0.7742598780288338,
+      "grad_norm": 0.35031184554100037,
+      "learning_rate": 2.4109875785205593e-06,
+      "loss": 0.25,
+      "step": 20710
+    },
+    {
+      "epoch": 0.7744468070191836,
+      "grad_norm": 0.21241429448127747,
+      "learning_rate": 2.407164632941773e-06,
+      "loss": 0.2752,
+      "step": 20715
+    },
+    {
+      "epoch": 0.7746337360095333,
+      "grad_norm": 0.2852620482444763,
+      "learning_rate": 2.403344305914346e-06,
+      "loss": 0.2454,
+      "step": 20720
+    },
+    {
+      "epoch": 0.7748206649998832,
+      "grad_norm": 0.20307046175003052,
+      "learning_rate": 2.3995265987557925e-06,
+      "loss": 0.3172,
+      "step": 20725
+    },
+    {
+      "epoch": 0.775007593990233,
+      "grad_norm": 0.38747474551200867,
+      "learning_rate": 2.395711512782738e-06,
+      "loss": 0.3205,
+      "step": 20730
+    },
+    {
+      "epoch": 0.7751945229805828,
+      "grad_norm": 0.2965136468410492,
+      "learning_rate": 2.3918990493108884e-06,
+      "loss": 0.2596,
+      "step": 20735
+    },
+    {
+      "epoch": 0.7753814519709326,
+      "grad_norm": 0.3963155150413513,
+      "learning_rate": 2.3880892096550578e-06,
+      "loss": 0.2905,
+      "step": 20740
+    },
+    {
+      "epoch": 0.7755683809612823,
+      "grad_norm": 0.31911271810531616,
+      "learning_rate": 2.384281995129153e-06,
+      "loss": 0.3044,
+      "step": 20745
+    },
+    {
+      "epoch": 0.7757553099516321,
+      "grad_norm": 0.8403980731964111,
+      "learning_rate": 2.380477407046169e-06,
+      "loss": 0.3053,
+      "step": 20750
+    },
+    {
+      "epoch": 0.7759422389419819,
+      "grad_norm": 0.35401323437690735,
+      "learning_rate": 2.3766754467182006e-06,
+      "loss": 0.2962,
+      "step": 20755
+    },
+    {
+      "epoch": 0.7761291679323317,
+      "grad_norm": 0.5200223922729492,
+      "learning_rate": 2.3728761154564326e-06,
+      "loss": 0.3908,
+      "step": 20760
+    },
+    {
+      "epoch": 0.7763160969226814,
+      "grad_norm": 0.4293026626110077,
+      "learning_rate": 2.3690794145711505e-06,
+      "loss": 0.3042,
+      "step": 20765
+    },
+    {
+      "epoch": 0.7765030259130313,
+      "grad_norm": 0.47955235838890076,
+      "learning_rate": 2.365285345371722e-06,
+      "loss": 0.2323,
+      "step": 20770
+    },
+    {
+      "epoch": 0.7766899549033811,
+      "grad_norm": 0.29286670684814453,
+      "learning_rate": 2.3614939091666177e-06,
+      "loss": 0.3391,
+      "step": 20775
+    },
+    {
+      "epoch": 0.7768768838937309,
+      "grad_norm": 0.21747294068336487,
+      "learning_rate": 2.3577051072633907e-06,
+      "loss": 0.3265,
+      "step": 20780
+    },
+    {
+      "epoch": 0.7770638128840807,
+      "grad_norm": 0.9644585251808167,
+      "learning_rate": 2.3539189409686937e-06,
+      "loss": 0.3585,
+      "step": 20785
+    },
+    {
+      "epoch": 0.7772507418744304,
+      "grad_norm": 0.21008965373039246,
+      "learning_rate": 2.350135411588267e-06,
+      "loss": 0.2923,
+      "step": 20790
+    },
+    {
+      "epoch": 0.7774376708647802,
+      "grad_norm": 0.4879824221134186,
+      "learning_rate": 2.346354520426942e-06,
+      "loss": 0.329,
+      "step": 20795
+    },
+    {
+      "epoch": 0.77762459985513,
+      "grad_norm": 0.26359811425209045,
+      "learning_rate": 2.3425762687886378e-06,
+      "loss": 0.2053,
+      "step": 20800
+    },
+    {
+      "epoch": 0.7778115288454798,
+      "grad_norm": 0.15240828692913055,
+      "learning_rate": 2.3388006579763623e-06,
+      "loss": 0.2589,
+      "step": 20805
+    },
+    {
+      "epoch": 0.7779984578358297,
+      "grad_norm": 0.27661359310150146,
+      "learning_rate": 2.3350276892922218e-06,
+      "loss": 0.2356,
+      "step": 20810
+    },
+    {
+      "epoch": 0.7781853868261794,
+      "grad_norm": 0.5237915515899658,
+      "learning_rate": 2.3312573640373994e-06,
+      "loss": 0.3032,
+      "step": 20815
+    },
+    {
+      "epoch": 0.7783723158165292,
+      "grad_norm": 0.33490797877311707,
+      "learning_rate": 2.3274896835121772e-06,
+      "loss": 0.2319,
+      "step": 20820
+    },
+    {
+      "epoch": 0.778559244806879,
+      "grad_norm": 0.5029880404472351,
+      "learning_rate": 2.323724649015916e-06,
+      "loss": 0.2439,
+      "step": 20825
+    },
+    {
+      "epoch": 0.7787461737972288,
+      "grad_norm": 0.47843530774116516,
+      "learning_rate": 2.31996226184707e-06,
+      "loss": 0.3145,
+      "step": 20830
+    },
+    {
+      "epoch": 0.7789331027875785,
+      "grad_norm": 0.5605517625808716,
+      "learning_rate": 2.3162025233031814e-06,
+      "loss": 0.3438,
+      "step": 20835
+    },
+    {
+      "epoch": 0.7791200317779283,
+      "grad_norm": 0.47056886553764343,
+      "learning_rate": 2.3124454346808713e-06,
+      "loss": 0.2864,
+      "step": 20840
+    },
+    {
+      "epoch": 0.7793069607682781,
+      "grad_norm": 0.5489882230758667,
+      "learning_rate": 2.3086909972758577e-06,
+      "loss": 0.272,
+      "step": 20845
+    },
+    {
+      "epoch": 0.779493889758628,
+      "grad_norm": 0.2519620656967163,
+      "learning_rate": 2.304939212382934e-06,
+      "loss": 0.2637,
+      "step": 20850
+    },
+    {
+      "epoch": 0.7796808187489778,
+      "grad_norm": 0.16588373482227325,
+      "learning_rate": 2.3011900812959855e-06,
+      "loss": 0.248,
+      "step": 20855
+    },
+    {
+      "epoch": 0.7798677477393275,
+      "grad_norm": 0.48377254605293274,
+      "learning_rate": 2.2974436053079764e-06,
+      "loss": 0.2892,
+      "step": 20860
+    },
+    {
+      "epoch": 0.7800546767296773,
+      "grad_norm": 0.32265156507492065,
+      "learning_rate": 2.2936997857109644e-06,
+      "loss": 0.2799,
+      "step": 20865
+    },
+    {
+      "epoch": 0.7802416057200271,
+      "grad_norm": 0.25994792580604553,
+      "learning_rate": 2.2899586237960793e-06,
+      "loss": 0.2736,
+      "step": 20870
+    },
+    {
+      "epoch": 0.7804285347103769,
+      "grad_norm": 0.4288146495819092,
+      "learning_rate": 2.286220120853545e-06,
+      "loss": 0.2761,
+      "step": 20875
+    },
+    {
+      "epoch": 0.7806154637007267,
+      "grad_norm": 0.2761852741241455,
+      "learning_rate": 2.2824842781726665e-06,
+      "loss": 0.2326,
+      "step": 20880
+    },
+    {
+      "epoch": 0.7808023926910764,
+      "grad_norm": 0.561490535736084,
+      "learning_rate": 2.2787510970418215e-06,
+      "loss": 0.2932,
+      "step": 20885
+    },
+    {
+      "epoch": 0.7809893216814263,
+      "grad_norm": 0.28700560331344604,
+      "learning_rate": 2.2750205787484846e-06,
+      "loss": 0.1987,
+      "step": 20890
+    },
+    {
+      "epoch": 0.7811762506717761,
+      "grad_norm": 0.2016209363937378,
+      "learning_rate": 2.271292724579203e-06,
+      "loss": 0.2507,
+      "step": 20895
+    },
+    {
+      "epoch": 0.7813631796621259,
+      "grad_norm": 0.3742198944091797,
+      "learning_rate": 2.2675675358196037e-06,
+      "loss": 0.2465,
+      "step": 20900
+    },
+    {
+      "epoch": 0.7815501086524757,
+      "grad_norm": 0.301685094833374,
+      "learning_rate": 2.2638450137543967e-06,
+      "loss": 0.281,
+      "step": 20905
+    },
+    {
+      "epoch": 0.7817370376428254,
+      "grad_norm": 0.6065481305122375,
+      "learning_rate": 2.2601251596673778e-06,
+      "loss": 0.249,
+      "step": 20910
+    },
+    {
+      "epoch": 0.7819239666331752,
+      "grad_norm": 0.41621100902557373,
+      "learning_rate": 2.2564079748414138e-06,
+      "loss": 0.254,
+      "step": 20915
+    },
+    {
+      "epoch": 0.782110895623525,
+      "grad_norm": 0.33886462450027466,
+      "learning_rate": 2.252693460558456e-06,
+      "loss": 0.2795,
+      "step": 20920
+    },
+    {
+      "epoch": 0.7822978246138748,
+      "grad_norm": 0.6175533533096313,
+      "learning_rate": 2.2489816180995395e-06,
+      "loss": 0.3793,
+      "step": 20925
+    },
+    {
+      "epoch": 0.7824847536042246,
+      "grad_norm": 0.7338931560516357,
+      "learning_rate": 2.245272448744765e-06,
+      "loss": 0.289,
+      "step": 20930
+    },
+    {
+      "epoch": 0.7826716825945744,
+      "grad_norm": 0.292777419090271,
+      "learning_rate": 2.241565953773325e-06,
+      "loss": 0.2248,
+      "step": 20935
+    },
+    {
+      "epoch": 0.7828586115849242,
+      "grad_norm": 0.5425524711608887,
+      "learning_rate": 2.237862134463479e-06,
+      "loss": 0.2553,
+      "step": 20940
+    },
+    {
+      "epoch": 0.783045540575274,
+      "grad_norm": 0.29631131887435913,
+      "learning_rate": 2.2341609920925698e-06,
+      "loss": 0.2826,
+      "step": 20945
+    },
+    {
+      "epoch": 0.7832324695656238,
+      "grad_norm": 0.313694566488266,
+      "learning_rate": 2.230462527937013e-06,
+      "loss": 0.3969,
+      "step": 20950
+    },
+    {
+      "epoch": 0.7834193985559735,
+      "grad_norm": 0.7053239941596985,
+      "learning_rate": 2.2267667432723073e-06,
+      "loss": 0.3284,
+      "step": 20955
+    },
+    {
+      "epoch": 0.7836063275463233,
+      "grad_norm": 0.22134394943714142,
+      "learning_rate": 2.2230736393730178e-06,
+      "loss": 0.276,
+      "step": 20960
+    },
+    {
+      "epoch": 0.7837932565366731,
+      "grad_norm": 0.2921614348888397,
+      "learning_rate": 2.2193832175127928e-06,
+      "loss": 0.247,
+      "step": 20965
+    },
+    {
+      "epoch": 0.783980185527023,
+      "grad_norm": 0.4508916735649109,
+      "learning_rate": 2.215695478964357e-06,
+      "loss": 0.2421,
+      "step": 20970
+    },
+    {
+      "epoch": 0.7841671145173728,
+      "grad_norm": 0.4369545578956604,
+      "learning_rate": 2.212010424999498e-06,
+      "loss": 0.2226,
+      "step": 20975
+    },
+    {
+      "epoch": 0.7843540435077225,
+      "grad_norm": 0.21277235448360443,
+      "learning_rate": 2.2083280568890918e-06,
+      "loss": 0.292,
+      "step": 20980
+    },
+    {
+      "epoch": 0.7845409724980723,
+      "grad_norm": 0.4735264778137207,
+      "learning_rate": 2.20464837590308e-06,
+      "loss": 0.2305,
+      "step": 20985
+    },
+    {
+      "epoch": 0.7847279014884221,
+      "grad_norm": 0.5336333513259888,
+      "learning_rate": 2.2009713833104785e-06,
+      "loss": 0.2775,
+      "step": 20990
+    },
+    {
+      "epoch": 0.7849148304787719,
+      "grad_norm": 0.4231540262699127,
+      "learning_rate": 2.1972970803793726e-06,
+      "loss": 0.2616,
+      "step": 20995
+    },
+    {
+      "epoch": 0.7851017594691216,
+      "grad_norm": 0.5644051432609558,
+      "learning_rate": 2.193625468376931e-06,
+      "loss": 0.2852,
+      "step": 21000
+    },
+    {
+      "epoch": 0.7852886884594714,
+      "grad_norm": 0.5196647047996521,
+      "learning_rate": 2.189956548569382e-06,
+      "loss": 0.3341,
+      "step": 21005
+    },
+    {
+      "epoch": 0.7854756174498212,
+      "grad_norm": 0.4049712121486664,
+      "learning_rate": 2.186290322222033e-06,
+      "loss": 0.4098,
+      "step": 21010
+    },
+    {
+      "epoch": 0.7856625464401711,
+      "grad_norm": 0.4484502375125885,
+      "learning_rate": 2.182626790599265e-06,
+      "loss": 0.2502,
+      "step": 21015
+    },
+    {
+      "epoch": 0.7858494754305209,
+      "grad_norm": 0.1917799860239029,
+      "learning_rate": 2.1789659549645158e-06,
+      "loss": 0.2588,
+      "step": 21020
+    },
+    {
+      "epoch": 0.7860364044208706,
+      "grad_norm": 0.3756994307041168,
+      "learning_rate": 2.175307816580312e-06,
+      "loss": 0.2758,
+      "step": 21025
+    },
+    {
+      "epoch": 0.7862233334112204,
+      "grad_norm": 0.4414009153842926,
+      "learning_rate": 2.171652376708233e-06,
+      "loss": 0.2278,
+      "step": 21030
+    },
+    {
+      "epoch": 0.7864102624015702,
+      "grad_norm": 0.34547463059425354,
+      "learning_rate": 2.1679996366089428e-06,
+      "loss": 0.3164,
+      "step": 21035
+    },
+    {
+      "epoch": 0.78659719139192,
+      "grad_norm": 0.3485727608203888,
+      "learning_rate": 2.1643495975421612e-06,
+      "loss": 0.3466,
+      "step": 21040
+    },
+    {
+      "epoch": 0.7867841203822697,
+      "grad_norm": 0.772977888584137,
+      "learning_rate": 2.160702260766684e-06,
+      "loss": 0.2409,
+      "step": 21045
+    },
+    {
+      "epoch": 0.7869710493726195,
+      "grad_norm": 0.5560535192489624,
+      "learning_rate": 2.157057627540371e-06,
+      "loss": 0.3805,
+      "step": 21050
+    },
+    {
+      "epoch": 0.7871579783629694,
+      "grad_norm": 0.5626355409622192,
+      "learning_rate": 2.1534156991201528e-06,
+      "loss": 0.2906,
+      "step": 21055
+    },
+    {
+      "epoch": 0.7873449073533192,
+      "grad_norm": 0.27478352189064026,
+      "learning_rate": 2.149776476762029e-06,
+      "loss": 0.2498,
+      "step": 21060
+    },
+    {
+      "epoch": 0.787531836343669,
+      "grad_norm": 0.8276031017303467,
+      "learning_rate": 2.146139961721059e-06,
+      "loss": 0.2522,
+      "step": 21065
+    },
+    {
+      "epoch": 0.7877187653340187,
+      "grad_norm": 0.459084153175354,
+      "learning_rate": 2.142506155251377e-06,
+      "loss": 0.3553,
+      "step": 21070
+    },
+    {
+      "epoch": 0.7879056943243685,
+      "grad_norm": 0.6312773823738098,
+      "learning_rate": 2.1388750586061735e-06,
+      "loss": 0.2408,
+      "step": 21075
+    },
+    {
+      "epoch": 0.7880926233147183,
+      "grad_norm": 0.3618089556694031,
+      "learning_rate": 2.1352466730377164e-06,
+      "loss": 0.3,
+      "step": 21080
+    },
+    {
+      "epoch": 0.7882795523050681,
+      "grad_norm": 0.5155732035636902,
+      "learning_rate": 2.131620999797327e-06,
+      "loss": 0.2281,
+      "step": 21085
+    },
+    {
+      "epoch": 0.7884664812954179,
+      "grad_norm": 0.5344289541244507,
+      "learning_rate": 2.1279980401353972e-06,
+      "loss": 0.2373,
+      "step": 21090
+    },
+    {
+      "epoch": 0.7886534102857677,
+      "grad_norm": 0.4500960111618042,
+      "learning_rate": 2.12437779530138e-06,
+      "loss": 0.3069,
+      "step": 21095
+    },
+    {
+      "epoch": 0.7888403392761175,
+      "grad_norm": 0.6218135952949524,
+      "learning_rate": 2.1207602665437953e-06,
+      "loss": 0.288,
+      "step": 21100
+    },
+    {
+      "epoch": 0.7890272682664673,
+      "grad_norm": 0.7415568828582764,
+      "learning_rate": 2.117145455110229e-06,
+      "loss": 0.2563,
+      "step": 21105
+    },
+    {
+      "epoch": 0.7892141972568171,
+      "grad_norm": 0.47236132621765137,
+      "learning_rate": 2.1135333622473208e-06,
+      "loss": 0.3075,
+      "step": 21110
+    },
+    {
+      "epoch": 0.7894011262471669,
+      "grad_norm": 0.3604547083377838,
+      "learning_rate": 2.1099239892007815e-06,
+      "loss": 0.3614,
+      "step": 21115
+    },
+    {
+      "epoch": 0.7895880552375166,
+      "grad_norm": 0.32508572936058044,
+      "learning_rate": 2.1063173372153778e-06,
+      "loss": 0.2853,
+      "step": 21120
+    },
+    {
+      "epoch": 0.7897749842278664,
+      "grad_norm": 0.40659722685813904,
+      "learning_rate": 2.102713407534943e-06,
+      "loss": 0.3314,
+      "step": 21125
+    },
+    {
+      "epoch": 0.7899619132182162,
+      "grad_norm": 0.5361529588699341,
+      "learning_rate": 2.099112201402369e-06,
+      "loss": 0.2541,
+      "step": 21130
+    },
+    {
+      "epoch": 0.7901488422085661,
+      "grad_norm": 0.32573097944259644,
+      "learning_rate": 2.0955137200596077e-06,
+      "loss": 0.2379,
+      "step": 21135
+    },
+    {
+      "epoch": 0.7903357711989158,
+      "grad_norm": 0.5124326348304749,
+      "learning_rate": 2.0919179647476694e-06,
+      "loss": 0.3336,
+      "step": 21140
+    },
+    {
+      "epoch": 0.7905227001892656,
+      "grad_norm": 0.6217731833457947,
+      "learning_rate": 2.0883249367066294e-06,
+      "loss": 0.341,
+      "step": 21145
+    },
+    {
+      "epoch": 0.7907096291796154,
+      "grad_norm": 0.44866645336151123,
+      "learning_rate": 2.0847346371756237e-06,
+      "loss": 0.2955,
+      "step": 21150
+    },
+    {
+      "epoch": 0.7908965581699652,
+      "grad_norm": 0.4434383809566498,
+      "learning_rate": 2.081147067392838e-06,
+      "loss": 0.3195,
+      "step": 21155
+    },
+    {
+      "epoch": 0.791083487160315,
+      "grad_norm": 0.5388154983520508,
+      "learning_rate": 2.0775622285955264e-06,
+      "loss": 0.2346,
+      "step": 21160
+    },
+    {
+      "epoch": 0.7912704161506647,
+      "grad_norm": 0.33477190136909485,
+      "learning_rate": 2.073980122019994e-06,
+      "loss": 0.3177,
+      "step": 21165
+    },
+    {
+      "epoch": 0.7914573451410145,
+      "grad_norm": 0.5977397561073303,
+      "learning_rate": 2.070400748901611e-06,
+      "loss": 0.3434,
+      "step": 21170
+    },
+    {
+      "epoch": 0.7916442741313644,
+      "grad_norm": 0.3142024874687195,
+      "learning_rate": 2.066824110474798e-06,
+      "loss": 0.2778,
+      "step": 21175
+    },
+    {
+      "epoch": 0.7918312031217142,
+      "grad_norm": 0.5590510964393616,
+      "learning_rate": 2.0632502079730356e-06,
+      "loss": 0.2392,
+      "step": 21180
+    },
+    {
+      "epoch": 0.792018132112064,
+      "grad_norm": 0.38960909843444824,
+      "learning_rate": 2.059679042628856e-06,
+      "loss": 0.2559,
+      "step": 21185
+    },
+    {
+      "epoch": 0.7922050611024137,
+      "grad_norm": 0.7024688720703125,
+      "learning_rate": 2.056110615673855e-06,
+      "loss": 0.2569,
+      "step": 21190
+    },
+    {
+      "epoch": 0.7923919900927635,
+      "grad_norm": 0.5414541959762573,
+      "learning_rate": 2.0525449283386855e-06,
+      "loss": 0.2865,
+      "step": 21195
+    },
+    {
+      "epoch": 0.7925789190831133,
+      "grad_norm": 0.4854254126548767,
+      "learning_rate": 2.0489819818530443e-06,
+      "loss": 0.267,
+      "step": 21200
+    },
+    {
+      "epoch": 0.7927658480734631,
+      "grad_norm": 0.34291431307792664,
+      "learning_rate": 2.045421777445694e-06,
+      "loss": 0.3289,
+      "step": 21205
+    },
+    {
+      "epoch": 0.7929527770638128,
+      "grad_norm": 0.366192489862442,
+      "learning_rate": 2.041864316344443e-06,
+      "loss": 0.3097,
+      "step": 21210
+    },
+    {
+      "epoch": 0.7931397060541627,
+      "grad_norm": 0.35338735580444336,
+      "learning_rate": 2.0383095997761628e-06,
+      "loss": 0.2791,
+      "step": 21215
+    },
+    {
+      "epoch": 0.7933266350445125,
+      "grad_norm": 0.5580964684486389,
+      "learning_rate": 2.0347576289667657e-06,
+      "loss": 0.2732,
+      "step": 21220
+    },
+    {
+      "epoch": 0.7935135640348623,
+      "grad_norm": 0.47220098972320557,
+      "learning_rate": 2.031208405141234e-06,
+      "loss": 0.243,
+      "step": 21225
+    },
+    {
+      "epoch": 0.7937004930252121,
+      "grad_norm": 0.6070525646209717,
+      "learning_rate": 2.027661929523588e-06,
+      "loss": 0.2208,
+      "step": 21230
+    },
+    {
+      "epoch": 0.7938874220155618,
+      "grad_norm": 0.4596133828163147,
+      "learning_rate": 2.0241182033369034e-06,
+      "loss": 0.2809,
+      "step": 21235
+    },
+    {
+      "epoch": 0.7940743510059116,
+      "grad_norm": 0.17286808788776398,
+      "learning_rate": 2.0205772278033153e-06,
+      "loss": 0.245,
+      "step": 21240
+    },
+    {
+      "epoch": 0.7942612799962614,
+      "grad_norm": 0.42331114411354065,
+      "learning_rate": 2.017039004143999e-06,
+      "loss": 0.2423,
+      "step": 21245
+    },
+    {
+      "epoch": 0.7944482089866112,
+      "grad_norm": 0.6041852235794067,
+      "learning_rate": 2.013503533579193e-06,
+      "loss": 0.2279,
+      "step": 21250
+    },
+    {
+      "epoch": 0.794635137976961,
+      "grad_norm": 0.4134674668312073,
+      "learning_rate": 2.009970817328173e-06,
+      "loss": 0.2674,
+      "step": 21255
+    },
+    {
+      "epoch": 0.7948220669673108,
+      "grad_norm": 0.44482535123825073,
+      "learning_rate": 2.0064408566092762e-06,
+      "loss": 0.2763,
+      "step": 21260
+    },
+    {
+      "epoch": 0.7950089959576606,
+      "grad_norm": 0.6239652037620544,
+      "learning_rate": 2.002913652639883e-06,
+      "loss": 0.2297,
+      "step": 21265
+    },
+    {
+      "epoch": 0.7951959249480104,
+      "grad_norm": 0.4844609498977661,
+      "learning_rate": 1.999389206636426e-06,
+      "loss": 0.2071,
+      "step": 21270
+    },
+    {
+      "epoch": 0.7953828539383602,
+      "grad_norm": 0.37075021862983704,
+      "learning_rate": 1.9958675198143873e-06,
+      "loss": 0.2251,
+      "step": 21275
+    },
+    {
+      "epoch": 0.79556978292871,
+      "grad_norm": 0.16581667959690094,
+      "learning_rate": 1.992348593388289e-06,
+      "loss": 0.1926,
+      "step": 21280
+    },
+    {
+      "epoch": 0.7957567119190597,
+      "grad_norm": 0.4773890972137451,
+      "learning_rate": 1.9888324285717166e-06,
+      "loss": 0.3475,
+      "step": 21285
+    },
+    {
+      "epoch": 0.7959436409094095,
+      "grad_norm": 0.57837975025177,
+      "learning_rate": 1.985319026577287e-06,
+      "loss": 0.3069,
+      "step": 21290
+    },
+    {
+      "epoch": 0.7961305698997593,
+      "grad_norm": 0.29405614733695984,
+      "learning_rate": 1.9818083886166795e-06,
+      "loss": 0.2997,
+      "step": 21295
+    },
+    {
+      "epoch": 0.7963174988901092,
+      "grad_norm": 0.2895260155200958,
+      "learning_rate": 1.978300515900604e-06,
+      "loss": 0.2789,
+      "step": 21300
+    },
+    {
+      "epoch": 0.7965044278804589,
+      "grad_norm": 0.36291879415512085,
+      "learning_rate": 1.9747954096388343e-06,
+      "loss": 0.2466,
+      "step": 21305
+    },
+    {
+      "epoch": 0.7966913568708087,
+      "grad_norm": 0.477522611618042,
+      "learning_rate": 1.9712930710401735e-06,
+      "loss": 0.2304,
+      "step": 21310
+    },
+    {
+      "epoch": 0.7968782858611585,
+      "grad_norm": 0.7150763273239136,
+      "learning_rate": 1.967793501312483e-06,
+      "loss": 0.2485,
+      "step": 21315
+    },
+    {
+      "epoch": 0.7970652148515083,
+      "grad_norm": 0.4385119378566742,
+      "learning_rate": 1.9642967016626624e-06,
+      "loss": 0.2579,
+      "step": 21320
+    },
+    {
+      "epoch": 0.797252143841858,
+      "grad_norm": 0.4212881624698639,
+      "learning_rate": 1.9608026732966544e-06,
+      "loss": 0.2947,
+      "step": 21325
+    },
+    {
+      "epoch": 0.7974390728322078,
+      "grad_norm": 0.54429692029953,
+      "learning_rate": 1.957311417419455e-06,
+      "loss": 0.2657,
+      "step": 21330
+    },
+    {
+      "epoch": 0.7976260018225576,
+      "grad_norm": 0.6954907178878784,
+      "learning_rate": 1.9538229352350924e-06,
+      "loss": 0.2948,
+      "step": 21335
+    },
+    {
+      "epoch": 0.7978129308129075,
+      "grad_norm": 0.56966632604599,
+      "learning_rate": 1.95033722794665e-06,
+      "loss": 0.2921,
+      "step": 21340
+    },
+    {
+      "epoch": 0.7979998598032573,
+      "grad_norm": 0.6537004113197327,
+      "learning_rate": 1.9468542967562443e-06,
+      "loss": 0.4463,
+      "step": 21345
+    },
+    {
+      "epoch": 0.798186788793607,
+      "grad_norm": 1.18008291721344,
+      "learning_rate": 1.943374142865042e-06,
+      "loss": 0.3565,
+      "step": 21350
+    },
+    {
+      "epoch": 0.7983737177839568,
+      "grad_norm": 0.28714442253112793,
+      "learning_rate": 1.939896767473243e-06,
+      "loss": 0.2965,
+      "step": 21355
+    },
+    {
+      "epoch": 0.7985606467743066,
+      "grad_norm": 1.268236756324768,
+      "learning_rate": 1.936422171780101e-06,
+      "loss": 0.3542,
+      "step": 21360
+    },
+    {
+      "epoch": 0.7987475757646564,
+      "grad_norm": 0.5523670315742493,
+      "learning_rate": 1.9329503569839002e-06,
+      "loss": 0.2992,
+      "step": 21365
+    },
+    {
+      "epoch": 0.7989345047550062,
+      "grad_norm": 0.391519695520401,
+      "learning_rate": 1.92948132428197e-06,
+      "loss": 0.3191,
+      "step": 21370
+    },
+    {
+      "epoch": 0.7991214337453559,
+      "grad_norm": 0.6156255602836609,
+      "learning_rate": 1.926015074870683e-06,
+      "loss": 0.2582,
+      "step": 21375
+    },
+    {
+      "epoch": 0.7993083627357058,
+      "grad_norm": 0.29384270310401917,
+      "learning_rate": 1.9225516099454456e-06,
+      "loss": 0.29,
+      "step": 21380
+    },
+    {
+      "epoch": 0.7994952917260556,
+      "grad_norm": 0.46407338976860046,
+      "learning_rate": 1.919090930700712e-06,
+      "loss": 0.2988,
+      "step": 21385
+    },
+    {
+      "epoch": 0.7996822207164054,
+      "grad_norm": 0.16256499290466309,
+      "learning_rate": 1.915633038329967e-06,
+      "loss": 0.3916,
+      "step": 21390
+    },
+    {
+      "epoch": 0.7998691497067552,
+      "grad_norm": 0.6829771399497986,
+      "learning_rate": 1.912177934025743e-06,
+      "loss": 0.2818,
+      "step": 21395
+    },
+    {
+      "epoch": 0.8000560786971049,
+      "grad_norm": 0.6164911389350891,
+      "learning_rate": 1.9087256189796012e-06,
+      "loss": 0.3556,
+      "step": 21400
+    },
+    {
+      "epoch": 0.8002430076874547,
+      "grad_norm": 0.5262479782104492,
+      "learning_rate": 1.9052760943821513e-06,
+      "loss": 0.2089,
+      "step": 21405
+    },
+    {
+      "epoch": 0.8004299366778045,
+      "grad_norm": 0.3013521730899811,
+      "learning_rate": 1.901829361423031e-06,
+      "loss": 0.2159,
+      "step": 21410
+    },
+    {
+      "epoch": 0.8006168656681543,
+      "grad_norm": 0.37884315848350525,
+      "learning_rate": 1.8983854212909247e-06,
+      "loss": 0.2776,
+      "step": 21415
+    },
+    {
+      "epoch": 0.8008037946585042,
+      "grad_norm": 0.27459797263145447,
+      "learning_rate": 1.894944275173547e-06,
+      "loss": 0.3337,
+      "step": 21420
+    },
+    {
+      "epoch": 0.8009907236488539,
+      "grad_norm": 0.6611989736557007,
+      "learning_rate": 1.8915059242576462e-06,
+      "loss": 0.2814,
+      "step": 21425
+    },
+    {
+      "epoch": 0.8011776526392037,
+      "grad_norm": 0.2187768518924713,
+      "learning_rate": 1.888070369729016e-06,
+      "loss": 0.2271,
+      "step": 21430
+    },
+    {
+      "epoch": 0.8013645816295535,
+      "grad_norm": 0.37497401237487793,
+      "learning_rate": 1.8846376127724775e-06,
+      "loss": 0.2489,
+      "step": 21435
+    },
+    {
+      "epoch": 0.8015515106199033,
+      "grad_norm": 0.44571053981781006,
+      "learning_rate": 1.881207654571895e-06,
+      "loss": 0.2394,
+      "step": 21440
+    },
+    {
+      "epoch": 0.801738439610253,
+      "grad_norm": 0.34269171953201294,
+      "learning_rate": 1.8777804963101553e-06,
+      "loss": 0.2689,
+      "step": 21445
+    },
+    {
+      "epoch": 0.8019253686006028,
+      "grad_norm": 0.3532237410545349,
+      "learning_rate": 1.8743561391691955e-06,
+      "loss": 0.2671,
+      "step": 21450
+    },
+    {
+      "epoch": 0.8021122975909526,
+      "grad_norm": 0.27271535992622375,
+      "learning_rate": 1.8709345843299708e-06,
+      "loss": 0.2663,
+      "step": 21455
+    },
+    {
+      "epoch": 0.8022992265813025,
+      "grad_norm": 0.3709903061389923,
+      "learning_rate": 1.867515832972484e-06,
+      "loss": 0.3473,
+      "step": 21460
+    },
+    {
+      "epoch": 0.8024861555716523,
+      "grad_norm": 0.5782517790794373,
+      "learning_rate": 1.864099886275761e-06,
+      "loss": 0.3205,
+      "step": 21465
+    },
+    {
+      "epoch": 0.802673084562002,
+      "grad_norm": 0.40941113233566284,
+      "learning_rate": 1.8606867454178612e-06,
+      "loss": 0.3808,
+      "step": 21470
+    },
+    {
+      "epoch": 0.8028600135523518,
+      "grad_norm": 0.29615747928619385,
+      "learning_rate": 1.8572764115758846e-06,
+      "loss": 0.3111,
+      "step": 21475
+    },
+    {
+      "epoch": 0.8030469425427016,
+      "grad_norm": 0.8128417134284973,
+      "learning_rate": 1.8538688859259534e-06,
+      "loss": 0.3365,
+      "step": 21480
+    },
+    {
+      "epoch": 0.8032338715330514,
+      "grad_norm": 0.5973396897315979,
+      "learning_rate": 1.850464169643229e-06,
+      "loss": 0.2514,
+      "step": 21485
+    },
+    {
+      "epoch": 0.8034208005234011,
+      "grad_norm": 0.46153488755226135,
+      "learning_rate": 1.8470622639018964e-06,
+      "loss": 0.3385,
+      "step": 21490
+    },
+    {
+      "epoch": 0.8036077295137509,
+      "grad_norm": 0.45193207263946533,
+      "learning_rate": 1.8436631698751806e-06,
+      "loss": 0.2695,
+      "step": 21495
+    },
+    {
+      "epoch": 0.8037946585041007,
+      "grad_norm": 0.5333691835403442,
+      "learning_rate": 1.840266888735326e-06,
+      "loss": 0.2737,
+      "step": 21500
+    },
+    {
+      "epoch": 0.8039815874944506,
+      "grad_norm": 0.46720612049102783,
+      "learning_rate": 1.8368734216536176e-06,
+      "loss": 0.2952,
+      "step": 21505
+    },
+    {
+      "epoch": 0.8041685164848004,
+      "grad_norm": 0.24607418477535248,
+      "learning_rate": 1.8334827698003644e-06,
+      "loss": 0.3072,
+      "step": 21510
+    },
+    {
+      "epoch": 0.8043554454751501,
+      "grad_norm": 0.3917045295238495,
+      "learning_rate": 1.8300949343449003e-06,
+      "loss": 0.2582,
+      "step": 21515
+    },
+    {
+      "epoch": 0.8045423744654999,
+      "grad_norm": 0.27164673805236816,
+      "learning_rate": 1.8267099164555978e-06,
+      "loss": 0.2381,
+      "step": 21520
+    },
+    {
+      "epoch": 0.8047293034558497,
+      "grad_norm": 0.7924597859382629,
+      "learning_rate": 1.8233277172998486e-06,
+      "loss": 0.2594,
+      "step": 21525
+    },
+    {
+      "epoch": 0.8049162324461995,
+      "grad_norm": 0.17336755990982056,
+      "learning_rate": 1.8199483380440808e-06,
+      "loss": 0.2524,
+      "step": 21530
+    },
+    {
+      "epoch": 0.8051031614365493,
+      "grad_norm": 0.5164810419082642,
+      "learning_rate": 1.8165717798537407e-06,
+      "loss": 0.31,
+      "step": 21535
+    },
+    {
+      "epoch": 0.805290090426899,
+      "grad_norm": 0.6605789065361023,
+      "learning_rate": 1.8131980438933117e-06,
+      "loss": 0.3017,
+      "step": 21540
+    },
+    {
+      "epoch": 0.8054770194172489,
+      "grad_norm": 0.31877684593200684,
+      "learning_rate": 1.809827131326294e-06,
+      "loss": 0.2414,
+      "step": 21545
+    },
+    {
+      "epoch": 0.8056639484075987,
+      "grad_norm": 0.4892379641532898,
+      "learning_rate": 1.8064590433152218e-06,
+      "loss": 0.2236,
+      "step": 21550
+    },
+    {
+      "epoch": 0.8058508773979485,
+      "grad_norm": 0.4304451644420624,
+      "learning_rate": 1.8030937810216486e-06,
+      "loss": 0.2741,
+      "step": 21555
+    },
+    {
+      "epoch": 0.8060378063882983,
+      "grad_norm": 0.5623677968978882,
+      "learning_rate": 1.7997313456061615e-06,
+      "loss": 0.2997,
+      "step": 21560
+    },
+    {
+      "epoch": 0.806224735378648,
+      "grad_norm": 0.24887755513191223,
+      "learning_rate": 1.7963717382283663e-06,
+      "loss": 0.2429,
+      "step": 21565
+    },
+    {
+      "epoch": 0.8064116643689978,
+      "grad_norm": 0.4200691282749176,
+      "learning_rate": 1.7930149600468927e-06,
+      "loss": 0.2047,
+      "step": 21570
+    },
+    {
+      "epoch": 0.8065985933593476,
+      "grad_norm": 0.5015040040016174,
+      "learning_rate": 1.7896610122194015e-06,
+      "loss": 0.3563,
+      "step": 21575
+    },
+    {
+      "epoch": 0.8067855223496974,
+      "grad_norm": 0.7050318121910095,
+      "learning_rate": 1.7863098959025692e-06,
+      "loss": 0.2453,
+      "step": 21580
+    },
+    {
+      "epoch": 0.8069724513400472,
+      "grad_norm": 0.39785218238830566,
+      "learning_rate": 1.7829616122521043e-06,
+      "loss": 0.2441,
+      "step": 21585
+    },
+    {
+      "epoch": 0.807159380330397,
+      "grad_norm": 0.5187550187110901,
+      "learning_rate": 1.7796161624227281e-06,
+      "loss": 0.3743,
+      "step": 21590
+    },
+    {
+      "epoch": 0.8073463093207468,
+      "grad_norm": 0.20182538032531738,
+      "learning_rate": 1.7762735475681947e-06,
+      "loss": 0.2975,
+      "step": 21595
+    },
+    {
+      "epoch": 0.8075332383110966,
+      "grad_norm": 0.5243845582008362,
+      "learning_rate": 1.7729337688412772e-06,
+      "loss": 0.2309,
+      "step": 21600
+    },
+    {
+      "epoch": 0.8077201673014464,
+      "grad_norm": 0.39806708693504333,
+      "learning_rate": 1.769596827393768e-06,
+      "loss": 0.3061,
+      "step": 21605
+    },
+    {
+      "epoch": 0.8079070962917961,
+      "grad_norm": 0.37377694249153137,
+      "learning_rate": 1.7662627243764808e-06,
+      "loss": 0.3372,
+      "step": 21610
+    },
+    {
+      "epoch": 0.8080940252821459,
+      "grad_norm": 0.4244140088558197,
+      "learning_rate": 1.7629314609392523e-06,
+      "loss": 0.2972,
+      "step": 21615
+    },
+    {
+      "epoch": 0.8082809542724957,
+      "grad_norm": 0.44188788533210754,
+      "learning_rate": 1.7596030382309436e-06,
+      "loss": 0.3291,
+      "step": 21620
+    },
+    {
+      "epoch": 0.8084678832628456,
+      "grad_norm": 0.37357720732688904,
+      "learning_rate": 1.7562774573994267e-06,
+      "loss": 0.2561,
+      "step": 21625
+    },
+    {
+      "epoch": 0.8086548122531954,
+      "grad_norm": 0.28990504145622253,
+      "learning_rate": 1.7529547195916052e-06,
+      "loss": 0.2949,
+      "step": 21630
+    },
+    {
+      "epoch": 0.8088417412435451,
+      "grad_norm": 0.49605655670166016,
+      "learning_rate": 1.7496348259533902e-06,
+      "loss": 0.3587,
+      "step": 21635
+    },
+    {
+      "epoch": 0.8090286702338949,
+      "grad_norm": 0.3280414044857025,
+      "learning_rate": 1.7463177776297202e-06,
+      "loss": 0.2554,
+      "step": 21640
+    },
+    {
+      "epoch": 0.8092155992242447,
+      "grad_norm": 0.4728814959526062,
+      "learning_rate": 1.7430035757645546e-06,
+      "loss": 0.3598,
+      "step": 21645
+    },
+    {
+      "epoch": 0.8094025282145945,
+      "grad_norm": 0.45704665780067444,
+      "learning_rate": 1.7396922215008628e-06,
+      "loss": 0.3224,
+      "step": 21650
+    },
+    {
+      "epoch": 0.8095894572049442,
+      "grad_norm": 0.47432267665863037,
+      "learning_rate": 1.7363837159806352e-06,
+      "loss": 0.276,
+      "step": 21655
+    },
+    {
+      "epoch": 0.809776386195294,
+      "grad_norm": 0.4984903037548065,
+      "learning_rate": 1.7330780603448794e-06,
+      "loss": 0.3783,
+      "step": 21660
+    },
+    {
+      "epoch": 0.8099633151856439,
+      "grad_norm": 0.4554220736026764,
+      "learning_rate": 1.7297752557336257e-06,
+      "loss": 0.2547,
+      "step": 21665
+    },
+    {
+      "epoch": 0.8101502441759937,
+      "grad_norm": 0.3859618008136749,
+      "learning_rate": 1.7264753032859115e-06,
+      "loss": 0.3227,
+      "step": 21670
+    },
+    {
+      "epoch": 0.8103371731663435,
+      "grad_norm": 0.2784779369831085,
+      "learning_rate": 1.7231782041398015e-06,
+      "loss": 0.3581,
+      "step": 21675
+    },
+    {
+      "epoch": 0.8105241021566932,
+      "grad_norm": 0.4222283363342285,
+      "learning_rate": 1.7198839594323658e-06,
+      "loss": 0.3204,
+      "step": 21680
+    },
+    {
+      "epoch": 0.810711031147043,
+      "grad_norm": 0.8711488842964172,
+      "learning_rate": 1.7165925702997e-06,
+      "loss": 0.3242,
+      "step": 21685
+    },
+    {
+      "epoch": 0.8108979601373928,
+      "grad_norm": 0.39160558581352234,
+      "learning_rate": 1.7133040378769039e-06,
+      "loss": 0.2511,
+      "step": 21690
+    },
+    {
+      "epoch": 0.8110848891277426,
+      "grad_norm": 1.2719933986663818,
+      "learning_rate": 1.7100183632981039e-06,
+      "loss": 0.3113,
+      "step": 21695
+    },
+    {
+      "epoch": 0.8112718181180923,
+      "grad_norm": 0.23821887373924255,
+      "learning_rate": 1.706735547696432e-06,
+      "loss": 0.3062,
+      "step": 21700
+    },
+    {
+      "epoch": 0.8114587471084422,
+      "grad_norm": 0.4011630117893219,
+      "learning_rate": 1.7034555922040351e-06,
+      "loss": 0.2491,
+      "step": 21705
+    },
+    {
+      "epoch": 0.811645676098792,
+      "grad_norm": 0.28980720043182373,
+      "learning_rate": 1.7001784979520808e-06,
+      "loss": 0.2797,
+      "step": 21710
+    },
+    {
+      "epoch": 0.8118326050891418,
+      "grad_norm": 0.41728466749191284,
+      "learning_rate": 1.6969042660707413e-06,
+      "loss": 0.3272,
+      "step": 21715
+    },
+    {
+      "epoch": 0.8120195340794916,
+      "grad_norm": 0.22036214172840118,
+      "learning_rate": 1.693632897689208e-06,
+      "loss": 0.2638,
+      "step": 21720
+    },
+    {
+      "epoch": 0.8122064630698413,
+      "grad_norm": 0.5138218402862549,
+      "learning_rate": 1.6903643939356784e-06,
+      "loss": 0.2993,
+      "step": 21725
+    },
+    {
+      "epoch": 0.8123933920601911,
+      "grad_norm": 0.6489676237106323,
+      "learning_rate": 1.6870987559373709e-06,
+      "loss": 0.2849,
+      "step": 21730
+    },
+    {
+      "epoch": 0.8125803210505409,
+      "grad_norm": 0.2285657376050949,
+      "learning_rate": 1.6838359848205055e-06,
+      "loss": 0.2543,
+      "step": 21735
+    },
+    {
+      "epoch": 0.8127672500408907,
+      "grad_norm": 0.29291707277297974,
+      "learning_rate": 1.6805760817103201e-06,
+      "loss": 0.2461,
+      "step": 21740
+    },
+    {
+      "epoch": 0.8129541790312405,
+      "grad_norm": 0.4174777865409851,
+      "learning_rate": 1.6773190477310652e-06,
+      "loss": 0.2149,
+      "step": 21745
+    },
+    {
+      "epoch": 0.8131411080215903,
+      "grad_norm": 0.2554204761981964,
+      "learning_rate": 1.6740648840059958e-06,
+      "loss": 0.2763,
+      "step": 21750
+    },
+    {
+      "epoch": 0.8133280370119401,
+      "grad_norm": 0.8167181611061096,
+      "learning_rate": 1.6708135916573797e-06,
+      "loss": 0.2746,
+      "step": 21755
+    },
+    {
+      "epoch": 0.8135149660022899,
+      "grad_norm": 0.304171085357666,
+      "learning_rate": 1.6675651718064922e-06,
+      "loss": 0.2495,
+      "step": 21760
+    },
+    {
+      "epoch": 0.8137018949926397,
+      "grad_norm": 0.39825260639190674,
+      "learning_rate": 1.6643196255736239e-06,
+      "loss": 0.2251,
+      "step": 21765
+    },
+    {
+      "epoch": 0.8138888239829895,
+      "grad_norm": 0.36910250782966614,
+      "learning_rate": 1.661076954078068e-06,
+      "loss": 0.3272,
+      "step": 21770
+    },
+    {
+      "epoch": 0.8140757529733392,
+      "grad_norm": 0.31329768896102905,
+      "learning_rate": 1.6578371584381326e-06,
+      "loss": 0.2188,
+      "step": 21775
+    },
+    {
+      "epoch": 0.814262681963689,
+      "grad_norm": 0.2181299477815628,
+      "learning_rate": 1.6546002397711247e-06,
+      "loss": 0.3479,
+      "step": 21780
+    },
+    {
+      "epoch": 0.8144496109540388,
+      "grad_norm": 0.8821444511413574,
+      "learning_rate": 1.6513661991933694e-06,
+      "loss": 0.2365,
+      "step": 21785
+    },
+    {
+      "epoch": 0.8146365399443887,
+      "grad_norm": 0.3816254436969757,
+      "learning_rate": 1.6481350378201954e-06,
+      "loss": 0.348,
+      "step": 21790
+    },
+    {
+      "epoch": 0.8148234689347384,
+      "grad_norm": 0.7494456171989441,
+      "learning_rate": 1.644906756765935e-06,
+      "loss": 0.2938,
+      "step": 21795
+    },
+    {
+      "epoch": 0.8150103979250882,
+      "grad_norm": 0.5195353031158447,
+      "learning_rate": 1.6416813571439305e-06,
+      "loss": 0.3243,
+      "step": 21800
+    },
+    {
+      "epoch": 0.815197326915438,
+      "grad_norm": 0.3925173282623291,
+      "learning_rate": 1.638458840066528e-06,
+      "loss": 0.2577,
+      "step": 21805
+    },
+    {
+      "epoch": 0.8153842559057878,
+      "grad_norm": 0.5360116958618164,
+      "learning_rate": 1.635239206645085e-06,
+      "loss": 0.3019,
+      "step": 21810
+    },
+    {
+      "epoch": 0.8155711848961376,
+      "grad_norm": 0.384461909532547,
+      "learning_rate": 1.632022457989958e-06,
+      "loss": 0.2346,
+      "step": 21815
+    },
+    {
+      "epoch": 0.8157581138864873,
+      "grad_norm": 0.3784307539463043,
+      "learning_rate": 1.6288085952105126e-06,
+      "loss": 0.2576,
+      "step": 21820
+    },
+    {
+      "epoch": 0.8159450428768371,
+      "grad_norm": 0.3324683606624603,
+      "learning_rate": 1.6255976194151168e-06,
+      "loss": 0.3859,
+      "step": 21825
+    },
+    {
+      "epoch": 0.816131971867187,
+      "grad_norm": 0.6262183785438538,
+      "learning_rate": 1.6223895317111449e-06,
+      "loss": 0.2211,
+      "step": 21830
+    },
+    {
+      "epoch": 0.8163189008575368,
+      "grad_norm": 0.3607613444328308,
+      "learning_rate": 1.6191843332049762e-06,
+      "loss": 0.2672,
+      "step": 21835
+    },
+    {
+      "epoch": 0.8165058298478866,
+      "grad_norm": 0.33100825548171997,
+      "learning_rate": 1.615982025001991e-06,
+      "loss": 0.2289,
+      "step": 21840
+    },
+    {
+      "epoch": 0.8166927588382363,
+      "grad_norm": 0.7562757134437561,
+      "learning_rate": 1.6127826082065723e-06,
+      "loss": 0.2542,
+      "step": 21845
+    },
+    {
+      "epoch": 0.8168796878285861,
+      "grad_norm": 0.359550803899765,
+      "learning_rate": 1.6095860839221055e-06,
+      "loss": 0.3022,
+      "step": 21850
+    },
+    {
+      "epoch": 0.8170666168189359,
+      "grad_norm": 0.8539102077484131,
+      "learning_rate": 1.6063924532509856e-06,
+      "loss": 0.3232,
+      "step": 21855
+    },
+    {
+      "epoch": 0.8172535458092857,
+      "grad_norm": 0.23160411417484283,
+      "learning_rate": 1.603201717294599e-06,
+      "loss": 0.2307,
+      "step": 21860
+    },
+    {
+      "epoch": 0.8174404747996354,
+      "grad_norm": 0.5298076272010803,
+      "learning_rate": 1.6000138771533424e-06,
+      "loss": 0.223,
+      "step": 21865
+    },
+    {
+      "epoch": 0.8176274037899853,
+      "grad_norm": 0.5675511360168457,
+      "learning_rate": 1.5968289339266084e-06,
+      "loss": 0.3385,
+      "step": 21870
+    },
+    {
+      "epoch": 0.8178143327803351,
+      "grad_norm": 0.8937890529632568,
+      "learning_rate": 1.5936468887127932e-06,
+      "loss": 0.3094,
+      "step": 21875
+    },
+    {
+      "epoch": 0.8180012617706849,
+      "grad_norm": 0.418905645608902,
+      "learning_rate": 1.5904677426092964e-06,
+      "loss": 0.3251,
+      "step": 21880
+    },
+    {
+      "epoch": 0.8181881907610347,
+      "grad_norm": 0.26117852330207825,
+      "learning_rate": 1.587291496712512e-06,
+      "loss": 0.3468,
+      "step": 21885
+    },
+    {
+      "epoch": 0.8183751197513844,
+      "grad_norm": 0.9076757431030273,
+      "learning_rate": 1.584118152117835e-06,
+      "loss": 0.26,
+      "step": 21890
+    },
+    {
+      "epoch": 0.8185620487417342,
+      "grad_norm": 0.46946465969085693,
+      "learning_rate": 1.5809477099196592e-06,
+      "loss": 0.3238,
+      "step": 21895
+    },
+    {
+      "epoch": 0.818748977732084,
+      "grad_norm": 0.28860294818878174,
+      "learning_rate": 1.577780171211385e-06,
+      "loss": 0.2435,
+      "step": 21900
+    },
+    {
+      "epoch": 0.8189359067224338,
+      "grad_norm": 0.24636733531951904,
+      "learning_rate": 1.5746155370853998e-06,
+      "loss": 0.2566,
+      "step": 21905
+    },
+    {
+      "epoch": 0.8191228357127837,
+      "grad_norm": 0.12167595326900482,
+      "learning_rate": 1.571453808633101e-06,
+      "loss": 0.3848,
+      "step": 21910
+    },
+    {
+      "epoch": 0.8193097647031334,
+      "grad_norm": 0.6894955039024353,
+      "learning_rate": 1.5682949869448715e-06,
+      "loss": 0.3028,
+      "step": 21915
+    },
+    {
+      "epoch": 0.8194966936934832,
+      "grad_norm": 0.3803219199180603,
+      "learning_rate": 1.5651390731101035e-06,
+      "loss": 0.2547,
+      "step": 21920
+    },
+    {
+      "epoch": 0.819683622683833,
+      "grad_norm": 0.28584155440330505,
+      "learning_rate": 1.5619860682171817e-06,
+      "loss": 0.2718,
+      "step": 21925
+    },
+    {
+      "epoch": 0.8198705516741828,
+      "grad_norm": 0.5630354285240173,
+      "learning_rate": 1.558835973353483e-06,
+      "loss": 0.2787,
+      "step": 21930
+    },
+    {
+      "epoch": 0.8200574806645325,
+      "grad_norm": 0.2100723534822464,
+      "learning_rate": 1.5556887896053896e-06,
+      "loss": 0.3048,
+      "step": 21935
+    },
+    {
+      "epoch": 0.8202444096548823,
+      "grad_norm": 0.38809457421302795,
+      "learning_rate": 1.5525445180582721e-06,
+      "loss": 0.2565,
+      "step": 21940
+    },
+    {
+      "epoch": 0.8204313386452321,
+      "grad_norm": 0.4679727554321289,
+      "learning_rate": 1.549403159796501e-06,
+      "loss": 0.2861,
+      "step": 21945
+    },
+    {
+      "epoch": 0.820618267635582,
+      "grad_norm": 0.39624401926994324,
+      "learning_rate": 1.5462647159034362e-06,
+      "loss": 0.2531,
+      "step": 21950
+    },
+    {
+      "epoch": 0.8208051966259318,
+      "grad_norm": 0.32841363549232483,
+      "learning_rate": 1.543129187461444e-06,
+      "loss": 0.2593,
+      "step": 21955
+    },
+    {
+      "epoch": 0.8209921256162815,
+      "grad_norm": 0.3810194134712219,
+      "learning_rate": 1.539996575551872e-06,
+      "loss": 0.2991,
+      "step": 21960
+    },
+    {
+      "epoch": 0.8211790546066313,
+      "grad_norm": 0.37340226769447327,
+      "learning_rate": 1.5368668812550724e-06,
+      "loss": 0.2569,
+      "step": 21965
+    },
+    {
+      "epoch": 0.8213659835969811,
+      "grad_norm": 0.511825680732727,
+      "learning_rate": 1.5337401056503876e-06,
+      "loss": 0.2649,
+      "step": 21970
+    },
+    {
+      "epoch": 0.8215529125873309,
+      "grad_norm": 0.5602962374687195,
+      "learning_rate": 1.5306162498161493e-06,
+      "loss": 0.2674,
+      "step": 21975
+    },
+    {
+      "epoch": 0.8217398415776807,
+      "grad_norm": 0.27479755878448486,
+      "learning_rate": 1.52749531482969e-06,
+      "loss": 0.304,
+      "step": 21980
+    },
+    {
+      "epoch": 0.8219267705680304,
+      "grad_norm": 0.3318054676055908,
+      "learning_rate": 1.5243773017673292e-06,
+      "loss": 0.2914,
+      "step": 21985
+    },
+    {
+      "epoch": 0.8221136995583802,
+      "grad_norm": 0.44144922494888306,
+      "learning_rate": 1.5212622117043784e-06,
+      "loss": 0.262,
+      "step": 21990
+    },
+    {
+      "epoch": 0.8223006285487301,
+      "grad_norm": 0.8043171763420105,
+      "learning_rate": 1.5181500457151432e-06,
+      "loss": 0.3467,
+      "step": 21995
+    },
+    {
+      "epoch": 0.8224875575390799,
+      "grad_norm": 0.4182523787021637,
+      "learning_rate": 1.5150408048729226e-06,
+      "loss": 0.3729,
+      "step": 22000
+    },
+    {
+      "epoch": 0.8226744865294296,
+      "grad_norm": 0.22673340141773224,
+      "learning_rate": 1.5119344902500022e-06,
+      "loss": 0.3049,
+      "step": 22005
+    },
+    {
+      "epoch": 0.8228614155197794,
+      "grad_norm": 0.3719673454761505,
+      "learning_rate": 1.5088311029176628e-06,
+      "loss": 0.2519,
+      "step": 22010
+    },
+    {
+      "epoch": 0.8230483445101292,
+      "grad_norm": 0.3984321355819702,
+      "learning_rate": 1.5057306439461738e-06,
+      "loss": 0.2315,
+      "step": 22015
+    },
+    {
+      "epoch": 0.823235273500479,
+      "grad_norm": 0.24741186201572418,
+      "learning_rate": 1.5026331144047935e-06,
+      "loss": 0.3394,
+      "step": 22020
+    },
+    {
+      "epoch": 0.8234222024908288,
+      "grad_norm": 0.43265968561172485,
+      "learning_rate": 1.4995385153617725e-06,
+      "loss": 0.3895,
+      "step": 22025
+    },
+    {
+      "epoch": 0.8236091314811785,
+      "grad_norm": 0.34110236167907715,
+      "learning_rate": 1.4964468478843496e-06,
+      "loss": 0.2754,
+      "step": 22030
+    },
+    {
+      "epoch": 0.8237960604715284,
+      "grad_norm": 0.21157729625701904,
+      "learning_rate": 1.4933581130387509e-06,
+      "loss": 0.253,
+      "step": 22035
+    },
+    {
+      "epoch": 0.8239829894618782,
+      "grad_norm": 0.5705170035362244,
+      "learning_rate": 1.4902723118901907e-06,
+      "loss": 0.2533,
+      "step": 22040
+    },
+    {
+      "epoch": 0.824169918452228,
+      "grad_norm": 0.41971489787101746,
+      "learning_rate": 1.4871894455028778e-06,
+      "loss": 0.2568,
+      "step": 22045
+    },
+    {
+      "epoch": 0.8243568474425778,
+      "grad_norm": 0.5276783108711243,
+      "learning_rate": 1.4841095149399998e-06,
+      "loss": 0.3132,
+      "step": 22050
+    },
+    {
+      "epoch": 0.8245437764329275,
+      "grad_norm": 0.2946646213531494,
+      "learning_rate": 1.481032521263739e-06,
+      "loss": 0.304,
+      "step": 22055
+    },
+    {
+      "epoch": 0.8247307054232773,
+      "grad_norm": 0.6871627569198608,
+      "learning_rate": 1.4779584655352652e-06,
+      "loss": 0.322,
+      "step": 22060
+    },
+    {
+      "epoch": 0.8249176344136271,
+      "grad_norm": 0.37137484550476074,
+      "learning_rate": 1.4748873488147264e-06,
+      "loss": 0.3303,
+      "step": 22065
+    },
+    {
+      "epoch": 0.8251045634039769,
+      "grad_norm": 0.34239670634269714,
+      "learning_rate": 1.4718191721612684e-06,
+      "loss": 0.297,
+      "step": 22070
+    },
+    {
+      "epoch": 0.8252914923943268,
+      "grad_norm": 0.6226434707641602,
+      "learning_rate": 1.468753936633014e-06,
+      "loss": 0.3276,
+      "step": 22075
+    },
+    {
+      "epoch": 0.8254784213846765,
+      "grad_norm": 0.5200907588005066,
+      "learning_rate": 1.4656916432870737e-06,
+      "loss": 0.2458,
+      "step": 22080
+    },
+    {
+      "epoch": 0.8256653503750263,
+      "grad_norm": 0.513783872127533,
+      "learning_rate": 1.4626322931795489e-06,
+      "loss": 0.2929,
+      "step": 22085
+    },
+    {
+      "epoch": 0.8258522793653761,
+      "grad_norm": 3.522709846496582,
+      "learning_rate": 1.4595758873655198e-06,
+      "loss": 0.2378,
+      "step": 22090
+    },
+    {
+      "epoch": 0.8260392083557259,
+      "grad_norm": 0.12881094217300415,
+      "learning_rate": 1.4565224268990507e-06,
+      "loss": 0.2293,
+      "step": 22095
+    },
+    {
+      "epoch": 0.8262261373460756,
+      "grad_norm": 0.5438992381095886,
+      "learning_rate": 1.4534719128331953e-06,
+      "loss": 0.2544,
+      "step": 22100
+    },
+    {
+      "epoch": 0.8264130663364254,
+      "grad_norm": 0.28744834661483765,
+      "learning_rate": 1.4504243462199896e-06,
+      "loss": 0.2217,
+      "step": 22105
+    },
+    {
+      "epoch": 0.8265999953267752,
+      "grad_norm": 0.39146536588668823,
+      "learning_rate": 1.4473797281104485e-06,
+      "loss": 0.2257,
+      "step": 22110
+    },
+    {
+      "epoch": 0.8267869243171251,
+      "grad_norm": 0.6850770115852356,
+      "learning_rate": 1.4443380595545787e-06,
+      "loss": 0.3273,
+      "step": 22115
+    },
+    {
+      "epoch": 0.8269738533074749,
+      "grad_norm": 0.3585718870162964,
+      "learning_rate": 1.4412993416013588e-06,
+      "loss": 0.2597,
+      "step": 22120
+    },
+    {
+      "epoch": 0.8271607822978246,
+      "grad_norm": 0.6018190979957581,
+      "learning_rate": 1.4382635752987606e-06,
+      "loss": 0.2203,
+      "step": 22125
+    },
+    {
+      "epoch": 0.8273477112881744,
+      "grad_norm": 0.48121556639671326,
+      "learning_rate": 1.43523076169373e-06,
+      "loss": 0.2179,
+      "step": 22130
+    },
+    {
+      "epoch": 0.8275346402785242,
+      "grad_norm": 0.3579247295856476,
+      "learning_rate": 1.432200901832198e-06,
+      "loss": 0.2088,
+      "step": 22135
+    },
+    {
+      "epoch": 0.827721569268874,
+      "grad_norm": 0.58817058801651,
+      "learning_rate": 1.4291739967590746e-06,
+      "loss": 0.2333,
+      "step": 22140
+    },
+    {
+      "epoch": 0.8279084982592237,
+      "grad_norm": 0.44945281744003296,
+      "learning_rate": 1.4261500475182543e-06,
+      "loss": 0.2591,
+      "step": 22145
+    },
+    {
+      "epoch": 0.8280954272495735,
+      "grad_norm": 0.6072959899902344,
+      "learning_rate": 1.4231290551526133e-06,
+      "loss": 0.2795,
+      "step": 22150
+    },
+    {
+      "epoch": 0.8282823562399234,
+      "grad_norm": 0.5549390912055969,
+      "learning_rate": 1.4201110207039993e-06,
+      "loss": 0.2617,
+      "step": 22155
+    },
+    {
+      "epoch": 0.8284692852302732,
+      "grad_norm": 0.419281929731369,
+      "learning_rate": 1.4170959452132526e-06,
+      "loss": 0.2794,
+      "step": 22160
+    },
+    {
+      "epoch": 0.828656214220623,
+      "grad_norm": 0.5446984171867371,
+      "learning_rate": 1.414083829720181e-06,
+      "loss": 0.2595,
+      "step": 22165
+    },
+    {
+      "epoch": 0.8288431432109727,
+      "grad_norm": 0.562941312789917,
+      "learning_rate": 1.4110746752635806e-06,
+      "loss": 0.3615,
+      "step": 22170
+    },
+    {
+      "epoch": 0.8290300722013225,
+      "grad_norm": 0.42342409491539,
+      "learning_rate": 1.4080684828812219e-06,
+      "loss": 0.2564,
+      "step": 22175
+    },
+    {
+      "epoch": 0.8292170011916723,
+      "grad_norm": 0.3505541682243347,
+      "learning_rate": 1.4050652536098518e-06,
+      "loss": 0.2427,
+      "step": 22180
+    },
+    {
+      "epoch": 0.8294039301820221,
+      "grad_norm": 0.5046654939651489,
+      "learning_rate": 1.4020649884851988e-06,
+      "loss": 0.2361,
+      "step": 22185
+    },
+    {
+      "epoch": 0.8295908591723719,
+      "grad_norm": 0.46937140822410583,
+      "learning_rate": 1.3990676885419685e-06,
+      "loss": 0.2695,
+      "step": 22190
+    },
+    {
+      "epoch": 0.8297777881627217,
+      "grad_norm": 1.1296278238296509,
+      "learning_rate": 1.3960733548138472e-06,
+      "loss": 0.3084,
+      "step": 22195
+    },
+    {
+      "epoch": 0.8299647171530715,
+      "grad_norm": 0.5058994889259338,
+      "learning_rate": 1.3930819883334901e-06,
+      "loss": 0.2994,
+      "step": 22200
+    },
+    {
+      "epoch": 0.8301516461434213,
+      "grad_norm": 0.26385819911956787,
+      "learning_rate": 1.3900935901325374e-06,
+      "loss": 0.2441,
+      "step": 22205
+    },
+    {
+      "epoch": 0.8303385751337711,
+      "grad_norm": 0.5458356738090515,
+      "learning_rate": 1.3871081612415982e-06,
+      "loss": 0.391,
+      "step": 22210
+    },
+    {
+      "epoch": 0.8305255041241209,
+      "grad_norm": 0.28343480825424194,
+      "learning_rate": 1.3841257026902665e-06,
+      "loss": 0.2884,
+      "step": 22215
+    },
+    {
+      "epoch": 0.8307124331144706,
+      "grad_norm": 0.3122907876968384,
+      "learning_rate": 1.3811462155071043e-06,
+      "loss": 0.2776,
+      "step": 22220
+    },
+    {
+      "epoch": 0.8308993621048204,
+      "grad_norm": 0.7874910831451416,
+      "learning_rate": 1.3781697007196493e-06,
+      "loss": 0.2769,
+      "step": 22225
+    },
+    {
+      "epoch": 0.8310862910951702,
+      "grad_norm": 0.5420087575912476,
+      "learning_rate": 1.3751961593544171e-06,
+      "loss": 0.2888,
+      "step": 22230
+    },
+    {
+      "epoch": 0.83127322008552,
+      "grad_norm": 0.2580297887325287,
+      "learning_rate": 1.3722255924368965e-06,
+      "loss": 0.3204,
+      "step": 22235
+    },
+    {
+      "epoch": 0.8314601490758698,
+      "grad_norm": 0.40570923686027527,
+      "learning_rate": 1.3692580009915557e-06,
+      "loss": 0.2792,
+      "step": 22240
+    },
+    {
+      "epoch": 0.8316470780662196,
+      "grad_norm": 0.3946312963962555,
+      "learning_rate": 1.3662933860418249e-06,
+      "loss": 0.2946,
+      "step": 22245
+    },
+    {
+      "epoch": 0.8318340070565694,
+      "grad_norm": 0.26744893193244934,
+      "learning_rate": 1.3633317486101205e-06,
+      "loss": 0.2792,
+      "step": 22250
+    },
+    {
+      "epoch": 0.8320209360469192,
+      "grad_norm": 0.41380831599235535,
+      "learning_rate": 1.3603730897178226e-06,
+      "loss": 0.2922,
+      "step": 22255
+    },
+    {
+      "epoch": 0.832207865037269,
+      "grad_norm": 0.27136343717575073,
+      "learning_rate": 1.3574174103852922e-06,
+      "loss": 0.315,
+      "step": 22260
+    },
+    {
+      "epoch": 0.8323947940276187,
+      "grad_norm": 0.6136317849159241,
+      "learning_rate": 1.3544647116318522e-06,
+      "loss": 0.2856,
+      "step": 22265
+    },
+    {
+      "epoch": 0.8325817230179685,
+      "grad_norm": 0.42765137553215027,
+      "learning_rate": 1.3515149944758098e-06,
+      "loss": 0.208,
+      "step": 22270
+    },
+    {
+      "epoch": 0.8327686520083183,
+      "grad_norm": 0.41445282101631165,
+      "learning_rate": 1.3485682599344351e-06,
+      "loss": 0.3367,
+      "step": 22275
+    },
+    {
+      "epoch": 0.8329555809986682,
+      "grad_norm": 0.2044340819120407,
+      "learning_rate": 1.3456245090239706e-06,
+      "loss": 0.2173,
+      "step": 22280
+    },
+    {
+      "epoch": 0.833142509989018,
+      "grad_norm": 0.3924507796764374,
+      "learning_rate": 1.3426837427596363e-06,
+      "loss": 0.3163,
+      "step": 22285
+    },
+    {
+      "epoch": 0.8333294389793677,
+      "grad_norm": 0.6259861588478088,
+      "learning_rate": 1.339745962155613e-06,
+      "loss": 0.3531,
+      "step": 22290
+    },
+    {
+      "epoch": 0.8335163679697175,
+      "grad_norm": 0.6004974246025085,
+      "learning_rate": 1.336811168225063e-06,
+      "loss": 0.2664,
+      "step": 22295
+    },
+    {
+      "epoch": 0.8337032969600673,
+      "grad_norm": 0.3294597268104553,
+      "learning_rate": 1.333879361980106e-06,
+      "loss": 0.2862,
+      "step": 22300
+    },
+    {
+      "epoch": 0.8338902259504171,
+      "grad_norm": 0.8053619861602783,
+      "learning_rate": 1.3309505444318439e-06,
+      "loss": 0.2701,
+      "step": 22305
+    },
+    {
+      "epoch": 0.8340771549407668,
+      "grad_norm": 0.47687965631484985,
+      "learning_rate": 1.328024716590336e-06,
+      "loss": 0.2936,
+      "step": 22310
+    },
+    {
+      "epoch": 0.8342640839311166,
+      "grad_norm": 0.2641568183898926,
+      "learning_rate": 1.3251018794646232e-06,
+      "loss": 0.2435,
+      "step": 22315
+    },
+    {
+      "epoch": 0.8344510129214665,
+      "grad_norm": 0.3692987859249115,
+      "learning_rate": 1.3221820340627044e-06,
+      "loss": 0.2775,
+      "step": 22320
+    },
+    {
+      "epoch": 0.8346379419118163,
+      "grad_norm": 0.6120216846466064,
+      "learning_rate": 1.319265181391549e-06,
+      "loss": 0.3437,
+      "step": 22325
+    },
+    {
+      "epoch": 0.8348248709021661,
+      "grad_norm": 0.5429954528808594,
+      "learning_rate": 1.3163513224571012e-06,
+      "loss": 0.3351,
+      "step": 22330
+    },
+    {
+      "epoch": 0.8350117998925158,
+      "grad_norm": 0.5047636032104492,
+      "learning_rate": 1.3134404582642612e-06,
+      "loss": 0.3325,
+      "step": 22335
+    },
+    {
+      "epoch": 0.8351987288828656,
+      "grad_norm": 0.5833032131195068,
+      "learning_rate": 1.3105325898169075e-06,
+      "loss": 0.2929,
+      "step": 22340
+    },
+    {
+      "epoch": 0.8353856578732154,
+      "grad_norm": 0.35585522651672363,
+      "learning_rate": 1.3076277181178775e-06,
+      "loss": 0.246,
+      "step": 22345
+    },
+    {
+      "epoch": 0.8355725868635652,
+      "grad_norm": 0.4089027941226959,
+      "learning_rate": 1.3047258441689815e-06,
+      "loss": 0.3233,
+      "step": 22350
+    },
+    {
+      "epoch": 0.835759515853915,
+      "grad_norm": 0.4178864657878876,
+      "learning_rate": 1.3018269689709883e-06,
+      "loss": 0.3107,
+      "step": 22355
+    },
+    {
+      "epoch": 0.8359464448442648,
+      "grad_norm": 0.6142130494117737,
+      "learning_rate": 1.2989310935236421e-06,
+      "loss": 0.2829,
+      "step": 22360
+    },
+    {
+      "epoch": 0.8361333738346146,
+      "grad_norm": 0.3945229947566986,
+      "learning_rate": 1.2960382188256438e-06,
+      "loss": 0.2744,
+      "step": 22365
+    },
+    {
+      "epoch": 0.8363203028249644,
+      "grad_norm": 0.3515075147151947,
+      "learning_rate": 1.2931483458746618e-06,
+      "loss": 0.2536,
+      "step": 22370
+    },
+    {
+      "epoch": 0.8365072318153142,
+      "grad_norm": 0.31164297461509705,
+      "learning_rate": 1.2902614756673348e-06,
+      "loss": 0.2761,
+      "step": 22375
+    },
+    {
+      "epoch": 0.836694160805664,
+      "grad_norm": 0.24446457624435425,
+      "learning_rate": 1.2873776091992574e-06,
+      "loss": 0.3083,
+      "step": 22380
+    },
+    {
+      "epoch": 0.8368810897960137,
+      "grad_norm": 0.8027161955833435,
+      "learning_rate": 1.284496747464996e-06,
+      "loss": 0.314,
+      "step": 22385
+    },
+    {
+      "epoch": 0.8370680187863635,
+      "grad_norm": 0.2881372570991516,
+      "learning_rate": 1.281618891458073e-06,
+      "loss": 0.2548,
+      "step": 22390
+    },
+    {
+      "epoch": 0.8372549477767133,
+      "grad_norm": 0.6596559286117554,
+      "learning_rate": 1.278744042170984e-06,
+      "loss": 0.2581,
+      "step": 22395
+    },
+    {
+      "epoch": 0.8374418767670632,
+      "grad_norm": 0.5899741649627686,
+      "learning_rate": 1.2758722005951773e-06,
+      "loss": 0.3011,
+      "step": 22400
+    },
+    {
+      "epoch": 0.8376288057574129,
+      "grad_norm": 0.2703937888145447,
+      "learning_rate": 1.273003367721073e-06,
+      "loss": 0.252,
+      "step": 22405
+    },
+    {
+      "epoch": 0.8378157347477627,
+      "grad_norm": 0.36412695050239563,
+      "learning_rate": 1.2701375445380459e-06,
+      "loss": 0.2738,
+      "step": 22410
+    },
+    {
+      "epoch": 0.8380026637381125,
+      "grad_norm": 0.35132333636283875,
+      "learning_rate": 1.2672747320344359e-06,
+      "loss": 0.2776,
+      "step": 22415
+    },
+    {
+      "epoch": 0.8381895927284623,
+      "grad_norm": 0.4513058662414551,
+      "learning_rate": 1.2644149311975494e-06,
+      "loss": 0.2587,
+      "step": 22420
+    },
+    {
+      "epoch": 0.838376521718812,
+      "grad_norm": 0.3663261830806732,
+      "learning_rate": 1.2615581430136449e-06,
+      "loss": 0.2893,
+      "step": 22425
+    },
+    {
+      "epoch": 0.8385634507091618,
+      "grad_norm": 0.7394602298736572,
+      "learning_rate": 1.25870436846795e-06,
+      "loss": 0.2527,
+      "step": 22430
+    },
+    {
+      "epoch": 0.8387503796995116,
+      "grad_norm": 0.3970726728439331,
+      "learning_rate": 1.255853608544647e-06,
+      "loss": 0.2582,
+      "step": 22435
+    },
+    {
+      "epoch": 0.8389373086898615,
+      "grad_norm": 0.6118816137313843,
+      "learning_rate": 1.253005864226885e-06,
+      "loss": 0.3117,
+      "step": 22440
+    },
+    {
+      "epoch": 0.8391242376802113,
+      "grad_norm": 0.561377763748169,
+      "learning_rate": 1.2501611364967647e-06,
+      "loss": 0.2964,
+      "step": 22445
+    },
+    {
+      "epoch": 0.839311166670561,
+      "grad_norm": 0.41377291083335876,
+      "learning_rate": 1.247319426335356e-06,
+      "loss": 0.2623,
+      "step": 22450
+    },
+    {
+      "epoch": 0.8394980956609108,
+      "grad_norm": 0.6088568568229675,
+      "learning_rate": 1.2444807347226795e-06,
+      "loss": 0.2458,
+      "step": 22455
+    },
+    {
+      "epoch": 0.8396850246512606,
+      "grad_norm": 0.2927430272102356,
+      "learning_rate": 1.241645062637723e-06,
+      "loss": 0.2587,
+      "step": 22460
+    },
+    {
+      "epoch": 0.8398719536416104,
+      "grad_norm": 0.4997739791870117,
+      "learning_rate": 1.2388124110584255e-06,
+      "loss": 0.2811,
+      "step": 22465
+    },
+    {
+      "epoch": 0.8400588826319602,
+      "grad_norm": 0.5061297416687012,
+      "learning_rate": 1.2359827809616843e-06,
+      "loss": 0.2929,
+      "step": 22470
+    },
+    {
+      "epoch": 0.8402458116223099,
+      "grad_norm": 0.5074501037597656,
+      "learning_rate": 1.2331561733233644e-06,
+      "loss": 0.3264,
+      "step": 22475
+    },
+    {
+      "epoch": 0.8404327406126597,
+      "grad_norm": 0.7985924482345581,
+      "learning_rate": 1.230332589118276e-06,
+      "loss": 0.3159,
+      "step": 22480
+    },
+    {
+      "epoch": 0.8406196696030096,
+      "grad_norm": 0.39413031935691833,
+      "learning_rate": 1.2275120293201969e-06,
+      "loss": 0.2703,
+      "step": 22485
+    },
+    {
+      "epoch": 0.8408065985933594,
+      "grad_norm": 0.5623971819877625,
+      "learning_rate": 1.2246944949018525e-06,
+      "loss": 0.3326,
+      "step": 22490
+    },
+    {
+      "epoch": 0.8409935275837092,
+      "grad_norm": 0.3700483739376068,
+      "learning_rate": 1.2218799868349362e-06,
+      "loss": 0.3001,
+      "step": 22495
+    },
+    {
+      "epoch": 0.8411804565740589,
+      "grad_norm": 0.42255735397338867,
+      "learning_rate": 1.2190685060900843e-06,
+      "loss": 0.3042,
+      "step": 22500
+    },
+    {
+      "epoch": 0.8413673855644087,
+      "grad_norm": 0.38053280115127563,
+      "learning_rate": 1.2162600536369018e-06,
+      "loss": 0.2707,
+      "step": 22505
+    },
+    {
+      "epoch": 0.8415543145547585,
+      "grad_norm": 0.42679354548454285,
+      "learning_rate": 1.2134546304439398e-06,
+      "loss": 0.2857,
+      "step": 22510
+    },
+    {
+      "epoch": 0.8417412435451083,
+      "grad_norm": 0.20769809186458588,
+      "learning_rate": 1.2106522374787078e-06,
+      "loss": 0.3186,
+      "step": 22515
+    },
+    {
+      "epoch": 0.841928172535458,
+      "grad_norm": 0.32969167828559875,
+      "learning_rate": 1.2078528757076746e-06,
+      "loss": 0.3293,
+      "step": 22520
+    },
+    {
+      "epoch": 0.8421151015258079,
+      "grad_norm": 0.40005597472190857,
+      "learning_rate": 1.205056546096256e-06,
+      "loss": 0.2552,
+      "step": 22525
+    },
+    {
+      "epoch": 0.8423020305161577,
+      "grad_norm": 0.500026524066925,
+      "learning_rate": 1.2022632496088294e-06,
+      "loss": 0.3076,
+      "step": 22530
+    },
+    {
+      "epoch": 0.8424889595065075,
+      "grad_norm": 0.34214553236961365,
+      "learning_rate": 1.1994729872087185e-06,
+      "loss": 0.2829,
+      "step": 22535
+    },
+    {
+      "epoch": 0.8426758884968573,
+      "grad_norm": 0.35944414138793945,
+      "learning_rate": 1.1966857598582104e-06,
+      "loss": 0.2663,
+      "step": 22540
+    },
+    {
+      "epoch": 0.842862817487207,
+      "grad_norm": 0.8880605101585388,
+      "learning_rate": 1.193901568518534e-06,
+      "loss": 0.3047,
+      "step": 22545
+    },
+    {
+      "epoch": 0.8430497464775568,
+      "grad_norm": 0.316691517829895,
+      "learning_rate": 1.1911204141498821e-06,
+      "loss": 0.2996,
+      "step": 22550
+    },
+    {
+      "epoch": 0.8432366754679066,
+      "grad_norm": 0.49218663573265076,
+      "learning_rate": 1.1883422977113935e-06,
+      "loss": 0.2315,
+      "step": 22555
+    },
+    {
+      "epoch": 0.8434236044582564,
+      "grad_norm": 0.27020570635795593,
+      "learning_rate": 1.1855672201611578e-06,
+      "loss": 0.2329,
+      "step": 22560
+    },
+    {
+      "epoch": 0.8436105334486063,
+      "grad_norm": 0.321402907371521,
+      "learning_rate": 1.1827951824562245e-06,
+      "loss": 0.3207,
+      "step": 22565
+    },
+    {
+      "epoch": 0.843797462438956,
+      "grad_norm": 0.8666585087776184,
+      "learning_rate": 1.1800261855525862e-06,
+      "loss": 0.2414,
+      "step": 22570
+    },
+    {
+      "epoch": 0.8439843914293058,
+      "grad_norm": 0.41892504692077637,
+      "learning_rate": 1.177260230405194e-06,
+      "loss": 0.2862,
+      "step": 22575
+    },
+    {
+      "epoch": 0.8441713204196556,
+      "grad_norm": 0.4117008149623871,
+      "learning_rate": 1.1744973179679431e-06,
+      "loss": 0.311,
+      "step": 22580
+    },
+    {
+      "epoch": 0.8443582494100054,
+      "grad_norm": 0.2825338840484619,
+      "learning_rate": 1.171737449193686e-06,
+      "loss": 0.3535,
+      "step": 22585
+    },
+    {
+      "epoch": 0.8445451784003551,
+      "grad_norm": 0.5039531588554382,
+      "learning_rate": 1.1689806250342196e-06,
+      "loss": 0.266,
+      "step": 22590
+    },
+    {
+      "epoch": 0.8447321073907049,
+      "grad_norm": 0.8699703216552734,
+      "learning_rate": 1.166226846440297e-06,
+      "loss": 0.2462,
+      "step": 22595
+    },
+    {
+      "epoch": 0.8449190363810547,
+      "grad_norm": 0.4708864986896515,
+      "learning_rate": 1.1634761143616159e-06,
+      "loss": 0.2971,
+      "step": 22600
+    },
+    {
+      "epoch": 0.8451059653714046,
+      "grad_norm": 0.34658095240592957,
+      "learning_rate": 1.1607284297468214e-06,
+      "loss": 0.2854,
+      "step": 22605
+    },
+    {
+      "epoch": 0.8452928943617544,
+      "grad_norm": 0.2925431430339813,
+      "learning_rate": 1.1579837935435168e-06,
+      "loss": 0.2688,
+      "step": 22610
+    },
+    {
+      "epoch": 0.8454798233521041,
+      "grad_norm": 0.48946163058280945,
+      "learning_rate": 1.1552422066982437e-06,
+      "loss": 0.2699,
+      "step": 22615
+    },
+    {
+      "epoch": 0.8456667523424539,
+      "grad_norm": 0.2582431733608246,
+      "learning_rate": 1.1525036701565018e-06,
+      "loss": 0.2689,
+      "step": 22620
+    },
+    {
+      "epoch": 0.8458536813328037,
+      "grad_norm": 0.26818692684173584,
+      "learning_rate": 1.1497681848627284e-06,
+      "loss": 0.2724,
+      "step": 22625
+    },
+    {
+      "epoch": 0.8460406103231535,
+      "grad_norm": 0.26108282804489136,
+      "learning_rate": 1.1470357517603192e-06,
+      "loss": 0.2324,
+      "step": 22630
+    },
+    {
+      "epoch": 0.8462275393135033,
+      "grad_norm": 0.3785153925418854,
+      "learning_rate": 1.1443063717916081e-06,
+      "loss": 0.2595,
+      "step": 22635
+    },
+    {
+      "epoch": 0.846414468303853,
+      "grad_norm": 0.4359034597873688,
+      "learning_rate": 1.141580045897881e-06,
+      "loss": 0.2702,
+      "step": 22640
+    },
+    {
+      "epoch": 0.8466013972942029,
+      "grad_norm": 0.41941553354263306,
+      "learning_rate": 1.1388567750193725e-06,
+      "loss": 0.2944,
+      "step": 22645
+    },
+    {
+      "epoch": 0.8467883262845527,
+      "grad_norm": 0.2629951536655426,
+      "learning_rate": 1.1361365600952589e-06,
+      "loss": 0.2684,
+      "step": 22650
+    },
+    {
+      "epoch": 0.8469752552749025,
+      "grad_norm": 0.32962939143180847,
+      "learning_rate": 1.1334194020636635e-06,
+      "loss": 0.3649,
+      "step": 22655
+    },
+    {
+      "epoch": 0.8471621842652522,
+      "grad_norm": 0.6077668070793152,
+      "learning_rate": 1.1307053018616543e-06,
+      "loss": 0.2784,
+      "step": 22660
+    },
+    {
+      "epoch": 0.847349113255602,
+      "grad_norm": 0.2332148402929306,
+      "learning_rate": 1.1279942604252514e-06,
+      "loss": 0.2693,
+      "step": 22665
+    },
+    {
+      "epoch": 0.8475360422459518,
+      "grad_norm": 0.3998556435108185,
+      "learning_rate": 1.1252862786894103e-06,
+      "loss": 0.241,
+      "step": 22670
+    },
+    {
+      "epoch": 0.8477229712363016,
+      "grad_norm": 0.6394714117050171,
+      "learning_rate": 1.1225813575880417e-06,
+      "loss": 0.3319,
+      "step": 22675
+    },
+    {
+      "epoch": 0.8479099002266514,
+      "grad_norm": 0.271589457988739,
+      "learning_rate": 1.1198794980539908e-06,
+      "loss": 0.2994,
+      "step": 22680
+    },
+    {
+      "epoch": 0.8480968292170012,
+      "grad_norm": 0.35430166125297546,
+      "learning_rate": 1.1171807010190528e-06,
+      "loss": 0.2702,
+      "step": 22685
+    },
+    {
+      "epoch": 0.848283758207351,
+      "grad_norm": 0.5060386061668396,
+      "learning_rate": 1.114484967413969e-06,
+      "loss": 0.3112,
+      "step": 22690
+    },
+    {
+      "epoch": 0.8484706871977008,
+      "grad_norm": 0.3983868956565857,
+      "learning_rate": 1.1117922981684172e-06,
+      "loss": 0.2696,
+      "step": 22695
+    },
+    {
+      "epoch": 0.8486576161880506,
+      "grad_norm": 0.569534957408905,
+      "learning_rate": 1.1091026942110217e-06,
+      "loss": 0.2857,
+      "step": 22700
+    },
+    {
+      "epoch": 0.8488445451784004,
+      "grad_norm": 0.9335552453994751,
+      "learning_rate": 1.1064161564693486e-06,
+      "loss": 0.2731,
+      "step": 22705
+    },
+    {
+      "epoch": 0.8490314741687501,
+      "grad_norm": 0.36156487464904785,
+      "learning_rate": 1.1037326858699126e-06,
+      "loss": 0.2699,
+      "step": 22710
+    },
+    {
+      "epoch": 0.8492184031590999,
+      "grad_norm": 0.31813082098960876,
+      "learning_rate": 1.101052283338161e-06,
+      "loss": 0.199,
+      "step": 22715
+    },
+    {
+      "epoch": 0.8494053321494497,
+      "grad_norm": 0.42129233479499817,
+      "learning_rate": 1.0983749497984908e-06,
+      "loss": 0.2301,
+      "step": 22720
+    },
+    {
+      "epoch": 0.8495922611397995,
+      "grad_norm": 0.39161425828933716,
+      "learning_rate": 1.0957006861742348e-06,
+      "loss": 0.2858,
+      "step": 22725
+    },
+    {
+      "epoch": 0.8497791901301494,
+      "grad_norm": 0.3193294107913971,
+      "learning_rate": 1.093029493387673e-06,
+      "loss": 0.2037,
+      "step": 22730
+    },
+    {
+      "epoch": 0.8499661191204991,
+      "grad_norm": 0.3217445909976959,
+      "learning_rate": 1.0903613723600225e-06,
+      "loss": 0.288,
+      "step": 22735
+    },
+    {
+      "epoch": 0.8501530481108489,
+      "grad_norm": 0.3758181035518646,
+      "learning_rate": 1.0876963240114413e-06,
+      "loss": 0.1978,
+      "step": 22740
+    },
+    {
+      "epoch": 0.8503399771011987,
+      "grad_norm": 0.4037764370441437,
+      "learning_rate": 1.0850343492610282e-06,
+      "loss": 0.2588,
+      "step": 22745
+    },
+    {
+      "epoch": 0.8505269060915485,
+      "grad_norm": 0.655259907245636,
+      "learning_rate": 1.0823754490268213e-06,
+      "loss": 0.2726,
+      "step": 22750
+    },
+    {
+      "epoch": 0.8507138350818982,
+      "grad_norm": 0.20327606797218323,
+      "learning_rate": 1.0797196242258002e-06,
+      "loss": 0.2396,
+      "step": 22755
+    },
+    {
+      "epoch": 0.850900764072248,
+      "grad_norm": 0.3590863049030304,
+      "learning_rate": 1.0770668757738811e-06,
+      "loss": 0.2711,
+      "step": 22760
+    },
+    {
+      "epoch": 0.8510876930625978,
+      "grad_norm": 0.5582608580589294,
+      "learning_rate": 1.0744172045859236e-06,
+      "loss": 0.2555,
+      "step": 22765
+    },
+    {
+      "epoch": 0.8512746220529477,
+      "grad_norm": 0.9252792596817017,
+      "learning_rate": 1.071770611575721e-06,
+      "loss": 0.3225,
+      "step": 22770
+    },
+    {
+      "epoch": 0.8514615510432975,
+      "grad_norm": 0.48875078558921814,
+      "learning_rate": 1.0691270976560075e-06,
+      "loss": 0.2456,
+      "step": 22775
+    },
+    {
+      "epoch": 0.8516484800336472,
+      "grad_norm": 0.691635251045227,
+      "learning_rate": 1.0664866637384574e-06,
+      "loss": 0.3181,
+      "step": 22780
+    },
+    {
+      "epoch": 0.851835409023997,
+      "grad_norm": 0.7403832077980042,
+      "learning_rate": 1.0638493107336811e-06,
+      "loss": 0.2726,
+      "step": 22785
+    },
+    {
+      "epoch": 0.8520223380143468,
+      "grad_norm": 0.37632712721824646,
+      "learning_rate": 1.0612150395512233e-06,
+      "loss": 0.315,
+      "step": 22790
+    },
+    {
+      "epoch": 0.8522092670046966,
+      "grad_norm": 0.3026890158653259,
+      "learning_rate": 1.0585838510995684e-06,
+      "loss": 0.2889,
+      "step": 22795
+    },
+    {
+      "epoch": 0.8523961959950463,
+      "grad_norm": 0.516616702079773,
+      "learning_rate": 1.0559557462861403e-06,
+      "loss": 0.2493,
+      "step": 22800
+    },
+    {
+      "epoch": 0.8525831249853961,
+      "grad_norm": 0.3008388876914978,
+      "learning_rate": 1.0533307260172954e-06,
+      "loss": 0.2922,
+      "step": 22805
+    },
+    {
+      "epoch": 0.852770053975746,
+      "grad_norm": 0.4109887182712555,
+      "learning_rate": 1.0507087911983293e-06,
+      "loss": 0.2831,
+      "step": 22810
+    },
+    {
+      "epoch": 0.8529569829660958,
+      "grad_norm": 0.3522479832172394,
+      "learning_rate": 1.04808994273347e-06,
+      "loss": 0.3247,
+      "step": 22815
+    },
+    {
+      "epoch": 0.8531439119564456,
+      "grad_norm": 0.5156667232513428,
+      "learning_rate": 1.045474181525885e-06,
+      "loss": 0.2889,
+      "step": 22820
+    },
+    {
+      "epoch": 0.8533308409467953,
+      "grad_norm": 0.4093747138977051,
+      "learning_rate": 1.0428615084776772e-06,
+      "loss": 0.243,
+      "step": 22825
+    },
+    {
+      "epoch": 0.8535177699371451,
+      "grad_norm": 0.3295019567012787,
+      "learning_rate": 1.0402519244898778e-06,
+      "loss": 0.2638,
+      "step": 22830
+    },
+    {
+      "epoch": 0.8537046989274949,
+      "grad_norm": 0.7684919238090515,
+      "learning_rate": 1.037645430462464e-06,
+      "loss": 0.2124,
+      "step": 22835
+    },
+    {
+      "epoch": 0.8538916279178447,
+      "grad_norm": 0.43423688411712646,
+      "learning_rate": 1.0350420272943362e-06,
+      "loss": 0.2397,
+      "step": 22840
+    },
+    {
+      "epoch": 0.8540785569081945,
+      "grad_norm": 0.6224250197410583,
+      "learning_rate": 1.0324417158833343e-06,
+      "loss": 0.2438,
+      "step": 22845
+    },
+    {
+      "epoch": 0.8542654858985443,
+      "grad_norm": 0.35096117854118347,
+      "learning_rate": 1.02984449712623e-06,
+      "loss": 0.2792,
+      "step": 22850
+    },
+    {
+      "epoch": 0.8544524148888941,
+      "grad_norm": 0.3952169716358185,
+      "learning_rate": 1.0272503719187332e-06,
+      "loss": 0.3134,
+      "step": 22855
+    },
+    {
+      "epoch": 0.8546393438792439,
+      "grad_norm": 0.18396589159965515,
+      "learning_rate": 1.0246593411554796e-06,
+      "loss": 0.2938,
+      "step": 22860
+    },
+    {
+      "epoch": 0.8548262728695937,
+      "grad_norm": 0.35608750581741333,
+      "learning_rate": 1.022071405730043e-06,
+      "loss": 0.2834,
+      "step": 22865
+    },
+    {
+      "epoch": 0.8550132018599435,
+      "grad_norm": 0.45381414890289307,
+      "learning_rate": 1.0194865665349296e-06,
+      "loss": 0.3398,
+      "step": 22870
+    },
+    {
+      "epoch": 0.8552001308502932,
+      "grad_norm": 0.5310588479042053,
+      "learning_rate": 1.0169048244615742e-06,
+      "loss": 0.4079,
+      "step": 22875
+    },
+    {
+      "epoch": 0.855387059840643,
+      "grad_norm": 0.2279624491930008,
+      "learning_rate": 1.0143261804003479e-06,
+      "loss": 0.2207,
+      "step": 22880
+    },
+    {
+      "epoch": 0.8555739888309928,
+      "grad_norm": 0.408582478761673,
+      "learning_rate": 1.0117506352405503e-06,
+      "loss": 0.2455,
+      "step": 22885
+    },
+    {
+      "epoch": 0.8557609178213427,
+      "grad_norm": 0.4538606107234955,
+      "learning_rate": 1.0091781898704123e-06,
+      "loss": 0.2467,
+      "step": 22890
+    },
+    {
+      "epoch": 0.8559478468116924,
+      "grad_norm": 0.36869016289711,
+      "learning_rate": 1.006608845177095e-06,
+      "loss": 0.298,
+      "step": 22895
+    },
+    {
+      "epoch": 0.8561347758020422,
+      "grad_norm": 0.3675679564476013,
+      "learning_rate": 1.0040426020466965e-06,
+      "loss": 0.2797,
+      "step": 22900
+    },
+    {
+      "epoch": 0.856321704792392,
+      "grad_norm": 0.5693662762641907,
+      "learning_rate": 1.0014794613642354e-06,
+      "loss": 0.2798,
+      "step": 22905
+    },
+    {
+      "epoch": 0.8565086337827418,
+      "grad_norm": 0.35014966130256653,
+      "learning_rate": 9.98919424013669e-07,
+      "loss": 0.2168,
+      "step": 22910
+    },
+    {
+      "epoch": 0.8566955627730916,
+      "grad_norm": 0.30344653129577637,
+      "learning_rate": 9.963624908778791e-07,
+      "loss": 0.2438,
+      "step": 22915
+    },
+    {
+      "epoch": 0.8568824917634413,
+      "grad_norm": 0.22757522761821747,
+      "learning_rate": 9.938086628386778e-07,
+      "loss": 0.2447,
+      "step": 22920
+    },
+    {
+      "epoch": 0.8570694207537911,
+      "grad_norm": 0.5425657629966736,
+      "learning_rate": 9.912579407768118e-07,
+      "loss": 0.2094,
+      "step": 22925
+    },
+    {
+      "epoch": 0.857256349744141,
+      "grad_norm": 0.6510607600212097,
+      "learning_rate": 9.887103255719489e-07,
+      "loss": 0.3355,
+      "step": 22930
+    },
+    {
+      "epoch": 0.8574432787344908,
+      "grad_norm": 0.4659566283226013,
+      "learning_rate": 9.861658181026878e-07,
+      "loss": 0.2521,
+      "step": 22935
+    },
+    {
+      "epoch": 0.8576302077248406,
+      "grad_norm": 0.40870416164398193,
+      "learning_rate": 9.83624419246555e-07,
+      "loss": 0.3023,
+      "step": 22940
+    },
+    {
+      "epoch": 0.8578171367151903,
+      "grad_norm": 0.24827082455158234,
+      "learning_rate": 9.810861298800111e-07,
+      "loss": 0.2788,
+      "step": 22945
+    },
+    {
+      "epoch": 0.8580040657055401,
+      "grad_norm": 0.6757010221481323,
+      "learning_rate": 9.78550950878433e-07,
+      "loss": 0.3783,
+      "step": 22950
+    },
+    {
+      "epoch": 0.8581909946958899,
+      "grad_norm": 0.3025006651878357,
+      "learning_rate": 9.760188831161376e-07,
+      "loss": 0.3317,
+      "step": 22955
+    },
+    {
+      "epoch": 0.8583779236862397,
+      "grad_norm": 0.45728322863578796,
+      "learning_rate": 9.734899274663578e-07,
+      "loss": 0.3688,
+      "step": 22960
+    },
+    {
+      "epoch": 0.8585648526765894,
+      "grad_norm": 0.5989465713500977,
+      "learning_rate": 9.709640848012602e-07,
+      "loss": 0.2258,
+      "step": 22965
+    },
+    {
+      "epoch": 0.8587517816669392,
+      "grad_norm": 0.2060554027557373,
+      "learning_rate": 9.684413559919358e-07,
+      "loss": 0.3065,
+      "step": 22970
+    },
+    {
+      "epoch": 0.8589387106572891,
+      "grad_norm": 0.3968755006790161,
+      "learning_rate": 9.65921741908402e-07,
+      "loss": 0.3119,
+      "step": 22975
+    },
+    {
+      "epoch": 0.8591256396476389,
+      "grad_norm": 0.4152187407016754,
+      "learning_rate": 9.634052434195983e-07,
+      "loss": 0.2567,
+      "step": 22980
+    },
+    {
+      "epoch": 0.8593125686379887,
+      "grad_norm": 0.34883493185043335,
+      "learning_rate": 9.60891861393396e-07,
+      "loss": 0.2796,
+      "step": 22985
+    },
+    {
+      "epoch": 0.8594994976283384,
+      "grad_norm": 0.13147194683551788,
+      "learning_rate": 9.583815966965882e-07,
+      "loss": 0.2369,
+      "step": 22990
+    },
+    {
+      "epoch": 0.8596864266186882,
+      "grad_norm": 0.49323028326034546,
+      "learning_rate": 9.558744501948903e-07,
+      "loss": 0.3564,
+      "step": 22995
+    },
+    {
+      "epoch": 0.859873355609038,
+      "grad_norm": 0.23689912259578705,
+      "learning_rate": 9.533704227529494e-07,
+      "loss": 0.2567,
+      "step": 23000
+    },
+    {
+      "epoch": 0.8600602845993878,
+      "grad_norm": 0.8092080950737,
+      "learning_rate": 9.508695152343295e-07,
+      "loss": 0.3302,
+      "step": 23005
+    },
+    {
+      "epoch": 0.8602472135897375,
+      "grad_norm": 0.3174912929534912,
+      "learning_rate": 9.483717285015237e-07,
+      "loss": 0.245,
+      "step": 23010
+    },
+    {
+      "epoch": 0.8604341425800874,
+      "grad_norm": 0.7012712955474854,
+      "learning_rate": 9.458770634159475e-07,
+      "loss": 0.3679,
+      "step": 23015
+    },
+    {
+      "epoch": 0.8606210715704372,
+      "grad_norm": 0.21312133967876434,
+      "learning_rate": 9.433855208379383e-07,
+      "loss": 0.2804,
+      "step": 23020
+    },
+    {
+      "epoch": 0.860808000560787,
+      "grad_norm": 0.3727993667125702,
+      "learning_rate": 9.408971016267588e-07,
+      "loss": 0.4381,
+      "step": 23025
+    },
+    {
+      "epoch": 0.8609949295511368,
+      "grad_norm": 0.3259866237640381,
+      "learning_rate": 9.384118066405934e-07,
+      "loss": 0.315,
+      "step": 23030
+    },
+    {
+      "epoch": 0.8611818585414865,
+      "grad_norm": 0.48218241333961487,
+      "learning_rate": 9.359296367365478e-07,
+      "loss": 0.3288,
+      "step": 23035
+    },
+    {
+      "epoch": 0.8613687875318363,
+      "grad_norm": 0.2950020134449005,
+      "learning_rate": 9.334505927706516e-07,
+      "loss": 0.3175,
+      "step": 23040
+    },
+    {
+      "epoch": 0.8615557165221861,
+      "grad_norm": 0.36184942722320557,
+      "learning_rate": 9.309746755978566e-07,
+      "loss": 0.2991,
+      "step": 23045
+    },
+    {
+      "epoch": 0.8617426455125359,
+      "grad_norm": 0.7623559236526489,
+      "learning_rate": 9.285018860720352e-07,
+      "loss": 0.2409,
+      "step": 23050
+    },
+    {
+      "epoch": 0.8619295745028858,
+      "grad_norm": 0.10221162438392639,
+      "learning_rate": 9.260322250459808e-07,
+      "loss": 0.2914,
+      "step": 23055
+    },
+    {
+      "epoch": 0.8621165034932355,
+      "grad_norm": 0.4181196093559265,
+      "learning_rate": 9.235656933714121e-07,
+      "loss": 0.3329,
+      "step": 23060
+    },
+    {
+      "epoch": 0.8623034324835853,
+      "grad_norm": 0.5724478960037231,
+      "learning_rate": 9.21102291898962e-07,
+      "loss": 0.3349,
+      "step": 23065
+    },
+    {
+      "epoch": 0.8624903614739351,
+      "grad_norm": 0.3799975514411926,
+      "learning_rate": 9.186420214781888e-07,
+      "loss": 0.3123,
+      "step": 23070
+    },
+    {
+      "epoch": 0.8626772904642849,
+      "grad_norm": 0.48635974526405334,
+      "learning_rate": 9.161848829575693e-07,
+      "loss": 0.2522,
+      "step": 23075
+    },
+    {
+      "epoch": 0.8628642194546347,
+      "grad_norm": 0.3440147340297699,
+      "learning_rate": 9.137308771844988e-07,
+      "loss": 0.2328,
+      "step": 23080
+    },
+    {
+      "epoch": 0.8630511484449844,
+      "grad_norm": 0.6380581259727478,
+      "learning_rate": 9.112800050052927e-07,
+      "loss": 0.3319,
+      "step": 23085
+    },
+    {
+      "epoch": 0.8632380774353342,
+      "grad_norm": 0.37005263566970825,
+      "learning_rate": 9.088322672651906e-07,
+      "loss": 0.3111,
+      "step": 23090
+    },
+    {
+      "epoch": 0.8634250064256841,
+      "grad_norm": 0.37442389130592346,
+      "learning_rate": 9.063876648083414e-07,
+      "loss": 0.2227,
+      "step": 23095
+    },
+    {
+      "epoch": 0.8636119354160339,
+      "grad_norm": 0.7940928339958191,
+      "learning_rate": 9.039461984778231e-07,
+      "loss": 0.3284,
+      "step": 23100
+    },
+    {
+      "epoch": 0.8637988644063836,
+      "grad_norm": 0.4885554015636444,
+      "learning_rate": 9.015078691156265e-07,
+      "loss": 0.2938,
+      "step": 23105
+    },
+    {
+      "epoch": 0.8639857933967334,
+      "grad_norm": 0.35439759492874146,
+      "learning_rate": 8.990726775626602e-07,
+      "loss": 0.2331,
+      "step": 23110
+    },
+    {
+      "epoch": 0.8641727223870832,
+      "grad_norm": 0.4514514207839966,
+      "learning_rate": 8.966406246587545e-07,
+      "loss": 0.2141,
+      "step": 23115
+    },
+    {
+      "epoch": 0.864359651377433,
+      "grad_norm": 0.5355989336967468,
+      "learning_rate": 8.942117112426529e-07,
+      "loss": 0.2842,
+      "step": 23120
+    },
+    {
+      "epoch": 0.8645465803677828,
+      "grad_norm": 2.2368109226226807,
+      "learning_rate": 8.917859381520189e-07,
+      "loss": 0.2435,
+      "step": 23125
+    },
+    {
+      "epoch": 0.8647335093581325,
+      "grad_norm": 0.5107064843177795,
+      "learning_rate": 8.893633062234285e-07,
+      "loss": 0.2857,
+      "step": 23130
+    },
+    {
+      "epoch": 0.8649204383484824,
+      "grad_norm": 0.4561639726161957,
+      "learning_rate": 8.869438162923838e-07,
+      "loss": 0.247,
+      "step": 23135
+    },
+    {
+      "epoch": 0.8651073673388322,
+      "grad_norm": 0.40335312485694885,
+      "learning_rate": 8.845274691932926e-07,
+      "loss": 0.3122,
+      "step": 23140
+    },
+    {
+      "epoch": 0.865294296329182,
+      "grad_norm": 0.25808554887771606,
+      "learning_rate": 8.821142657594861e-07,
+      "loss": 0.2516,
+      "step": 23145
+    },
+    {
+      "epoch": 0.8654812253195318,
+      "grad_norm": 0.4519065320491791,
+      "learning_rate": 8.797042068232098e-07,
+      "loss": 0.2665,
+      "step": 23150
+    },
+    {
+      "epoch": 0.8656681543098815,
+      "grad_norm": 0.7045873403549194,
+      "learning_rate": 8.772972932156221e-07,
+      "loss": 0.4179,
+      "step": 23155
+    },
+    {
+      "epoch": 0.8658550833002313,
+      "grad_norm": 0.4974367618560791,
+      "learning_rate": 8.748935257668012e-07,
+      "loss": 0.3216,
+      "step": 23160
+    },
+    {
+      "epoch": 0.8660420122905811,
+      "grad_norm": 0.4620605707168579,
+      "learning_rate": 8.72492905305734e-07,
+      "loss": 0.2646,
+      "step": 23165
+    },
+    {
+      "epoch": 0.8662289412809309,
+      "grad_norm": 0.45460769534111023,
+      "learning_rate": 8.700954326603295e-07,
+      "loss": 0.3191,
+      "step": 23170
+    },
+    {
+      "epoch": 0.8664158702712808,
+      "grad_norm": 0.484361857175827,
+      "learning_rate": 8.67701108657405e-07,
+      "loss": 0.2041,
+      "step": 23175
+    },
+    {
+      "epoch": 0.8666027992616305,
+      "grad_norm": 0.8345504999160767,
+      "learning_rate": 8.653099341226956e-07,
+      "loss": 0.382,
+      "step": 23180
+    },
+    {
+      "epoch": 0.8667897282519803,
+      "grad_norm": 0.5114479660987854,
+      "learning_rate": 8.62921909880845e-07,
+      "loss": 0.2313,
+      "step": 23185
+    },
+    {
+      "epoch": 0.8669766572423301,
+      "grad_norm": 0.35048356652259827,
+      "learning_rate": 8.605370367554178e-07,
+      "loss": 0.2199,
+      "step": 23190
+    },
+    {
+      "epoch": 0.8671635862326799,
+      "grad_norm": 0.5474756360054016,
+      "learning_rate": 8.581553155688894e-07,
+      "loss": 0.3222,
+      "step": 23195
+    },
+    {
+      "epoch": 0.8673505152230296,
+      "grad_norm": 0.4555470645427704,
+      "learning_rate": 8.557767471426448e-07,
+      "loss": 0.2262,
+      "step": 23200
+    },
+    {
+      "epoch": 0.8675374442133794,
+      "grad_norm": 0.5740458965301514,
+      "learning_rate": 8.534013322969859e-07,
+      "loss": 0.3478,
+      "step": 23205
+    },
+    {
+      "epoch": 0.8677243732037292,
+      "grad_norm": 0.4381164014339447,
+      "learning_rate": 8.510290718511227e-07,
+      "loss": 0.3054,
+      "step": 23210
+    },
+    {
+      "epoch": 0.867911302194079,
+      "grad_norm": 0.27854111790657043,
+      "learning_rate": 8.486599666231832e-07,
+      "loss": 0.2728,
+      "step": 23215
+    },
+    {
+      "epoch": 0.8680982311844289,
+      "grad_norm": 0.5790026187896729,
+      "learning_rate": 8.462940174302026e-07,
+      "loss": 0.3049,
+      "step": 23220
+    },
+    {
+      "epoch": 0.8682851601747786,
+      "grad_norm": 0.18455126881599426,
+      "learning_rate": 8.439312250881282e-07,
+      "loss": 0.266,
+      "step": 23225
+    },
+    {
+      "epoch": 0.8684720891651284,
+      "grad_norm": 0.4800267517566681,
+      "learning_rate": 8.415715904118171e-07,
+      "loss": 0.3171,
+      "step": 23230
+    },
+    {
+      "epoch": 0.8686590181554782,
+      "grad_norm": 0.36994990706443787,
+      "learning_rate": 8.392151142150428e-07,
+      "loss": 0.3076,
+      "step": 23235
+    },
+    {
+      "epoch": 0.868845947145828,
+      "grad_norm": 0.7822495698928833,
+      "learning_rate": 8.368617973104887e-07,
+      "loss": 0.2798,
+      "step": 23240
+    },
+    {
+      "epoch": 0.8690328761361777,
+      "grad_norm": 0.38529595732688904,
+      "learning_rate": 8.345116405097408e-07,
+      "loss": 0.321,
+      "step": 23245
+    },
+    {
+      "epoch": 0.8692198051265275,
+      "grad_norm": 0.5859056115150452,
+      "learning_rate": 8.321646446233056e-07,
+      "loss": 0.2563,
+      "step": 23250
+    },
+    {
+      "epoch": 0.8694067341168773,
+      "grad_norm": 0.5239048004150391,
+      "learning_rate": 8.29820810460591e-07,
+      "loss": 0.2624,
+      "step": 23255
+    },
+    {
+      "epoch": 0.8695936631072272,
+      "grad_norm": 0.5495345592498779,
+      "learning_rate": 8.274801388299225e-07,
+      "loss": 0.2581,
+      "step": 23260
+    },
+    {
+      "epoch": 0.869780592097577,
+      "grad_norm": 0.4601067900657654,
+      "learning_rate": 8.251426305385268e-07,
+      "loss": 0.2837,
+      "step": 23265
+    },
+    {
+      "epoch": 0.8699675210879267,
+      "grad_norm": 0.37067458033561707,
+      "learning_rate": 8.228082863925457e-07,
+      "loss": 0.2406,
+      "step": 23270
+    },
+    {
+      "epoch": 0.8701544500782765,
+      "grad_norm": 0.8990246653556824,
+      "learning_rate": 8.204771071970253e-07,
+      "loss": 0.303,
+      "step": 23275
+    },
+    {
+      "epoch": 0.8703413790686263,
+      "grad_norm": 0.23729777336120605,
+      "learning_rate": 8.181490937559234e-07,
+      "loss": 0.323,
+      "step": 23280
+    },
+    {
+      "epoch": 0.8705283080589761,
+      "grad_norm": 0.6693614721298218,
+      "learning_rate": 8.158242468721078e-07,
+      "loss": 0.277,
+      "step": 23285
+    },
+    {
+      "epoch": 0.8707152370493259,
+      "grad_norm": 0.7663869857788086,
+      "learning_rate": 8.135025673473474e-07,
+      "loss": 0.303,
+      "step": 23290
+    },
+    {
+      "epoch": 0.8709021660396756,
+      "grad_norm": 0.389027863740921,
+      "learning_rate": 8.11184055982327e-07,
+      "loss": 0.2435,
+      "step": 23295
+    },
+    {
+      "epoch": 0.8710890950300255,
+      "grad_norm": 0.38779979944229126,
+      "learning_rate": 8.088687135766316e-07,
+      "loss": 0.3011,
+      "step": 23300
+    },
+    {
+      "epoch": 0.8712760240203753,
+      "grad_norm": 0.33862584829330444,
+      "learning_rate": 8.06556540928759e-07,
+      "loss": 0.304,
+      "step": 23305
+    },
+    {
+      "epoch": 0.8714629530107251,
+      "grad_norm": 0.31352442502975464,
+      "learning_rate": 8.042475388361104e-07,
+      "loss": 0.2957,
+      "step": 23310
+    },
+    {
+      "epoch": 0.8716498820010748,
+      "grad_norm": 0.4300394058227539,
+      "learning_rate": 8.019417080949932e-07,
+      "loss": 0.2494,
+      "step": 23315
+    },
+    {
+      "epoch": 0.8718368109914246,
+      "grad_norm": 0.4149588942527771,
+      "learning_rate": 7.996390495006223e-07,
+      "loss": 0.2895,
+      "step": 23320
+    },
+    {
+      "epoch": 0.8720237399817744,
+      "grad_norm": 0.3871593177318573,
+      "learning_rate": 7.973395638471182e-07,
+      "loss": 0.2378,
+      "step": 23325
+    },
+    {
+      "epoch": 0.8722106689721242,
+      "grad_norm": 0.4816226363182068,
+      "learning_rate": 7.95043251927512e-07,
+      "loss": 0.2898,
+      "step": 23330
+    },
+    {
+      "epoch": 0.872397597962474,
+      "grad_norm": 0.6048807501792908,
+      "learning_rate": 7.927501145337302e-07,
+      "loss": 0.2711,
+      "step": 23335
+    },
+    {
+      "epoch": 0.8725845269528238,
+      "grad_norm": 0.3330458700656891,
+      "learning_rate": 7.904601524566157e-07,
+      "loss": 0.278,
+      "step": 23340
+    },
+    {
+      "epoch": 0.8727714559431736,
+      "grad_norm": 0.5402839183807373,
+      "learning_rate": 7.881733664859048e-07,
+      "loss": 0.2908,
+      "step": 23345
+    },
+    {
+      "epoch": 0.8729583849335234,
+      "grad_norm": 0.3132112920284271,
+      "learning_rate": 7.858897574102508e-07,
+      "loss": 0.2525,
+      "step": 23350
+    },
+    {
+      "epoch": 0.8731453139238732,
+      "grad_norm": 0.5713052153587341,
+      "learning_rate": 7.836093260171995e-07,
+      "loss": 0.2595,
+      "step": 23355
+    },
+    {
+      "epoch": 0.873332242914223,
+      "grad_norm": 0.39907294511795044,
+      "learning_rate": 7.813320730932094e-07,
+      "loss": 0.2347,
+      "step": 23360
+    },
+    {
+      "epoch": 0.8735191719045727,
+      "grad_norm": 0.6258445978164673,
+      "learning_rate": 7.790579994236402e-07,
+      "loss": 0.3143,
+      "step": 23365
+    },
+    {
+      "epoch": 0.8737061008949225,
+      "grad_norm": 0.5387104153633118,
+      "learning_rate": 7.767871057927512e-07,
+      "loss": 0.2598,
+      "step": 23370
+    },
+    {
+      "epoch": 0.8738930298852723,
+      "grad_norm": 0.32804661989212036,
+      "learning_rate": 7.745193929837136e-07,
+      "loss": 0.2496,
+      "step": 23375
+    },
+    {
+      "epoch": 0.8740799588756222,
+      "grad_norm": 0.17273744940757751,
+      "learning_rate": 7.722548617785907e-07,
+      "loss": 0.3359,
+      "step": 23380
+    },
+    {
+      "epoch": 0.874266887865972,
+      "grad_norm": 0.2972503900527954,
+      "learning_rate": 7.699935129583602e-07,
+      "loss": 0.3159,
+      "step": 23385
+    },
+    {
+      "epoch": 0.8744538168563217,
+      "grad_norm": 0.49442118406295776,
+      "learning_rate": 7.677353473028926e-07,
+      "loss": 0.3058,
+      "step": 23390
+    },
+    {
+      "epoch": 0.8746407458466715,
+      "grad_norm": 0.5777555108070374,
+      "learning_rate": 7.654803655909671e-07,
+      "loss": 0.2823,
+      "step": 23395
+    },
+    {
+      "epoch": 0.8748276748370213,
+      "grad_norm": 0.549485445022583,
+      "learning_rate": 7.632285686002594e-07,
+      "loss": 0.2715,
+      "step": 23400
+    },
+    {
+      "epoch": 0.8750146038273711,
+      "grad_norm": 0.44350236654281616,
+      "learning_rate": 7.609799571073529e-07,
+      "loss": 0.2928,
+      "step": 23405
+    },
+    {
+      "epoch": 0.8752015328177208,
+      "grad_norm": 0.3184381425380707,
+      "learning_rate": 7.587345318877282e-07,
+      "loss": 0.3997,
+      "step": 23410
+    },
+    {
+      "epoch": 0.8753884618080706,
+      "grad_norm": 0.5490055084228516,
+      "learning_rate": 7.564922937157659e-07,
+      "loss": 0.2365,
+      "step": 23415
+    },
+    {
+      "epoch": 0.8755753907984205,
+      "grad_norm": 0.2554487884044647,
+      "learning_rate": 7.542532433647532e-07,
+      "loss": 0.2517,
+      "step": 23420
+    },
+    {
+      "epoch": 0.8757623197887703,
+      "grad_norm": 0.6824100613594055,
+      "learning_rate": 7.52017381606871e-07,
+      "loss": 0.3281,
+      "step": 23425
+    },
+    {
+      "epoch": 0.8759492487791201,
+      "grad_norm": 0.33024176955223083,
+      "learning_rate": 7.497847092132071e-07,
+      "loss": 0.2404,
+      "step": 23430
+    },
+    {
+      "epoch": 0.8761361777694698,
+      "grad_norm": 0.5682893395423889,
+      "learning_rate": 7.475552269537434e-07,
+      "loss": 0.3247,
+      "step": 23435
+    },
+    {
+      "epoch": 0.8763231067598196,
+      "grad_norm": 0.5878921747207642,
+      "learning_rate": 7.453289355973669e-07,
+      "loss": 0.2347,
+      "step": 23440
+    },
+    {
+      "epoch": 0.8765100357501694,
+      "grad_norm": 0.5948970913887024,
+      "learning_rate": 7.431058359118593e-07,
+      "loss": 0.223,
+      "step": 23445
+    },
+    {
+      "epoch": 0.8766969647405192,
+      "grad_norm": 0.4269593358039856,
+      "learning_rate": 7.408859286639069e-07,
+      "loss": 0.2496,
+      "step": 23450
+    },
+    {
+      "epoch": 0.876883893730869,
+      "grad_norm": 0.35104551911354065,
+      "learning_rate": 7.38669214619091e-07,
+      "loss": 0.2273,
+      "step": 23455
+    },
+    {
+      "epoch": 0.8770708227212187,
+      "grad_norm": 0.5351817607879639,
+      "learning_rate": 7.36455694541891e-07,
+      "loss": 0.2464,
+      "step": 23460
+    },
+    {
+      "epoch": 0.8772577517115686,
+      "grad_norm": 0.3460119962692261,
+      "learning_rate": 7.342453691956886e-07,
+      "loss": 0.286,
+      "step": 23465
+    },
+    {
+      "epoch": 0.8774446807019184,
+      "grad_norm": 0.25786739587783813,
+      "learning_rate": 7.320382393427595e-07,
+      "loss": 0.3703,
+      "step": 23470
+    },
+    {
+      "epoch": 0.8776316096922682,
+      "grad_norm": 0.47995084524154663,
+      "learning_rate": 7.298343057442825e-07,
+      "loss": 0.2853,
+      "step": 23475
+    },
+    {
+      "epoch": 0.8778185386826179,
+      "grad_norm": 0.31466034054756165,
+      "learning_rate": 7.276335691603276e-07,
+      "loss": 0.2484,
+      "step": 23480
+    },
+    {
+      "epoch": 0.8780054676729677,
+      "grad_norm": 0.42213770747184753,
+      "learning_rate": 7.254360303498697e-07,
+      "loss": 0.2988,
+      "step": 23485
+    },
+    {
+      "epoch": 0.8781923966633175,
+      "grad_norm": 0.5641787052154541,
+      "learning_rate": 7.232416900707739e-07,
+      "loss": 0.3115,
+      "step": 23490
+    },
+    {
+      "epoch": 0.8783793256536673,
+      "grad_norm": 0.3340662717819214,
+      "learning_rate": 7.210505490798081e-07,
+      "loss": 0.2782,
+      "step": 23495
+    },
+    {
+      "epoch": 0.878566254644017,
+      "grad_norm": 0.4800506830215454,
+      "learning_rate": 7.188626081326322e-07,
+      "loss": 0.2921,
+      "step": 23500
+    },
+    {
+      "epoch": 0.8787531836343669,
+      "grad_norm": 1.0108988285064697,
+      "learning_rate": 7.166778679838026e-07,
+      "loss": 0.2956,
+      "step": 23505
+    },
+    {
+      "epoch": 0.8789401126247167,
+      "grad_norm": 0.4938255846500397,
+      "learning_rate": 7.144963293867779e-07,
+      "loss": 0.3058,
+      "step": 23510
+    },
+    {
+      "epoch": 0.8791270416150665,
+      "grad_norm": 0.2628172039985657,
+      "learning_rate": 7.123179930939028e-07,
+      "loss": 0.3074,
+      "step": 23515
+    },
+    {
+      "epoch": 0.8793139706054163,
+      "grad_norm": 0.3008750081062317,
+      "learning_rate": 7.101428598564286e-07,
+      "loss": 0.2431,
+      "step": 23520
+    },
+    {
+      "epoch": 0.879500899595766,
+      "grad_norm": 0.5366243124008179,
+      "learning_rate": 7.07970930424493e-07,
+      "loss": 0.3393,
+      "step": 23525
+    },
+    {
+      "epoch": 0.8796878285861158,
+      "grad_norm": 0.3495098054409027,
+      "learning_rate": 7.058022055471337e-07,
+      "loss": 0.1941,
+      "step": 23530
+    },
+    {
+      "epoch": 0.8798747575764656,
+      "grad_norm": 0.5412205457687378,
+      "learning_rate": 7.036366859722798e-07,
+      "loss": 0.2646,
+      "step": 23535
+    },
+    {
+      "epoch": 0.8800616865668154,
+      "grad_norm": 0.4367334842681885,
+      "learning_rate": 7.014743724467609e-07,
+      "loss": 0.2841,
+      "step": 23540
+    },
+    {
+      "epoch": 0.8802486155571653,
+      "grad_norm": 0.2953815460205078,
+      "learning_rate": 6.993152657162916e-07,
+      "loss": 0.2756,
+      "step": 23545
+    },
+    {
+      "epoch": 0.880435544547515,
+      "grad_norm": 0.5892349481582642,
+      "learning_rate": 6.971593665254917e-07,
+      "loss": 0.2218,
+      "step": 23550
+    },
+    {
+      "epoch": 0.8806224735378648,
+      "grad_norm": 0.4493967890739441,
+      "learning_rate": 6.950066756178653e-07,
+      "loss": 0.2927,
+      "step": 23555
+    },
+    {
+      "epoch": 0.8808094025282146,
+      "grad_norm": 0.4241293668746948,
+      "learning_rate": 6.928571937358131e-07,
+      "loss": 0.2596,
+      "step": 23560
+    },
+    {
+      "epoch": 0.8809963315185644,
+      "grad_norm": 0.36346501111984253,
+      "learning_rate": 6.907109216206342e-07,
+      "loss": 0.2376,
+      "step": 23565
+    },
+    {
+      "epoch": 0.8811832605089142,
+      "grad_norm": 0.5046384334564209,
+      "learning_rate": 6.885678600125101e-07,
+      "loss": 0.3006,
+      "step": 23570
+    },
+    {
+      "epoch": 0.8813701894992639,
+      "grad_norm": 0.5586752891540527,
+      "learning_rate": 6.864280096505283e-07,
+      "loss": 0.2569,
+      "step": 23575
+    },
+    {
+      "epoch": 0.8815571184896137,
+      "grad_norm": 0.2976597547531128,
+      "learning_rate": 6.842913712726551e-07,
+      "loss": 0.2507,
+      "step": 23580
+    },
+    {
+      "epoch": 0.8817440474799636,
+      "grad_norm": 0.5597246289253235,
+      "learning_rate": 6.821579456157612e-07,
+      "loss": 0.3023,
+      "step": 23585
+    },
+    {
+      "epoch": 0.8819309764703134,
+      "grad_norm": 0.6922735571861267,
+      "learning_rate": 6.800277334156013e-07,
+      "loss": 0.2563,
+      "step": 23590
+    },
+    {
+      "epoch": 0.8821179054606632,
+      "grad_norm": 0.32556474208831787,
+      "learning_rate": 6.779007354068257e-07,
+      "loss": 0.2421,
+      "step": 23595
+    },
+    {
+      "epoch": 0.8823048344510129,
+      "grad_norm": 0.2897225618362427,
+      "learning_rate": 6.757769523229751e-07,
+      "loss": 0.255,
+      "step": 23600
+    },
+    {
+      "epoch": 0.8824917634413627,
+      "grad_norm": 0.45104631781578064,
+      "learning_rate": 6.736563848964784e-07,
+      "loss": 0.2586,
+      "step": 23605
+    },
+    {
+      "epoch": 0.8826786924317125,
+      "grad_norm": 0.5284656286239624,
+      "learning_rate": 6.715390338586636e-07,
+      "loss": 0.2726,
+      "step": 23610
+    },
+    {
+      "epoch": 0.8828656214220623,
+      "grad_norm": 0.42080292105674744,
+      "learning_rate": 6.694248999397402e-07,
+      "loss": 0.1887,
+      "step": 23615
+    },
+    {
+      "epoch": 0.883052550412412,
+      "grad_norm": 0.7159053087234497,
+      "learning_rate": 6.673139838688148e-07,
+      "loss": 0.3204,
+      "step": 23620
+    },
+    {
+      "epoch": 0.8832394794027619,
+      "grad_norm": 0.635303795337677,
+      "learning_rate": 6.652062863738795e-07,
+      "loss": 0.3029,
+      "step": 23625
+    },
+    {
+      "epoch": 0.8834264083931117,
+      "grad_norm": 0.4003882110118866,
+      "learning_rate": 6.63101808181823e-07,
+      "loss": 0.3126,
+      "step": 23630
+    },
+    {
+      "epoch": 0.8836133373834615,
+      "grad_norm": 0.35992348194122314,
+      "learning_rate": 6.610005500184147e-07,
+      "loss": 0.2064,
+      "step": 23635
+    },
+    {
+      "epoch": 0.8838002663738113,
+      "grad_norm": 0.29172858595848083,
+      "learning_rate": 6.589025126083216e-07,
+      "loss": 0.3099,
+      "step": 23640
+    },
+    {
+      "epoch": 0.883987195364161,
+      "grad_norm": 0.43935173749923706,
+      "learning_rate": 6.568076966750958e-07,
+      "loss": 0.3544,
+      "step": 23645
+    },
+    {
+      "epoch": 0.8841741243545108,
+      "grad_norm": 0.6253902316093445,
+      "learning_rate": 6.547161029411775e-07,
+      "loss": 0.3469,
+      "step": 23650
+    },
+    {
+      "epoch": 0.8843610533448606,
+      "grad_norm": 0.570335865020752,
+      "learning_rate": 6.526277321279006e-07,
+      "loss": 0.2235,
+      "step": 23655
+    },
+    {
+      "epoch": 0.8845479823352104,
+      "grad_norm": 1.0959036350250244,
+      "learning_rate": 6.505425849554825e-07,
+      "loss": 0.2871,
+      "step": 23660
+    },
+    {
+      "epoch": 0.8847349113255603,
+      "grad_norm": 0.6984147429466248,
+      "learning_rate": 6.484606621430312e-07,
+      "loss": 0.2487,
+      "step": 23665
+    },
+    {
+      "epoch": 0.88492184031591,
+      "grad_norm": 0.6359334588050842,
+      "learning_rate": 6.463819644085412e-07,
+      "loss": 0.284,
+      "step": 23670
+    },
+    {
+      "epoch": 0.8851087693062598,
+      "grad_norm": 0.4095711410045624,
+      "learning_rate": 6.443064924688969e-07,
+      "loss": 0.2513,
+      "step": 23675
+    },
+    {
+      "epoch": 0.8852956982966096,
+      "grad_norm": 0.28603771328926086,
+      "learning_rate": 6.422342470398679e-07,
+      "loss": 0.2653,
+      "step": 23680
+    },
+    {
+      "epoch": 0.8854826272869594,
+      "grad_norm": 0.36115431785583496,
+      "learning_rate": 6.40165228836116e-07,
+      "loss": 0.2031,
+      "step": 23685
+    },
+    {
+      "epoch": 0.8856695562773091,
+      "grad_norm": 0.3040491044521332,
+      "learning_rate": 6.380994385711803e-07,
+      "loss": 0.3044,
+      "step": 23690
+    },
+    {
+      "epoch": 0.8858564852676589,
+      "grad_norm": 0.4708629548549652,
+      "learning_rate": 6.360368769574977e-07,
+      "loss": 0.3036,
+      "step": 23695
+    },
+    {
+      "epoch": 0.8860434142580087,
+      "grad_norm": 0.19724823534488678,
+      "learning_rate": 6.339775447063856e-07,
+      "loss": 0.2344,
+      "step": 23700
+    },
+    {
+      "epoch": 0.8862303432483585,
+      "grad_norm": 0.6166355609893799,
+      "learning_rate": 6.319214425280451e-07,
+      "loss": 0.2477,
+      "step": 23705
+    },
+    {
+      "epoch": 0.8864172722387084,
+      "grad_norm": 0.4069211184978485,
+      "learning_rate": 6.298685711315722e-07,
+      "loss": 0.2525,
+      "step": 23710
+    },
+    {
+      "epoch": 0.8866042012290581,
+      "grad_norm": 0.6085258722305298,
+      "learning_rate": 6.278189312249395e-07,
+      "loss": 0.3372,
+      "step": 23715
+    },
+    {
+      "epoch": 0.8867911302194079,
+      "grad_norm": 0.3506983518600464,
+      "learning_rate": 6.257725235150113e-07,
+      "loss": 0.2526,
+      "step": 23720
+    },
+    {
+      "epoch": 0.8869780592097577,
+      "grad_norm": 0.6167741417884827,
+      "learning_rate": 6.237293487075324e-07,
+      "loss": 0.2678,
+      "step": 23725
+    },
+    {
+      "epoch": 0.8871649882001075,
+      "grad_norm": 0.3610391318798065,
+      "learning_rate": 6.216894075071378e-07,
+      "loss": 0.1962,
+      "step": 23730
+    },
+    {
+      "epoch": 0.8873519171904573,
+      "grad_norm": 0.5724371075630188,
+      "learning_rate": 6.196527006173447e-07,
+      "loss": 0.2723,
+      "step": 23735
+    },
+    {
+      "epoch": 0.887538846180807,
+      "grad_norm": 0.4358518719673157,
+      "learning_rate": 6.176192287405547e-07,
+      "loss": 0.2568,
+      "step": 23740
+    },
+    {
+      "epoch": 0.8877257751711568,
+      "grad_norm": 0.34022319316864014,
+      "learning_rate": 6.155889925780534e-07,
+      "loss": 0.25,
+      "step": 23745
+    },
+    {
+      "epoch": 0.8879127041615067,
+      "grad_norm": 0.5654661059379578,
+      "learning_rate": 6.135619928300096e-07,
+      "loss": 0.3062,
+      "step": 23750
+    },
+    {
+      "epoch": 0.8880996331518565,
+      "grad_norm": 0.27159449458122253,
+      "learning_rate": 6.115382301954809e-07,
+      "loss": 0.2635,
+      "step": 23755
+    },
+    {
+      "epoch": 0.8882865621422062,
+      "grad_norm": 0.33312445878982544,
+      "learning_rate": 6.095177053724011e-07,
+      "loss": 0.2983,
+      "step": 23760
+    },
+    {
+      "epoch": 0.888473491132556,
+      "grad_norm": 0.7186973094940186,
+      "learning_rate": 6.07500419057595e-07,
+      "loss": 0.2881,
+      "step": 23765
+    },
+    {
+      "epoch": 0.8886604201229058,
+      "grad_norm": 0.4082525074481964,
+      "learning_rate": 6.054863719467641e-07,
+      "loss": 0.2782,
+      "step": 23770
+    },
+    {
+      "epoch": 0.8888473491132556,
+      "grad_norm": 0.22623296082019806,
+      "learning_rate": 6.034755647344958e-07,
+      "loss": 0.198,
+      "step": 23775
+    },
+    {
+      "epoch": 0.8890342781036054,
+      "grad_norm": 0.26331770420074463,
+      "learning_rate": 6.014679981142635e-07,
+      "loss": 0.2145,
+      "step": 23780
+    },
+    {
+      "epoch": 0.8892212070939551,
+      "grad_norm": 0.613768458366394,
+      "learning_rate": 5.994636727784153e-07,
+      "loss": 0.282,
+      "step": 23785
+    },
+    {
+      "epoch": 0.889408136084305,
+      "grad_norm": 0.27621743083000183,
+      "learning_rate": 5.974625894181874e-07,
+      "loss": 0.2704,
+      "step": 23790
+    },
+    {
+      "epoch": 0.8895950650746548,
+      "grad_norm": 1.3323360681533813,
+      "learning_rate": 5.954647487236942e-07,
+      "loss": 0.2972,
+      "step": 23795
+    },
+    {
+      "epoch": 0.8897819940650046,
+      "grad_norm": 0.46217793226242065,
+      "learning_rate": 5.934701513839369e-07,
+      "loss": 0.3143,
+      "step": 23800
+    },
+    {
+      "epoch": 0.8899689230553544,
+      "grad_norm": 0.6428271532058716,
+      "learning_rate": 5.91478798086792e-07,
+      "loss": 0.2476,
+      "step": 23805
+    },
+    {
+      "epoch": 0.8901558520457041,
+      "grad_norm": 0.5378536581993103,
+      "learning_rate": 5.894906895190222e-07,
+      "loss": 0.2954,
+      "step": 23810
+    },
+    {
+      "epoch": 0.8903427810360539,
+      "grad_norm": 0.34355923533439636,
+      "learning_rate": 5.875058263662669e-07,
+      "loss": 0.2932,
+      "step": 23815
+    },
+    {
+      "epoch": 0.8905297100264037,
+      "grad_norm": 0.42726051807403564,
+      "learning_rate": 5.855242093130498e-07,
+      "loss": 0.2709,
+      "step": 23820
+    },
+    {
+      "epoch": 0.8907166390167535,
+      "grad_norm": 0.4561409652233124,
+      "learning_rate": 5.835458390427762e-07,
+      "loss": 0.2824,
+      "step": 23825
+    },
+    {
+      "epoch": 0.8909035680071034,
+      "grad_norm": 0.523114800453186,
+      "learning_rate": 5.815707162377271e-07,
+      "loss": 0.2704,
+      "step": 23830
+    },
+    {
+      "epoch": 0.8910904969974531,
+      "grad_norm": 3.815091848373413,
+      "learning_rate": 5.795988415790655e-07,
+      "loss": 0.3295,
+      "step": 23835
+    },
+    {
+      "epoch": 0.8912774259878029,
+      "grad_norm": 0.3353985548019409,
+      "learning_rate": 5.776302157468338e-07,
+      "loss": 0.3024,
+      "step": 23840
+    },
+    {
+      "epoch": 0.8914643549781527,
+      "grad_norm": 0.4886054992675781,
+      "learning_rate": 5.756648394199571e-07,
+      "loss": 0.3643,
+      "step": 23845
+    },
+    {
+      "epoch": 0.8916512839685025,
+      "grad_norm": 0.28375887870788574,
+      "learning_rate": 5.737027132762341e-07,
+      "loss": 0.2608,
+      "step": 23850
+    },
+    {
+      "epoch": 0.8918382129588522,
+      "grad_norm": 0.39414143562316895,
+      "learning_rate": 5.71743837992349e-07,
+      "loss": 0.3279,
+      "step": 23855
+    },
+    {
+      "epoch": 0.892025141949202,
+      "grad_norm": 0.29853355884552,
+      "learning_rate": 5.697882142438594e-07,
+      "loss": 0.2457,
+      "step": 23860
+    },
+    {
+      "epoch": 0.8922120709395518,
+      "grad_norm": 0.49426108598709106,
+      "learning_rate": 5.678358427052045e-07,
+      "loss": 0.2018,
+      "step": 23865
+    },
+    {
+      "epoch": 0.8923989999299017,
+      "grad_norm": 0.48164868354797363,
+      "learning_rate": 5.658867240497034e-07,
+      "loss": 0.2593,
+      "step": 23870
+    },
+    {
+      "epoch": 0.8925859289202515,
+      "grad_norm": 0.3493533432483673,
+      "learning_rate": 5.639408589495476e-07,
+      "loss": 0.3238,
+      "step": 23875
+    },
+    {
+      "epoch": 0.8927728579106012,
+      "grad_norm": 0.40912967920303345,
+      "learning_rate": 5.619982480758146e-07,
+      "loss": 0.2422,
+      "step": 23880
+    },
+    {
+      "epoch": 0.892959786900951,
+      "grad_norm": 0.50095134973526,
+      "learning_rate": 5.600588920984529e-07,
+      "loss": 0.2204,
+      "step": 23885
+    },
+    {
+      "epoch": 0.8931467158913008,
+      "grad_norm": 0.3179953992366791,
+      "learning_rate": 5.581227916862907e-07,
+      "loss": 0.2518,
+      "step": 23890
+    },
+    {
+      "epoch": 0.8933336448816506,
+      "grad_norm": 0.39287543296813965,
+      "learning_rate": 5.56189947507032e-07,
+      "loss": 0.291,
+      "step": 23895
+    },
+    {
+      "epoch": 0.8935205738720003,
+      "grad_norm": 0.44171902537345886,
+      "learning_rate": 5.542603602272622e-07,
+      "loss": 0.2908,
+      "step": 23900
+    },
+    {
+      "epoch": 0.8937075028623501,
+      "grad_norm": 0.8764915466308594,
+      "learning_rate": 5.523340305124381e-07,
+      "loss": 0.3125,
+      "step": 23905
+    },
+    {
+      "epoch": 0.8938944318527,
+      "grad_norm": 0.3236830234527588,
+      "learning_rate": 5.50410959026898e-07,
+      "loss": 0.2637,
+      "step": 23910
+    },
+    {
+      "epoch": 0.8940813608430498,
+      "grad_norm": 0.5709930062294006,
+      "learning_rate": 5.484911464338539e-07,
+      "loss": 0.2993,
+      "step": 23915
+    },
+    {
+      "epoch": 0.8942682898333996,
+      "grad_norm": 0.22036674618721008,
+      "learning_rate": 5.465745933953914e-07,
+      "loss": 0.3086,
+      "step": 23920
+    },
+    {
+      "epoch": 0.8944552188237493,
+      "grad_norm": 0.36622971296310425,
+      "learning_rate": 5.446613005724788e-07,
+      "loss": 0.287,
+      "step": 23925
+    },
+    {
+      "epoch": 0.8946421478140991,
+      "grad_norm": 0.37628886103630066,
+      "learning_rate": 5.427512686249537e-07,
+      "loss": 0.2985,
+      "step": 23930
+    },
+    {
+      "epoch": 0.8948290768044489,
+      "grad_norm": 0.3855689764022827,
+      "learning_rate": 5.408444982115313e-07,
+      "loss": 0.2803,
+      "step": 23935
+    },
+    {
+      "epoch": 0.8950160057947987,
+      "grad_norm": 0.0829666256904602,
+      "learning_rate": 5.389409899898013e-07,
+      "loss": 0.2945,
+      "step": 23940
+    },
+    {
+      "epoch": 0.8952029347851485,
+      "grad_norm": 0.5317602753639221,
+      "learning_rate": 5.370407446162318e-07,
+      "loss": 0.2388,
+      "step": 23945
+    },
+    {
+      "epoch": 0.8953898637754982,
+      "grad_norm": 0.635329008102417,
+      "learning_rate": 5.351437627461598e-07,
+      "loss": 0.351,
+      "step": 23950
+    },
+    {
+      "epoch": 0.8955767927658481,
+      "grad_norm": 0.48536187410354614,
+      "learning_rate": 5.332500450338018e-07,
+      "loss": 0.3298,
+      "step": 23955
+    },
+    {
+      "epoch": 0.8957637217561979,
+      "grad_norm": 0.3373602330684662,
+      "learning_rate": 5.313595921322479e-07,
+      "loss": 0.2615,
+      "step": 23960
+    },
+    {
+      "epoch": 0.8959506507465477,
+      "grad_norm": 0.49494582414627075,
+      "learning_rate": 5.294724046934585e-07,
+      "loss": 0.2406,
+      "step": 23965
+    },
+    {
+      "epoch": 0.8961375797368974,
+      "grad_norm": 0.36144334077835083,
+      "learning_rate": 5.275884833682721e-07,
+      "loss": 0.2532,
+      "step": 23970
+    },
+    {
+      "epoch": 0.8963245087272472,
+      "grad_norm": 0.6861150860786438,
+      "learning_rate": 5.257078288064e-07,
+      "loss": 0.2478,
+      "step": 23975
+    },
+    {
+      "epoch": 0.896511437717597,
+      "grad_norm": 0.22255679965019226,
+      "learning_rate": 5.238304416564243e-07,
+      "loss": 0.3317,
+      "step": 23980
+    },
+    {
+      "epoch": 0.8966983667079468,
+      "grad_norm": 0.2977222800254822,
+      "learning_rate": 5.21956322565802e-07,
+      "loss": 0.2917,
+      "step": 23985
+    },
+    {
+      "epoch": 0.8968852956982966,
+      "grad_norm": 0.7013628482818604,
+      "learning_rate": 5.200854721808645e-07,
+      "loss": 0.3236,
+      "step": 23990
+    },
+    {
+      "epoch": 0.8970722246886464,
+      "grad_norm": 0.46044737100601196,
+      "learning_rate": 5.182178911468128e-07,
+      "loss": 0.2712,
+      "step": 23995
+    },
+    {
+      "epoch": 0.8972591536789962,
+      "grad_norm": 0.21883058547973633,
+      "learning_rate": 5.163535801077235e-07,
+      "loss": 0.2518,
+      "step": 24000
+    },
+    {
+      "epoch": 0.897446082669346,
+      "grad_norm": 0.3243284523487091,
+      "learning_rate": 5.144925397065437e-07,
+      "loss": 0.3362,
+      "step": 24005
+    },
+    {
+      "epoch": 0.8976330116596958,
+      "grad_norm": 0.5240607261657715,
+      "learning_rate": 5.12634770585092e-07,
+      "loss": 0.2729,
+      "step": 24010
+    },
+    {
+      "epoch": 0.8978199406500456,
+      "grad_norm": 0.318803608417511,
+      "learning_rate": 5.107802733840616e-07,
+      "loss": 0.2773,
+      "step": 24015
+    },
+    {
+      "epoch": 0.8980068696403953,
+      "grad_norm": 0.41666528582572937,
+      "learning_rate": 5.089290487430154e-07,
+      "loss": 0.233,
+      "step": 24020
+    },
+    {
+      "epoch": 0.8981937986307451,
+      "grad_norm": 0.5397652387619019,
+      "learning_rate": 5.070810973003859e-07,
+      "loss": 0.2415,
+      "step": 24025
+    },
+    {
+      "epoch": 0.8983807276210949,
+      "grad_norm": 0.4510786533355713,
+      "learning_rate": 5.052364196934779e-07,
+      "loss": 0.2595,
+      "step": 24030
+    },
+    {
+      "epoch": 0.8985676566114448,
+      "grad_norm": 0.4586493670940399,
+      "learning_rate": 5.033950165584711e-07,
+      "loss": 0.4253,
+      "step": 24035
+    },
+    {
+      "epoch": 0.8987545856017946,
+      "grad_norm": 0.6832088828086853,
+      "learning_rate": 5.01556888530409e-07,
+      "loss": 0.3542,
+      "step": 24040
+    },
+    {
+      "epoch": 0.8989415145921443,
+      "grad_norm": 0.2557532489299774,
+      "learning_rate": 4.99722036243212e-07,
+      "loss": 0.2235,
+      "step": 24045
+    },
+    {
+      "epoch": 0.8991284435824941,
+      "grad_norm": 0.2579525113105774,
+      "learning_rate": 4.978904603296686e-07,
+      "loss": 0.2518,
+      "step": 24050
+    },
+    {
+      "epoch": 0.8993153725728439,
+      "grad_norm": 0.7935723066329956,
+      "learning_rate": 4.960621614214334e-07,
+      "loss": 0.2589,
+      "step": 24055
+    },
+    {
+      "epoch": 0.8995023015631937,
+      "grad_norm": 0.6913199424743652,
+      "learning_rate": 4.942371401490386e-07,
+      "loss": 0.2652,
+      "step": 24060
+    },
+    {
+      "epoch": 0.8996892305535434,
+      "grad_norm": 0.604274570941925,
+      "learning_rate": 4.924153971418777e-07,
+      "loss": 0.2887,
+      "step": 24065
+    },
+    {
+      "epoch": 0.8998761595438932,
+      "grad_norm": 0.35394561290740967,
+      "learning_rate": 4.905969330282212e-07,
+      "loss": 0.2466,
+      "step": 24070
+    },
+    {
+      "epoch": 0.9000630885342431,
+      "grad_norm": 0.39302387833595276,
+      "learning_rate": 4.887817484352031e-07,
+      "loss": 0.3516,
+      "step": 24075
+    },
+    {
+      "epoch": 0.9002500175245929,
+      "grad_norm": 0.5640286207199097,
+      "learning_rate": 4.869698439888304e-07,
+      "loss": 0.3531,
+      "step": 24080
+    },
+    {
+      "epoch": 0.9004369465149427,
+      "grad_norm": 0.4714716076850891,
+      "learning_rate": 4.851612203139733e-07,
+      "loss": 0.2918,
+      "step": 24085
+    },
+    {
+      "epoch": 0.9006238755052924,
+      "grad_norm": 0.41106295585632324,
+      "learning_rate": 4.833558780343772e-07,
+      "loss": 0.3051,
+      "step": 24090
+    },
+    {
+      "epoch": 0.9008108044956422,
+      "grad_norm": 0.48679038882255554,
+      "learning_rate": 4.815538177726531e-07,
+      "loss": 0.2615,
+      "step": 24095
+    },
+    {
+      "epoch": 0.900997733485992,
+      "grad_norm": 0.5820590257644653,
+      "learning_rate": 4.797550401502782e-07,
+      "loss": 0.2669,
+      "step": 24100
+    },
+    {
+      "epoch": 0.9011846624763418,
+      "grad_norm": 0.5170806050300598,
+      "learning_rate": 4.779595457876019e-07,
+      "loss": 0.2829,
+      "step": 24105
+    },
+    {
+      "epoch": 0.9013715914666915,
+      "grad_norm": 0.4748556613922119,
+      "learning_rate": 4.761673353038354e-07,
+      "loss": 0.2686,
+      "step": 24110
+    },
+    {
+      "epoch": 0.9015585204570414,
+      "grad_norm": 0.3551413118839264,
+      "learning_rate": 4.743784093170645e-07,
+      "loss": 0.3079,
+      "step": 24115
+    },
+    {
+      "epoch": 0.9017454494473912,
+      "grad_norm": 0.47169989347457886,
+      "learning_rate": 4.725927684442366e-07,
+      "loss": 0.2218,
+      "step": 24120
+    },
+    {
+      "epoch": 0.901932378437741,
+      "grad_norm": 0.39478862285614014,
+      "learning_rate": 4.7081041330116816e-07,
+      "loss": 0.2159,
+      "step": 24125
+    },
+    {
+      "epoch": 0.9021193074280908,
+      "grad_norm": 0.33222270011901855,
+      "learning_rate": 4.6903134450254186e-07,
+      "loss": 0.3028,
+      "step": 24130
+    },
+    {
+      "epoch": 0.9023062364184405,
+      "grad_norm": 0.39596304297447205,
+      "learning_rate": 4.6725556266190687e-07,
+      "loss": 0.1949,
+      "step": 24135
+    },
+    {
+      "epoch": 0.9024931654087903,
+      "grad_norm": 0.31857696175575256,
+      "learning_rate": 4.6548306839168224e-07,
+      "loss": 0.2729,
+      "step": 24140
+    },
+    {
+      "epoch": 0.9026800943991401,
+      "grad_norm": 0.4975537359714508,
+      "learning_rate": 4.6371386230314785e-07,
+      "loss": 0.3893,
+      "step": 24145
+    },
+    {
+      "epoch": 0.9028670233894899,
+      "grad_norm": 0.4041035771369934,
+      "learning_rate": 4.619479450064535e-07,
+      "loss": 0.3135,
+      "step": 24150
+    },
+    {
+      "epoch": 0.9030539523798398,
+      "grad_norm": 0.5101954936981201,
+      "learning_rate": 4.6018531711061297e-07,
+      "loss": 0.2647,
+      "step": 24155
+    },
+    {
+      "epoch": 0.9032408813701895,
+      "grad_norm": 0.4400761127471924,
+      "learning_rate": 4.5842597922350683e-07,
+      "loss": 0.2631,
+      "step": 24160
+    },
+    {
+      "epoch": 0.9034278103605393,
+      "grad_norm": 0.4869976341724396,
+      "learning_rate": 4.566699319518808e-07,
+      "loss": 0.2841,
+      "step": 24165
+    },
+    {
+      "epoch": 0.9036147393508891,
+      "grad_norm": 0.6493037343025208,
+      "learning_rate": 4.5491717590134377e-07,
+      "loss": 0.3233,
+      "step": 24170
+    },
+    {
+      "epoch": 0.9038016683412389,
+      "grad_norm": 0.2384498566389084,
+      "learning_rate": 4.5316771167637e-07,
+      "loss": 0.2755,
+      "step": 24175
+    },
+    {
+      "epoch": 0.9039885973315887,
+      "grad_norm": 0.5434466004371643,
+      "learning_rate": 4.5142153988030236e-07,
+      "loss": 0.2591,
+      "step": 24180
+    },
+    {
+      "epoch": 0.9041755263219384,
+      "grad_norm": 0.2663435935974121,
+      "learning_rate": 4.4967866111534254e-07,
+      "loss": 0.2451,
+      "step": 24185
+    },
+    {
+      "epoch": 0.9043624553122882,
+      "grad_norm": 0.48140108585357666,
+      "learning_rate": 4.4793907598256193e-07,
+      "loss": 0.3392,
+      "step": 24190
+    },
+    {
+      "epoch": 0.904549384302638,
+      "grad_norm": 0.29158666729927063,
+      "learning_rate": 4.4620278508189395e-07,
+      "loss": 0.2499,
+      "step": 24195
+    },
+    {
+      "epoch": 0.9047363132929879,
+      "grad_norm": 0.8527382612228394,
+      "learning_rate": 4.44469789012133e-07,
+      "loss": 0.2645,
+      "step": 24200
+    },
+    {
+      "epoch": 0.9049232422833376,
+      "grad_norm": 0.469309538602829,
+      "learning_rate": 4.4274008837094316e-07,
+      "loss": 0.3296,
+      "step": 24205
+    },
+    {
+      "epoch": 0.9051101712736874,
+      "grad_norm": 0.7557592988014221,
+      "learning_rate": 4.410136837548462e-07,
+      "loss": 0.2996,
+      "step": 24210
+    },
+    {
+      "epoch": 0.9052971002640372,
+      "grad_norm": 0.36758342385292053,
+      "learning_rate": 4.392905757592303e-07,
+      "loss": 0.2624,
+      "step": 24215
+    },
+    {
+      "epoch": 0.905484029254387,
+      "grad_norm": 0.6781210899353027,
+      "learning_rate": 4.3757076497834337e-07,
+      "loss": 0.3425,
+      "step": 24220
+    },
+    {
+      "epoch": 0.9056709582447368,
+      "grad_norm": 0.4783806800842285,
+      "learning_rate": 4.358542520053044e-07,
+      "loss": 0.3332,
+      "step": 24225
+    },
+    {
+      "epoch": 0.9058578872350865,
+      "grad_norm": 0.33384254574775696,
+      "learning_rate": 4.3414103743208426e-07,
+      "loss": 0.2122,
+      "step": 24230
+    },
+    {
+      "epoch": 0.9060448162254363,
+      "grad_norm": 0.31750744581222534,
+      "learning_rate": 4.3243112184952365e-07,
+      "loss": 0.31,
+      "step": 24235
+    },
+    {
+      "epoch": 0.9062317452157862,
+      "grad_norm": 0.32413434982299805,
+      "learning_rate": 4.307245058473253e-07,
+      "loss": 0.2356,
+      "step": 24240
+    },
+    {
+      "epoch": 0.906418674206136,
+      "grad_norm": 0.5775876641273499,
+      "learning_rate": 4.290211900140495e-07,
+      "loss": 0.3082,
+      "step": 24245
+    },
+    {
+      "epoch": 0.9066056031964858,
+      "grad_norm": 0.3418591618537903,
+      "learning_rate": 4.27321174937122e-07,
+      "loss": 0.3737,
+      "step": 24250
+    },
+    {
+      "epoch": 0.9067925321868355,
+      "grad_norm": 0.48873278498649597,
+      "learning_rate": 4.256244612028293e-07,
+      "loss": 0.3144,
+      "step": 24255
+    },
+    {
+      "epoch": 0.9069794611771853,
+      "grad_norm": 0.43501177430152893,
+      "learning_rate": 4.2393104939632e-07,
+      "loss": 0.3015,
+      "step": 24260
+    },
+    {
+      "epoch": 0.9071663901675351,
+      "grad_norm": 0.46144506335258484,
+      "learning_rate": 4.222409401016025e-07,
+      "loss": 0.3018,
+      "step": 24265
+    },
+    {
+      "epoch": 0.9073533191578849,
+      "grad_norm": 0.4738104045391083,
+      "learning_rate": 4.205541339015484e-07,
+      "loss": 0.2621,
+      "step": 24270
+    },
+    {
+      "epoch": 0.9075402481482346,
+      "grad_norm": 0.5475369691848755,
+      "learning_rate": 4.1887063137788565e-07,
+      "loss": 0.3754,
+      "step": 24275
+    },
+    {
+      "epoch": 0.9077271771385845,
+      "grad_norm": 0.26058679819107056,
+      "learning_rate": 4.1719043311120757e-07,
+      "loss": 0.2475,
+      "step": 24280
+    },
+    {
+      "epoch": 0.9079141061289343,
+      "grad_norm": 0.6181069016456604,
+      "learning_rate": 4.155135396809684e-07,
+      "loss": 0.3012,
+      "step": 24285
+    },
+    {
+      "epoch": 0.9081010351192841,
+      "grad_norm": 0.4764721095561981,
+      "learning_rate": 4.13839951665479e-07,
+      "loss": 0.28,
+      "step": 24290
+    },
+    {
+      "epoch": 0.9082879641096339,
+      "grad_norm": 0.20308902859687805,
+      "learning_rate": 4.1216966964191194e-07,
+      "loss": 0.341,
+      "step": 24295
+    },
+    {
+      "epoch": 0.9084748930999836,
+      "grad_norm": 0.3182278573513031,
+      "learning_rate": 4.1050269418629887e-07,
+      "loss": 0.3403,
+      "step": 24300
+    },
+    {
+      "epoch": 0.9086618220903334,
+      "grad_norm": 0.4016958475112915,
+      "learning_rate": 4.088390258735342e-07,
+      "loss": 0.2859,
+      "step": 24305
+    },
+    {
+      "epoch": 0.9088487510806832,
+      "grad_norm": 0.5920534729957581,
+      "learning_rate": 4.071786652773679e-07,
+      "loss": 0.3373,
+      "step": 24310
+    },
+    {
+      "epoch": 0.909035680071033,
+      "grad_norm": 0.5127608180046082,
+      "learning_rate": 4.0552161297041094e-07,
+      "loss": 0.3156,
+      "step": 24315
+    },
+    {
+      "epoch": 0.9092226090613829,
+      "grad_norm": 0.40801453590393066,
+      "learning_rate": 4.038678695241316e-07,
+      "loss": 0.3016,
+      "step": 24320
+    },
+    {
+      "epoch": 0.9094095380517326,
+      "grad_norm": 0.5536433458328247,
+      "learning_rate": 4.022174355088593e-07,
+      "loss": 0.3749,
+      "step": 24325
+    },
+    {
+      "epoch": 0.9095964670420824,
+      "grad_norm": 0.3775192201137543,
+      "learning_rate": 4.005703114937842e-07,
+      "loss": 0.3362,
+      "step": 24330
+    },
+    {
+      "epoch": 0.9097833960324322,
+      "grad_norm": 0.3780711591243744,
+      "learning_rate": 3.989264980469498e-07,
+      "loss": 0.2515,
+      "step": 24335
+    },
+    {
+      "epoch": 0.909970325022782,
+      "grad_norm": 0.27277469635009766,
+      "learning_rate": 3.972859957352604e-07,
+      "loss": 0.2751,
+      "step": 24340
+    },
+    {
+      "epoch": 0.9101572540131317,
+      "grad_norm": 0.28428831696510315,
+      "learning_rate": 3.956488051244789e-07,
+      "loss": 0.3447,
+      "step": 24345
+    },
+    {
+      "epoch": 0.9103441830034815,
+      "grad_norm": 0.3745942711830139,
+      "learning_rate": 3.9401492677922483e-07,
+      "loss": 0.2715,
+      "step": 24350
+    },
+    {
+      "epoch": 0.9105311119938313,
+      "grad_norm": 0.33236637711524963,
+      "learning_rate": 3.9238436126297743e-07,
+      "loss": 0.3158,
+      "step": 24355
+    },
+    {
+      "epoch": 0.9107180409841812,
+      "grad_norm": 0.4709293246269226,
+      "learning_rate": 3.9075710913807016e-07,
+      "loss": 0.2759,
+      "step": 24360
+    },
+    {
+      "epoch": 0.910904969974531,
+      "grad_norm": 0.43469110131263733,
+      "learning_rate": 3.8913317096569427e-07,
+      "loss": 0.2859,
+      "step": 24365
+    },
+    {
+      "epoch": 0.9110918989648807,
+      "grad_norm": 0.3980669379234314,
+      "learning_rate": 3.875125473059027e-07,
+      "loss": 0.3309,
+      "step": 24370
+    },
+    {
+      "epoch": 0.9112788279552305,
+      "grad_norm": 0.5004674196243286,
+      "learning_rate": 3.8589523871760183e-07,
+      "loss": 0.1799,
+      "step": 24375
+    },
+    {
+      "epoch": 0.9114657569455803,
+      "grad_norm": 0.28025054931640625,
+      "learning_rate": 3.8428124575855317e-07,
+      "loss": 0.2318,
+      "step": 24380
+    },
+    {
+      "epoch": 0.9116526859359301,
+      "grad_norm": 0.6791032552719116,
+      "learning_rate": 3.826705689853782e-07,
+      "loss": 0.2544,
+      "step": 24385
+    },
+    {
+      "epoch": 0.9118396149262799,
+      "grad_norm": 0.44797998666763306,
+      "learning_rate": 3.810632089535526e-07,
+      "loss": 0.2337,
+      "step": 24390
+    },
+    {
+      "epoch": 0.9120265439166296,
+      "grad_norm": 0.6669244766235352,
+      "learning_rate": 3.794591662174096e-07,
+      "loss": 0.1961,
+      "step": 24395
+    },
+    {
+      "epoch": 0.9122134729069794,
+      "grad_norm": 0.9826619625091553,
+      "learning_rate": 3.778584413301356e-07,
+      "loss": 0.24,
+      "step": 24400
+    },
+    {
+      "epoch": 0.9124004018973293,
+      "grad_norm": 0.5112035870552063,
+      "learning_rate": 3.7626103484377674e-07,
+      "loss": 0.2262,
+      "step": 24405
+    },
+    {
+      "epoch": 0.9125873308876791,
+      "grad_norm": 0.3886997699737549,
+      "learning_rate": 3.7466694730923124e-07,
+      "loss": 0.2582,
+      "step": 24410
+    },
+    {
+      "epoch": 0.9127742598780288,
+      "grad_norm": 0.26052147150039673,
+      "learning_rate": 3.7307617927625494e-07,
+      "loss": 0.3355,
+      "step": 24415
+    },
+    {
+      "epoch": 0.9129611888683786,
+      "grad_norm": 0.18169300258159637,
+      "learning_rate": 3.7148873129345896e-07,
+      "loss": 0.2485,
+      "step": 24420
+    },
+    {
+      "epoch": 0.9131481178587284,
+      "grad_norm": 0.6013249158859253,
+      "learning_rate": 3.6990460390830653e-07,
+      "loss": 0.2622,
+      "step": 24425
+    },
+    {
+      "epoch": 0.9133350468490782,
+      "grad_norm": 0.12792934477329254,
+      "learning_rate": 3.6832379766712057e-07,
+      "loss": 0.2445,
+      "step": 24430
+    },
+    {
+      "epoch": 0.913521975839428,
+      "grad_norm": 0.6494380235671997,
+      "learning_rate": 3.667463131150728e-07,
+      "loss": 0.2618,
+      "step": 24435
+    },
+    {
+      "epoch": 0.9137089048297777,
+      "grad_norm": 1.2967865467071533,
+      "learning_rate": 3.6517215079619583e-07,
+      "loss": 0.3326,
+      "step": 24440
+    },
+    {
+      "epoch": 0.9138958338201276,
+      "grad_norm": 0.43164533376693726,
+      "learning_rate": 3.6360131125336983e-07,
+      "loss": 0.2116,
+      "step": 24445
+    },
+    {
+      "epoch": 0.9140827628104774,
+      "grad_norm": 0.2929776608943939,
+      "learning_rate": 3.620337950283348e-07,
+      "loss": 0.3048,
+      "step": 24450
+    },
+    {
+      "epoch": 0.9142696918008272,
+      "grad_norm": 0.4373328685760498,
+      "learning_rate": 3.6046960266168163e-07,
+      "loss": 0.2887,
+      "step": 24455
+    },
+    {
+      "epoch": 0.914456620791177,
+      "grad_norm": 0.29323285818099976,
+      "learning_rate": 3.5890873469285325e-07,
+      "loss": 0.3293,
+      "step": 24460
+    },
+    {
+      "epoch": 0.9146435497815267,
+      "grad_norm": 0.3576362729072571,
+      "learning_rate": 3.573511916601513e-07,
+      "loss": 0.2113,
+      "step": 24465
+    },
+    {
+      "epoch": 0.9148304787718765,
+      "grad_norm": 0.5483099222183228,
+      "learning_rate": 3.55796974100725e-07,
+      "loss": 0.2588,
+      "step": 24470
+    },
+    {
+      "epoch": 0.9150174077622263,
+      "grad_norm": 0.30468040704727173,
+      "learning_rate": 3.5424608255058334e-07,
+      "loss": 0.2478,
+      "step": 24475
+    },
+    {
+      "epoch": 0.9152043367525761,
+      "grad_norm": 0.3078984320163727,
+      "learning_rate": 3.526985175445796e-07,
+      "loss": 0.2239,
+      "step": 24480
+    },
+    {
+      "epoch": 0.915391265742926,
+      "grad_norm": 0.2722003757953644,
+      "learning_rate": 3.511542796164291e-07,
+      "loss": 0.2702,
+      "step": 24485
+    },
+    {
+      "epoch": 0.9155781947332757,
+      "grad_norm": 0.317244291305542,
+      "learning_rate": 3.496133692986914e-07,
+      "loss": 0.2063,
+      "step": 24490
+    },
+    {
+      "epoch": 0.9157651237236255,
+      "grad_norm": 0.23974300920963287,
+      "learning_rate": 3.480757871227858e-07,
+      "loss": 0.2249,
+      "step": 24495
+    },
+    {
+      "epoch": 0.9159520527139753,
+      "grad_norm": 0.8161414265632629,
+      "learning_rate": 3.4654153361897815e-07,
+      "loss": 0.3797,
+      "step": 24500
+    },
+    {
+      "epoch": 0.9161389817043251,
+      "grad_norm": 0.28393059968948364,
+      "learning_rate": 3.4501060931638743e-07,
+      "loss": 0.2648,
+      "step": 24505
+    },
+    {
+      "epoch": 0.9163259106946748,
+      "grad_norm": 0.44209572672843933,
+      "learning_rate": 3.4348301474298906e-07,
+      "loss": 0.3289,
+      "step": 24510
+    },
+    {
+      "epoch": 0.9165128396850246,
+      "grad_norm": 0.35628998279571533,
+      "learning_rate": 3.4195875042560276e-07,
+      "loss": 0.2093,
+      "step": 24515
+    },
+    {
+      "epoch": 0.9166997686753744,
+      "grad_norm": 0.3359036445617676,
+      "learning_rate": 3.4043781688990696e-07,
+      "loss": 0.2941,
+      "step": 24520
+    },
+    {
+      "epoch": 0.9168866976657243,
+      "grad_norm": 0.6143574118614197,
+      "learning_rate": 3.3892021466042646e-07,
+      "loss": 0.303,
+      "step": 24525
+    },
+    {
+      "epoch": 0.9170736266560741,
+      "grad_norm": 0.24244384467601776,
+      "learning_rate": 3.374059442605393e-07,
+      "loss": 0.2823,
+      "step": 24530
+    },
+    {
+      "epoch": 0.9172605556464238,
+      "grad_norm": 0.8676662445068359,
+      "learning_rate": 3.3589500621247215e-07,
+      "loss": 0.3931,
+      "step": 24535
+    },
+    {
+      "epoch": 0.9174474846367736,
+      "grad_norm": 0.5118948817253113,
+      "learning_rate": 3.3438740103730716e-07,
+      "loss": 0.2997,
+      "step": 24540
+    },
+    {
+      "epoch": 0.9176344136271234,
+      "grad_norm": 1.1253736019134521,
+      "learning_rate": 3.3288312925497277e-07,
+      "loss": 0.274,
+      "step": 24545
+    },
+    {
+      "epoch": 0.9178213426174732,
+      "grad_norm": 0.32212045788764954,
+      "learning_rate": 3.3138219138424856e-07,
+      "loss": 0.2363,
+      "step": 24550
+    },
+    {
+      "epoch": 0.918008271607823,
+      "grad_norm": 0.5436800122261047,
+      "learning_rate": 3.298845879427659e-07,
+      "loss": 0.2476,
+      "step": 24555
+    },
+    {
+      "epoch": 0.9181952005981727,
+      "grad_norm": 0.6046520471572876,
+      "learning_rate": 3.2839031944700394e-07,
+      "loss": 0.2633,
+      "step": 24560
+    },
+    {
+      "epoch": 0.9183821295885226,
+      "grad_norm": 0.5511953234672546,
+      "learning_rate": 3.2689938641229603e-07,
+      "loss": 0.3388,
+      "step": 24565
+    },
+    {
+      "epoch": 0.9185690585788724,
+      "grad_norm": 0.30859842896461487,
+      "learning_rate": 3.254117893528186e-07,
+      "loss": 0.2729,
+      "step": 24570
+    },
+    {
+      "epoch": 0.9187559875692222,
+      "grad_norm": 0.39766547083854675,
+      "learning_rate": 3.239275287816035e-07,
+      "loss": 0.3074,
+      "step": 24575
+    },
+    {
+      "epoch": 0.9189429165595719,
+      "grad_norm": 0.3398101031780243,
+      "learning_rate": 3.224466052105291e-07,
+      "loss": 0.3279,
+      "step": 24580
+    },
+    {
+      "epoch": 0.9191298455499217,
+      "grad_norm": 0.7181034684181213,
+      "learning_rate": 3.209690191503245e-07,
+      "loss": 0.3621,
+      "step": 24585
+    },
+    {
+      "epoch": 0.9193167745402715,
+      "grad_norm": 0.541123628616333,
+      "learning_rate": 3.194947711105645e-07,
+      "loss": 0.2336,
+      "step": 24590
+    },
+    {
+      "epoch": 0.9195037035306213,
+      "grad_norm": 0.24448862671852112,
+      "learning_rate": 3.180238615996778e-07,
+      "loss": 0.3324,
+      "step": 24595
+    },
+    {
+      "epoch": 0.919690632520971,
+      "grad_norm": 0.8702734112739563,
+      "learning_rate": 3.165562911249376e-07,
+      "loss": 0.2881,
+      "step": 24600
+    },
+    {
+      "epoch": 0.9198775615113209,
+      "grad_norm": 0.38017889857292175,
+      "learning_rate": 3.1509206019246564e-07,
+      "loss": 0.2344,
+      "step": 24605
+    },
+    {
+      "epoch": 0.9200644905016707,
+      "grad_norm": 0.48280656337738037,
+      "learning_rate": 3.136311693072347e-07,
+      "loss": 0.2581,
+      "step": 24610
+    },
+    {
+      "epoch": 0.9202514194920205,
+      "grad_norm": 0.37911441922187805,
+      "learning_rate": 3.1217361897306395e-07,
+      "loss": 0.2333,
+      "step": 24615
+    },
+    {
+      "epoch": 0.9204383484823703,
+      "grad_norm": 0.4707018733024597,
+      "learning_rate": 3.107194096926214e-07,
+      "loss": 0.2859,
+      "step": 24620
+    },
+    {
+      "epoch": 0.92062527747272,
+      "grad_norm": 0.308724969625473,
+      "learning_rate": 3.0926854196742017e-07,
+      "loss": 0.28,
+      "step": 24625
+    },
+    {
+      "epoch": 0.9208122064630698,
+      "grad_norm": 0.3189759850502014,
+      "learning_rate": 3.0782101629782456e-07,
+      "loss": 0.2821,
+      "step": 24630
+    },
+    {
+      "epoch": 0.9209991354534196,
+      "grad_norm": 0.24063967168331146,
+      "learning_rate": 3.063768331830441e-07,
+      "loss": 0.2809,
+      "step": 24635
+    },
+    {
+      "epoch": 0.9211860644437694,
+      "grad_norm": 0.5102589726448059,
+      "learning_rate": 3.049359931211382e-07,
+      "loss": 0.3776,
+      "step": 24640
+    },
+    {
+      "epoch": 0.9213729934341192,
+      "grad_norm": 0.5993364453315735,
+      "learning_rate": 3.034984966090082e-07,
+      "loss": 0.3378,
+      "step": 24645
+    },
+    {
+      "epoch": 0.921559922424469,
+      "grad_norm": 0.5947995185852051,
+      "learning_rate": 3.020643441424065e-07,
+      "loss": 0.2313,
+      "step": 24650
+    },
+    {
+      "epoch": 0.9217468514148188,
+      "grad_norm": 0.35502389073371887,
+      "learning_rate": 3.006335362159329e-07,
+      "loss": 0.2421,
+      "step": 24655
+    },
+    {
+      "epoch": 0.9219337804051686,
+      "grad_norm": 0.5855832099914551,
+      "learning_rate": 2.9920607332302844e-07,
+      "loss": 0.2156,
+      "step": 24660
+    },
+    {
+      "epoch": 0.9221207093955184,
+      "grad_norm": 0.2624458074569702,
+      "learning_rate": 2.97781955955988e-07,
+      "loss": 0.2104,
+      "step": 24665
+    },
+    {
+      "epoch": 0.9223076383858682,
+      "grad_norm": 0.41134822368621826,
+      "learning_rate": 2.9636118460594667e-07,
+      "loss": 0.3424,
+      "step": 24670
+    },
+    {
+      "epoch": 0.9224945673762179,
+      "grad_norm": 0.28406789898872375,
+      "learning_rate": 2.94943759762889e-07,
+      "loss": 0.2791,
+      "step": 24675
+    },
+    {
+      "epoch": 0.9226814963665677,
+      "grad_norm": 0.6110890507698059,
+      "learning_rate": 2.93529681915643e-07,
+      "loss": 0.3006,
+      "step": 24680
+    },
+    {
+      "epoch": 0.9228684253569175,
+      "grad_norm": 0.4325246810913086,
+      "learning_rate": 2.9211895155188406e-07,
+      "loss": 0.2617,
+      "step": 24685
+    },
+    {
+      "epoch": 0.9230553543472674,
+      "grad_norm": 0.3159783184528351,
+      "learning_rate": 2.9071156915813413e-07,
+      "loss": 0.2363,
+      "step": 24690
+    },
+    {
+      "epoch": 0.9232422833376172,
+      "grad_norm": 0.3180788457393646,
+      "learning_rate": 2.8930753521975496e-07,
+      "loss": 0.2499,
+      "step": 24695
+    },
+    {
+      "epoch": 0.9234292123279669,
+      "grad_norm": 0.4603706896305084,
+      "learning_rate": 2.879068502209625e-07,
+      "loss": 0.2631,
+      "step": 24700
+    },
+    {
+      "epoch": 0.9236161413183167,
+      "grad_norm": 0.5309453010559082,
+      "learning_rate": 2.865095146448105e-07,
+      "loss": 0.2576,
+      "step": 24705
+    },
+    {
+      "epoch": 0.9238030703086665,
+      "grad_norm": 0.5529757738113403,
+      "learning_rate": 2.8511552897319997e-07,
+      "loss": 0.229,
+      "step": 24710
+    },
+    {
+      "epoch": 0.9239899992990163,
+      "grad_norm": 0.38355591893196106,
+      "learning_rate": 2.837248936868764e-07,
+      "loss": 0.2191,
+      "step": 24715
+    },
+    {
+      "epoch": 0.924176928289366,
+      "grad_norm": 0.40068626403808594,
+      "learning_rate": 2.823376092654306e-07,
+      "loss": 0.3506,
+      "step": 24720
+    },
+    {
+      "epoch": 0.9243638572797158,
+      "grad_norm": 0.46175092458724976,
+      "learning_rate": 2.8095367618729664e-07,
+      "loss": 0.2266,
+      "step": 24725
+    },
+    {
+      "epoch": 0.9245507862700657,
+      "grad_norm": 0.4619698226451874,
+      "learning_rate": 2.7957309492975483e-07,
+      "loss": 0.2883,
+      "step": 24730
+    },
+    {
+      "epoch": 0.9247377152604155,
+      "grad_norm": 0.3867049217224121,
+      "learning_rate": 2.781958659689277e-07,
+      "loss": 0.2381,
+      "step": 24735
+    },
+    {
+      "epoch": 0.9249246442507653,
+      "grad_norm": 0.34441909193992615,
+      "learning_rate": 2.768219897797797e-07,
+      "loss": 0.2983,
+      "step": 24740
+    },
+    {
+      "epoch": 0.925111573241115,
+      "grad_norm": 0.1862328201532364,
+      "learning_rate": 2.7545146683612413e-07,
+      "loss": 0.2152,
+      "step": 24745
+    },
+    {
+      "epoch": 0.9252985022314648,
+      "grad_norm": 0.4825479984283447,
+      "learning_rate": 2.7408429761061393e-07,
+      "loss": 0.2617,
+      "step": 24750
+    },
+    {
+      "epoch": 0.9254854312218146,
+      "grad_norm": 0.40720510482788086,
+      "learning_rate": 2.727204825747476e-07,
+      "loss": 0.3691,
+      "step": 24755
+    },
+    {
+      "epoch": 0.9256723602121644,
+      "grad_norm": 0.48773303627967834,
+      "learning_rate": 2.7136002219886326e-07,
+      "loss": 0.3045,
+      "step": 24760
+    },
+    {
+      "epoch": 0.9258592892025141,
+      "grad_norm": 0.5098922848701477,
+      "learning_rate": 2.700029169521479e-07,
+      "loss": 0.2355,
+      "step": 24765
+    },
+    {
+      "epoch": 0.926046218192864,
+      "grad_norm": 0.3420339524745941,
+      "learning_rate": 2.6864916730262593e-07,
+      "loss": 0.2178,
+      "step": 24770
+    },
+    {
+      "epoch": 0.9262331471832138,
+      "grad_norm": 0.8477463722229004,
+      "learning_rate": 2.672987737171673e-07,
+      "loss": 0.2843,
+      "step": 24775
+    },
+    {
+      "epoch": 0.9264200761735636,
+      "grad_norm": 0.481887549161911,
+      "learning_rate": 2.65951736661485e-07,
+      "loss": 0.2758,
+      "step": 24780
+    },
+    {
+      "epoch": 0.9266070051639134,
+      "grad_norm": 0.2788504660129547,
+      "learning_rate": 2.646080566001341e-07,
+      "loss": 0.289,
+      "step": 24785
+    },
+    {
+      "epoch": 0.9267939341542631,
+      "grad_norm": 0.46086716651916504,
+      "learning_rate": 2.632677339965095e-07,
+      "loss": 0.285,
+      "step": 24790
+    },
+    {
+      "epoch": 0.9269808631446129,
+      "grad_norm": 0.7587642669677734,
+      "learning_rate": 2.6193076931285035e-07,
+      "loss": 0.3542,
+      "step": 24795
+    },
+    {
+      "epoch": 0.9271677921349627,
+      "grad_norm": 0.32218098640441895,
+      "learning_rate": 2.6059716301023885e-07,
+      "loss": 0.2192,
+      "step": 24800
+    },
+    {
+      "epoch": 0.9273547211253125,
+      "grad_norm": 0.6089930534362793,
+      "learning_rate": 2.5926691554859497e-07,
+      "loss": 0.2849,
+      "step": 24805
+    },
+    {
+      "epoch": 0.9275416501156624,
+      "grad_norm": 0.8761005401611328,
+      "learning_rate": 2.579400273866861e-07,
+      "loss": 0.2688,
+      "step": 24810
+    },
+    {
+      "epoch": 0.9277285791060121,
+      "grad_norm": 0.4644549787044525,
+      "learning_rate": 2.5661649898211515e-07,
+      "loss": 0.2952,
+      "step": 24815
+    },
+    {
+      "epoch": 0.9279155080963619,
+      "grad_norm": 0.19469016790390015,
+      "learning_rate": 2.552963307913303e-07,
+      "loss": 0.2141,
+      "step": 24820
+    },
+    {
+      "epoch": 0.9281024370867117,
+      "grad_norm": 0.5360405445098877,
+      "learning_rate": 2.5397952326962183e-07,
+      "loss": 0.2797,
+      "step": 24825
+    },
+    {
+      "epoch": 0.9282893660770615,
+      "grad_norm": 0.5923690795898438,
+      "learning_rate": 2.526660768711153e-07,
+      "loss": 0.2603,
+      "step": 24830
+    },
+    {
+      "epoch": 0.9284762950674113,
+      "grad_norm": 0.47409185767173767,
+      "learning_rate": 2.5135599204878403e-07,
+      "loss": 0.2873,
+      "step": 24835
+    },
+    {
+      "epoch": 0.928663224057761,
+      "grad_norm": 0.7585999965667725,
+      "learning_rate": 2.500492692544354e-07,
+      "loss": 0.2297,
+      "step": 24840
+    },
+    {
+      "epoch": 0.9288501530481108,
+      "grad_norm": 0.5913994908332825,
+      "learning_rate": 2.487459089387234e-07,
+      "loss": 0.2202,
+      "step": 24845
+    },
+    {
+      "epoch": 0.9290370820384607,
+      "grad_norm": 0.5756853818893433,
+      "learning_rate": 2.474459115511374e-07,
+      "loss": 0.2813,
+      "step": 24850
+    },
+    {
+      "epoch": 0.9292240110288105,
+      "grad_norm": 0.8577892780303955,
+      "learning_rate": 2.461492775400121e-07,
+      "loss": 0.2344,
+      "step": 24855
+    },
+    {
+      "epoch": 0.9294109400191602,
+      "grad_norm": 0.26759210228919983,
+      "learning_rate": 2.448560073525164e-07,
+      "loss": 0.3403,
+      "step": 24860
+    },
+    {
+      "epoch": 0.92959786900951,
+      "grad_norm": 1.6753215789794922,
+      "learning_rate": 2.4356610143466353e-07,
+      "loss": 0.2956,
+      "step": 24865
+    },
+    {
+      "epoch": 0.9297847979998598,
+      "grad_norm": 0.6404578685760498,
+      "learning_rate": 2.422795602313066e-07,
+      "loss": 0.2465,
+      "step": 24870
+    },
+    {
+      "epoch": 0.9299717269902096,
+      "grad_norm": 0.33225345611572266,
+      "learning_rate": 2.40996384186134e-07,
+      "loss": 0.2499,
+      "step": 24875
+    },
+    {
+      "epoch": 0.9301586559805594,
+      "grad_norm": 0.5298144221305847,
+      "learning_rate": 2.3971657374167956e-07,
+      "loss": 0.2753,
+      "step": 24880
+    },
+    {
+      "epoch": 0.9303455849709091,
+      "grad_norm": 0.3931209444999695,
+      "learning_rate": 2.3844012933930906e-07,
+      "loss": 0.2161,
+      "step": 24885
+    },
+    {
+      "epoch": 0.9305325139612589,
+      "grad_norm": 0.8054724931716919,
+      "learning_rate": 2.371670514192348e-07,
+      "loss": 0.2659,
+      "step": 24890
+    },
+    {
+      "epoch": 0.9307194429516088,
+      "grad_norm": 0.33258944749832153,
+      "learning_rate": 2.358973404205034e-07,
+      "loss": 0.2871,
+      "step": 24895
+    },
+    {
+      "epoch": 0.9309063719419586,
+      "grad_norm": 0.475870281457901,
+      "learning_rate": 2.3463099678100344e-07,
+      "loss": 0.2974,
+      "step": 24900
+    },
+    {
+      "epoch": 0.9310933009323084,
+      "grad_norm": 0.43682655692100525,
+      "learning_rate": 2.333680209374578e-07,
+      "loss": 0.2509,
+      "step": 24905
+    },
+    {
+      "epoch": 0.9312802299226581,
+      "grad_norm": 0.3373750150203705,
+      "learning_rate": 2.3210841332543254e-07,
+      "loss": 0.2143,
+      "step": 24910
+    },
+    {
+      "epoch": 0.9314671589130079,
+      "grad_norm": 0.33678069710731506,
+      "learning_rate": 2.3085217437933127e-07,
+      "loss": 0.2806,
+      "step": 24915
+    },
+    {
+      "epoch": 0.9316540879033577,
+      "grad_norm": 0.47912338376045227,
+      "learning_rate": 2.295993045323941e-07,
+      "loss": 0.2977,
+      "step": 24920
+    },
+    {
+      "epoch": 0.9318410168937075,
+      "grad_norm": 0.3937797248363495,
+      "learning_rate": 2.2834980421669872e-07,
+      "loss": 0.2392,
+      "step": 24925
+    },
+    {
+      "epoch": 0.9320279458840572,
+      "grad_norm": 0.6482548117637634,
+      "learning_rate": 2.2710367386316156e-07,
+      "loss": 0.2579,
+      "step": 24930
+    },
+    {
+      "epoch": 0.9322148748744071,
+      "grad_norm": 1.0093536376953125,
+      "learning_rate": 2.2586091390153996e-07,
+      "loss": 0.3091,
+      "step": 24935
+    },
+    {
+      "epoch": 0.9324018038647569,
+      "grad_norm": 0.29904359579086304,
+      "learning_rate": 2.246215247604233e-07,
+      "loss": 0.3163,
+      "step": 24940
+    },
+    {
+      "epoch": 0.9325887328551067,
+      "grad_norm": 0.6364412903785706,
+      "learning_rate": 2.2338550686724413e-07,
+      "loss": 0.2685,
+      "step": 24945
+    },
+    {
+      "epoch": 0.9327756618454565,
+      "grad_norm": 0.7019013166427612,
+      "learning_rate": 2.221528606482659e-07,
+      "loss": 0.3034,
+      "step": 24950
+    },
+    {
+      "epoch": 0.9329625908358062,
+      "grad_norm": 0.5660467743873596,
+      "learning_rate": 2.2092358652859634e-07,
+      "loss": 0.25,
+      "step": 24955
+    },
+    {
+      "epoch": 0.933149519826156,
+      "grad_norm": 0.5666435956954956,
+      "learning_rate": 2.1969768493217747e-07,
+      "loss": 0.2411,
+      "step": 24960
+    },
+    {
+      "epoch": 0.9333364488165058,
+      "grad_norm": 0.3779709041118622,
+      "learning_rate": 2.1847515628178328e-07,
+      "loss": 0.3184,
+      "step": 24965
+    },
+    {
+      "epoch": 0.9335233778068556,
+      "grad_norm": 0.6057195067405701,
+      "learning_rate": 2.1725600099903433e-07,
+      "loss": 0.23,
+      "step": 24970
+    },
+    {
+      "epoch": 0.9337103067972055,
+      "grad_norm": 0.5047149062156677,
+      "learning_rate": 2.160402195043776e-07,
+      "loss": 0.3242,
+      "step": 24975
+    },
+    {
+      "epoch": 0.9338972357875552,
+      "grad_norm": 1.0450160503387451,
+      "learning_rate": 2.1482781221710437e-07,
+      "loss": 0.2839,
+      "step": 24980
+    },
+    {
+      "epoch": 0.934084164777905,
+      "grad_norm": 0.44621530175209045,
+      "learning_rate": 2.1361877955533682e-07,
+      "loss": 0.2289,
+      "step": 24985
+    },
+    {
+      "epoch": 0.9342710937682548,
+      "grad_norm": 0.547173023223877,
+      "learning_rate": 2.1241312193603814e-07,
+      "loss": 0.3041,
+      "step": 24990
+    },
+    {
+      "epoch": 0.9344580227586046,
+      "grad_norm": 0.4551314413547516,
+      "learning_rate": 2.1121083977500346e-07,
+      "loss": 0.2025,
+      "step": 24995
+    },
+    {
+      "epoch": 0.9346449517489543,
+      "grad_norm": 0.19889132678508759,
+      "learning_rate": 2.1001193348686444e-07,
+      "loss": 0.2547,
+      "step": 25000
+    },
+    {
+      "epoch": 0.9348318807393041,
+      "grad_norm": 0.4394857585430145,
+      "learning_rate": 2.0881640348509258e-07,
+      "loss": 0.3064,
+      "step": 25005
+    },
+    {
+      "epoch": 0.9350188097296539,
+      "grad_norm": 0.29480358958244324,
+      "learning_rate": 2.0762425018199028e-07,
+      "loss": 0.2712,
+      "step": 25010
+    },
+    {
+      "epoch": 0.9352057387200038,
+      "grad_norm": 0.306598037481308,
+      "learning_rate": 2.0643547398869646e-07,
+      "loss": 0.2319,
+      "step": 25015
+    },
+    {
+      "epoch": 0.9353926677103536,
+      "grad_norm": 0.18189042806625366,
+      "learning_rate": 2.052500753151876e-07,
+      "loss": 0.2516,
+      "step": 25020
+    },
+    {
+      "epoch": 0.9355795967007033,
+      "grad_norm": 0.3686217963695526,
+      "learning_rate": 2.0406805457027225e-07,
+      "loss": 0.2465,
+      "step": 25025
+    },
+    {
+      "epoch": 0.9357665256910531,
+      "grad_norm": 0.3795625567436218,
+      "learning_rate": 2.028894121615943e-07,
+      "loss": 0.3719,
+      "step": 25030
+    },
+    {
+      "epoch": 0.9359534546814029,
+      "grad_norm": 0.3678499758243561,
+      "learning_rate": 2.0171414849563753e-07,
+      "loss": 0.2745,
+      "step": 25035
+    },
+    {
+      "epoch": 0.9361403836717527,
+      "grad_norm": 0.20017379522323608,
+      "learning_rate": 2.0054226397771216e-07,
+      "loss": 0.2769,
+      "step": 25040
+    },
+    {
+      "epoch": 0.9363273126621025,
+      "grad_norm": 0.42065608501434326,
+      "learning_rate": 1.9937375901197154e-07,
+      "loss": 0.3136,
+      "step": 25045
+    },
+    {
+      "epoch": 0.9365142416524522,
+      "grad_norm": 0.4758951663970947,
+      "learning_rate": 1.9820863400139778e-07,
+      "loss": 0.2509,
+      "step": 25050
+    },
+    {
+      "epoch": 0.9367011706428021,
+      "grad_norm": 0.30327993631362915,
+      "learning_rate": 1.9704688934780946e-07,
+      "loss": 0.2372,
+      "step": 25055
+    },
+    {
+      "epoch": 0.9368880996331519,
+      "grad_norm": 0.4054654836654663,
+      "learning_rate": 1.9588852545185831e-07,
+      "loss": 0.3371,
+      "step": 25060
+    },
+    {
+      "epoch": 0.9370750286235017,
+      "grad_norm": 0.40502023696899414,
+      "learning_rate": 1.9473354271303258e-07,
+      "loss": 0.2511,
+      "step": 25065
+    },
+    {
+      "epoch": 0.9372619576138514,
+      "grad_norm": 0.4846619665622711,
+      "learning_rate": 1.9358194152965139e-07,
+      "loss": 0.2792,
+      "step": 25070
+    },
+    {
+      "epoch": 0.9374488866042012,
+      "grad_norm": 0.9523441195487976,
+      "learning_rate": 1.9243372229886704e-07,
+      "loss": 0.3035,
+      "step": 25075
+    },
+    {
+      "epoch": 0.937635815594551,
+      "grad_norm": 0.7273759245872498,
+      "learning_rate": 1.9128888541667167e-07,
+      "loss": 0.2516,
+      "step": 25080
+    },
+    {
+      "epoch": 0.9378227445849008,
+      "grad_norm": 0.32927486300468445,
+      "learning_rate": 1.9014743127788392e-07,
+      "loss": 0.2888,
+      "step": 25085
+    },
+    {
+      "epoch": 0.9380096735752506,
+      "grad_norm": 0.31658056378364563,
+      "learning_rate": 1.890093602761589e-07,
+      "loss": 0.2707,
+      "step": 25090
+    },
+    {
+      "epoch": 0.9381966025656004,
+      "grad_norm": 0.41149401664733887,
+      "learning_rate": 1.8787467280398597e-07,
+      "loss": 0.2212,
+      "step": 25095
+    },
+    {
+      "epoch": 0.9383835315559502,
+      "grad_norm": 0.3825901746749878,
+      "learning_rate": 1.8674336925268434e-07,
+      "loss": 0.3109,
+      "step": 25100
+    },
+    {
+      "epoch": 0.9385704605463,
+      "grad_norm": 0.28844115138053894,
+      "learning_rate": 1.8561545001240967e-07,
+      "loss": 0.315,
+      "step": 25105
+    },
+    {
+      "epoch": 0.9387573895366498,
+      "grad_norm": 0.6432269811630249,
+      "learning_rate": 1.844909154721497e-07,
+      "loss": 0.3375,
+      "step": 25110
+    },
+    {
+      "epoch": 0.9389443185269996,
+      "grad_norm": 0.3793524205684662,
+      "learning_rate": 1.83369766019722e-07,
+      "loss": 0.3181,
+      "step": 25115
+    },
+    {
+      "epoch": 0.9391312475173493,
+      "grad_norm": 0.42755433917045593,
+      "learning_rate": 1.8225200204177952e-07,
+      "loss": 0.3514,
+      "step": 25120
+    },
+    {
+      "epoch": 0.9393181765076991,
+      "grad_norm": 0.2594048082828522,
+      "learning_rate": 1.811376239238083e-07,
+      "loss": 0.256,
+      "step": 25125
+    },
+    {
+      "epoch": 0.9395051054980489,
+      "grad_norm": 0.6199470162391663,
+      "learning_rate": 1.8002663205012428e-07,
+      "loss": 0.2888,
+      "step": 25130
+    },
+    {
+      "epoch": 0.9396920344883987,
+      "grad_norm": 0.5659986734390259,
+      "learning_rate": 1.7891902680387652e-07,
+      "loss": 0.2381,
+      "step": 25135
+    },
+    {
+      "epoch": 0.9398789634787486,
+      "grad_norm": 0.31770560145378113,
+      "learning_rate": 1.7781480856704835e-07,
+      "loss": 0.3309,
+      "step": 25140
+    },
+    {
+      "epoch": 0.9400658924690983,
+      "grad_norm": 0.44480013847351074,
+      "learning_rate": 1.7671397772044962e-07,
+      "loss": 0.3511,
+      "step": 25145
+    },
+    {
+      "epoch": 0.9402528214594481,
+      "grad_norm": 0.48626697063446045,
+      "learning_rate": 1.7561653464372885e-07,
+      "loss": 0.2877,
+      "step": 25150
+    },
+    {
+      "epoch": 0.9404397504497979,
+      "grad_norm": 0.4127686023712158,
+      "learning_rate": 1.7452247971535995e-07,
+      "loss": 0.2773,
+      "step": 25155
+    },
+    {
+      "epoch": 0.9406266794401477,
+      "grad_norm": 0.9780619144439697,
+      "learning_rate": 1.7343181331265336e-07,
+      "loss": 0.2247,
+      "step": 25160
+    },
+    {
+      "epoch": 0.9408136084304974,
+      "grad_norm": 0.4334399104118347,
+      "learning_rate": 1.7234453581174704e-07,
+      "loss": 0.2965,
+      "step": 25165
+    },
+    {
+      "epoch": 0.9410005374208472,
+      "grad_norm": 0.37765777111053467,
+      "learning_rate": 1.7126064758761217e-07,
+      "loss": 0.3012,
+      "step": 25170
+    },
+    {
+      "epoch": 0.941187466411197,
+      "grad_norm": 0.258969247341156,
+      "learning_rate": 1.701801490140509e-07,
+      "loss": 0.215,
+      "step": 25175
+    },
+    {
+      "epoch": 0.9413743954015469,
+      "grad_norm": 0.762693464756012,
+      "learning_rate": 1.6910304046369618e-07,
+      "loss": 0.3182,
+      "step": 25180
+    },
+    {
+      "epoch": 0.9415613243918967,
+      "grad_norm": 0.2748546600341797,
+      "learning_rate": 1.6802932230801205e-07,
+      "loss": 0.2761,
+      "step": 25185
+    },
+    {
+      "epoch": 0.9417482533822464,
+      "grad_norm": 0.556158721446991,
+      "learning_rate": 1.669589949172934e-07,
+      "loss": 0.2127,
+      "step": 25190
+    },
+    {
+      "epoch": 0.9419351823725962,
+      "grad_norm": 0.3723142445087433,
+      "learning_rate": 1.6589205866066493e-07,
+      "loss": 0.277,
+      "step": 25195
+    },
+    {
+      "epoch": 0.942122111362946,
+      "grad_norm": 0.42889419198036194,
+      "learning_rate": 1.6482851390608235e-07,
+      "loss": 0.3108,
+      "step": 25200
+    },
+    {
+      "epoch": 0.9423090403532958,
+      "grad_norm": 1.3041778802871704,
+      "learning_rate": 1.6376836102033223e-07,
+      "loss": 0.3281,
+      "step": 25205
+    },
+    {
+      "epoch": 0.9424959693436455,
+      "grad_norm": 0.5724756121635437,
+      "learning_rate": 1.6271160036903099e-07,
+      "loss": 0.3212,
+      "step": 25210
+    },
+    {
+      "epoch": 0.9426828983339953,
+      "grad_norm": 0.2716809809207916,
+      "learning_rate": 1.616582323166249e-07,
+      "loss": 0.2562,
+      "step": 25215
+    },
+    {
+      "epoch": 0.9428698273243452,
+      "grad_norm": 0.5220561027526855,
+      "learning_rate": 1.6060825722639005e-07,
+      "loss": 0.3082,
+      "step": 25220
+    },
+    {
+      "epoch": 0.943056756314695,
+      "grad_norm": 0.3205987811088562,
+      "learning_rate": 1.5956167546043234e-07,
+      "loss": 0.2507,
+      "step": 25225
+    },
+    {
+      "epoch": 0.9432436853050448,
+      "grad_norm": 0.36693549156188965,
+      "learning_rate": 1.5851848737968968e-07,
+      "loss": 0.2144,
+      "step": 25230
+    },
+    {
+      "epoch": 0.9434306142953945,
+      "grad_norm": 0.26295924186706543,
+      "learning_rate": 1.574786933439254e-07,
+      "loss": 0.3514,
+      "step": 25235
+    },
+    {
+      "epoch": 0.9436175432857443,
+      "grad_norm": 0.5648385286331177,
+      "learning_rate": 1.5644229371173714e-07,
+      "loss": 0.3295,
+      "step": 25240
+    },
+    {
+      "epoch": 0.9438044722760941,
+      "grad_norm": 0.43656599521636963,
+      "learning_rate": 1.5540928884054674e-07,
+      "loss": 0.2869,
+      "step": 25245
+    },
+    {
+      "epoch": 0.9439914012664439,
+      "grad_norm": 0.5460593104362488,
+      "learning_rate": 1.5437967908661143e-07,
+      "loss": 0.2633,
+      "step": 25250
+    },
+    {
+      "epoch": 0.9441783302567937,
+      "grad_norm": 0.4732210040092468,
+      "learning_rate": 1.5335346480501056e-07,
+      "loss": 0.2891,
+      "step": 25255
+    },
+    {
+      "epoch": 0.9443652592471435,
+      "grad_norm": 0.2938053011894226,
+      "learning_rate": 1.5233064634965878e-07,
+      "loss": 0.3107,
+      "step": 25260
+    },
+    {
+      "epoch": 0.9445521882374933,
+      "grad_norm": 0.28973618149757385,
+      "learning_rate": 1.513112240732939e-07,
+      "loss": 0.3202,
+      "step": 25265
+    },
+    {
+      "epoch": 0.9447391172278431,
+      "grad_norm": 0.34332942962646484,
+      "learning_rate": 1.5029519832748807e-07,
+      "loss": 0.3028,
+      "step": 25270
+    },
+    {
+      "epoch": 0.9449260462181929,
+      "grad_norm": 0.37853550910949707,
+      "learning_rate": 1.4928256946263875e-07,
+      "loss": 0.3199,
+      "step": 25275
+    },
+    {
+      "epoch": 0.9451129752085426,
+      "grad_norm": 0.8875368237495422,
+      "learning_rate": 1.4827333782797216e-07,
+      "loss": 0.2674,
+      "step": 25280
+    },
+    {
+      "epoch": 0.9452999041988924,
+      "grad_norm": 0.5367876291275024,
+      "learning_rate": 1.472675037715443e-07,
+      "loss": 0.2802,
+      "step": 25285
+    },
+    {
+      "epoch": 0.9454868331892422,
+      "grad_norm": 0.2964636981487274,
+      "learning_rate": 1.4626506764023663e-07,
+      "loss": 0.2168,
+      "step": 25290
+    },
+    {
+      "epoch": 0.945673762179592,
+      "grad_norm": 0.25260117650032043,
+      "learning_rate": 1.4526602977976368e-07,
+      "loss": 0.3053,
+      "step": 25295
+    },
+    {
+      "epoch": 0.9458606911699419,
+      "grad_norm": 0.49495014548301697,
+      "learning_rate": 1.4427039053466207e-07,
+      "loss": 0.2923,
+      "step": 25300
+    },
+    {
+      "epoch": 0.9460476201602916,
+      "grad_norm": 0.4212638735771179,
+      "learning_rate": 1.432781502483005e-07,
+      "loss": 0.2597,
+      "step": 25305
+    },
+    {
+      "epoch": 0.9462345491506414,
+      "grad_norm": 0.5896459817886353,
+      "learning_rate": 1.422893092628741e-07,
+      "loss": 0.2654,
+      "step": 25310
+    },
+    {
+      "epoch": 0.9464214781409912,
+      "grad_norm": 0.9404630661010742,
+      "learning_rate": 1.4130386791940564e-07,
+      "loss": 0.2247,
+      "step": 25315
+    },
+    {
+      "epoch": 0.946608407131341,
+      "grad_norm": 0.4611830413341522,
+      "learning_rate": 1.4032182655774661e-07,
+      "loss": 0.2525,
+      "step": 25320
+    },
+    {
+      "epoch": 0.9467953361216908,
+      "grad_norm": 0.5101840496063232,
+      "learning_rate": 1.3934318551657277e-07,
+      "loss": 0.2983,
+      "step": 25325
+    },
+    {
+      "epoch": 0.9469822651120405,
+      "grad_norm": 0.6482624411582947,
+      "learning_rate": 1.383679451333919e-07,
+      "loss": 0.3196,
+      "step": 25330
+    },
+    {
+      "epoch": 0.9471691941023903,
+      "grad_norm": 0.2823061943054199,
+      "learning_rate": 1.373961057445339e-07,
+      "loss": 0.2113,
+      "step": 25335
+    },
+    {
+      "epoch": 0.9473561230927402,
+      "grad_norm": 0.6522537469863892,
+      "learning_rate": 1.364276676851617e-07,
+      "loss": 0.2637,
+      "step": 25340
+    },
+    {
+      "epoch": 0.94754305208309,
+      "grad_norm": 0.338235080242157,
+      "learning_rate": 1.354626312892582e-07,
+      "loss": 0.2926,
+      "step": 25345
+    },
+    {
+      "epoch": 0.9477299810734398,
+      "grad_norm": 0.3630608022212982,
+      "learning_rate": 1.3450099688963823e-07,
+      "loss": 0.232,
+      "step": 25350
+    },
+    {
+      "epoch": 0.9479169100637895,
+      "grad_norm": 0.32054030895233154,
+      "learning_rate": 1.3354276481794325e-07,
+      "loss": 0.2809,
+      "step": 25355
+    },
+    {
+      "epoch": 0.9481038390541393,
+      "grad_norm": 0.40766042470932007,
+      "learning_rate": 1.3258793540463778e-07,
+      "loss": 0.2871,
+      "step": 25360
+    },
+    {
+      "epoch": 0.9482907680444891,
+      "grad_norm": 0.6181418299674988,
+      "learning_rate": 1.3163650897901724e-07,
+      "loss": 0.3037,
+      "step": 25365
+    },
+    {
+      "epoch": 0.9484776970348389,
+      "grad_norm": 0.39881226420402527,
+      "learning_rate": 1.3068848586920035e-07,
+      "loss": 0.27,
+      "step": 25370
+    },
+    {
+      "epoch": 0.9486646260251886,
+      "grad_norm": 0.548456609249115,
+      "learning_rate": 1.2974386640213333e-07,
+      "loss": 0.2401,
+      "step": 25375
+    },
+    {
+      "epoch": 0.9488515550155384,
+      "grad_norm": 0.3512159585952759,
+      "learning_rate": 1.2880265090358668e-07,
+      "loss": 0.3958,
+      "step": 25380
+    },
+    {
+      "epoch": 0.9490384840058883,
+      "grad_norm": 0.33197692036628723,
+      "learning_rate": 1.278648396981619e-07,
+      "loss": 0.2512,
+      "step": 25385
+    },
+    {
+      "epoch": 0.9492254129962381,
+      "grad_norm": 0.4836597144603729,
+      "learning_rate": 1.2693043310928022e-07,
+      "loss": 0.3093,
+      "step": 25390
+    },
+    {
+      "epoch": 0.9494123419865879,
+      "grad_norm": 0.3625621795654297,
+      "learning_rate": 1.2599943145919392e-07,
+      "loss": 0.3484,
+      "step": 25395
+    },
+    {
+      "epoch": 0.9495992709769376,
+      "grad_norm": 0.41878730058670044,
+      "learning_rate": 1.2507183506897723e-07,
+      "loss": 0.267,
+      "step": 25400
+    },
+    {
+      "epoch": 0.9497861999672874,
+      "grad_norm": 0.2932325303554535,
+      "learning_rate": 1.241476442585321e-07,
+      "loss": 0.2598,
+      "step": 25405
+    },
+    {
+      "epoch": 0.9499731289576372,
+      "grad_norm": 0.3391314744949341,
+      "learning_rate": 1.2322685934658573e-07,
+      "loss": 0.2433,
+      "step": 25410
+    },
+    {
+      "epoch": 0.950160057947987,
+      "grad_norm": 0.4093460142612457,
+      "learning_rate": 1.223094806506897e-07,
+      "loss": 0.2741,
+      "step": 25415
+    },
+    {
+      "epoch": 0.9503469869383367,
+      "grad_norm": 0.4316021800041199,
+      "learning_rate": 1.2139550848722203e-07,
+      "loss": 0.2655,
+      "step": 25420
+    },
+    {
+      "epoch": 0.9505339159286866,
+      "grad_norm": 0.7140107750892639,
+      "learning_rate": 1.2048494317138615e-07,
+      "loss": 0.3403,
+      "step": 25425
+    },
+    {
+      "epoch": 0.9507208449190364,
+      "grad_norm": 0.5122732520103455,
+      "learning_rate": 1.1957778501720973e-07,
+      "loss": 0.2101,
+      "step": 25430
+    },
+    {
+      "epoch": 0.9509077739093862,
+      "grad_norm": 1.210839867591858,
+      "learning_rate": 1.1867403433754476e-07,
+      "loss": 0.2857,
+      "step": 25435
+    },
+    {
+      "epoch": 0.951094702899736,
+      "grad_norm": 0.507588267326355,
+      "learning_rate": 1.177736914440697e-07,
+      "loss": 0.2956,
+      "step": 25440
+    },
+    {
+      "epoch": 0.9512816318900857,
+      "grad_norm": 0.4931941330432892,
+      "learning_rate": 1.1687675664728837e-07,
+      "loss": 0.2871,
+      "step": 25445
+    },
+    {
+      "epoch": 0.9514685608804355,
+      "grad_norm": 0.5191053748130798,
+      "learning_rate": 1.1598323025652447e-07,
+      "loss": 0.2739,
+      "step": 25450
+    },
+    {
+      "epoch": 0.9516554898707853,
+      "grad_norm": 0.37645572423934937,
+      "learning_rate": 1.150931125799315e-07,
+      "loss": 0.2422,
+      "step": 25455
+    },
+    {
+      "epoch": 0.9518424188611351,
+      "grad_norm": 0.46614864468574524,
+      "learning_rate": 1.1420640392448612e-07,
+      "loss": 0.2909,
+      "step": 25460
+    },
+    {
+      "epoch": 0.952029347851485,
+      "grad_norm": 0.3071075677871704,
+      "learning_rate": 1.1332310459598928e-07,
+      "loss": 0.1938,
+      "step": 25465
+    },
+    {
+      "epoch": 0.9522162768418347,
+      "grad_norm": 0.4759249985218048,
+      "learning_rate": 1.1244321489906285e-07,
+      "loss": 0.2831,
+      "step": 25470
+    },
+    {
+      "epoch": 0.9524032058321845,
+      "grad_norm": 0.4241246283054352,
+      "learning_rate": 1.1156673513715744e-07,
+      "loss": 0.2166,
+      "step": 25475
+    },
+    {
+      "epoch": 0.9525901348225343,
+      "grad_norm": 0.26274579763412476,
+      "learning_rate": 1.1069366561254679e-07,
+      "loss": 0.3136,
+      "step": 25480
+    },
+    {
+      "epoch": 0.9527770638128841,
+      "grad_norm": 0.3445034325122833,
+      "learning_rate": 1.0982400662632564e-07,
+      "loss": 0.2326,
+      "step": 25485
+    },
+    {
+      "epoch": 0.9529639928032339,
+      "grad_norm": 0.46550294756889343,
+      "learning_rate": 1.0895775847841516e-07,
+      "loss": 0.2672,
+      "step": 25490
+    },
+    {
+      "epoch": 0.9531509217935836,
+      "grad_norm": 0.3357786238193512,
+      "learning_rate": 1.0809492146755973e-07,
+      "loss": 0.2343,
+      "step": 25495
+    },
+    {
+      "epoch": 0.9533378507839334,
+      "grad_norm": 0.1831052303314209,
+      "learning_rate": 1.0723549589132687e-07,
+      "loss": 0.3252,
+      "step": 25500
+    },
+    {
+      "epoch": 0.9535247797742833,
+      "grad_norm": 0.726946234703064,
+      "learning_rate": 1.0637948204610837e-07,
+      "loss": 0.3023,
+      "step": 25505
+    },
+    {
+      "epoch": 0.9537117087646331,
+      "grad_norm": 0.5573680996894836,
+      "learning_rate": 1.055268802271181e-07,
+      "loss": 0.2786,
+      "step": 25510
+    },
+    {
+      "epoch": 0.9538986377549828,
+      "grad_norm": 0.5966600179672241,
+      "learning_rate": 1.0467769072839307e-07,
+      "loss": 0.3758,
+      "step": 25515
+    },
+    {
+      "epoch": 0.9540855667453326,
+      "grad_norm": 0.31094256043434143,
+      "learning_rate": 1.0383191384279789e-07,
+      "loss": 0.2691,
+      "step": 25520
+    },
+    {
+      "epoch": 0.9542724957356824,
+      "grad_norm": 0.6314263343811035,
+      "learning_rate": 1.029895498620126e-07,
+      "loss": 0.3314,
+      "step": 25525
+    },
+    {
+      "epoch": 0.9544594247260322,
+      "grad_norm": 0.6387194991111755,
+      "learning_rate": 1.0215059907654811e-07,
+      "loss": 0.2326,
+      "step": 25530
+    },
+    {
+      "epoch": 0.954646353716382,
+      "grad_norm": 0.27241405844688416,
+      "learning_rate": 1.013150617757308e-07,
+      "loss": 0.2931,
+      "step": 25535
+    },
+    {
+      "epoch": 0.9548332827067317,
+      "grad_norm": 0.6504563093185425,
+      "learning_rate": 1.0048293824771682e-07,
+      "loss": 0.3224,
+      "step": 25540
+    },
+    {
+      "epoch": 0.9550202116970816,
+      "grad_norm": 0.5727891325950623,
+      "learning_rate": 9.965422877948106e-08,
+      "loss": 0.2353,
+      "step": 25545
+    },
+    {
+      "epoch": 0.9552071406874314,
+      "grad_norm": 0.5274751782417297,
+      "learning_rate": 9.882893365681934e-08,
+      "loss": 0.3013,
+      "step": 25550
+    },
+    {
+      "epoch": 0.9553940696777812,
+      "grad_norm": 0.41574156284332275,
+      "learning_rate": 9.80070531643551e-08,
+      "loss": 0.2363,
+      "step": 25555
+    },
+    {
+      "epoch": 0.955580998668131,
+      "grad_norm": 0.5117254257202148,
+      "learning_rate": 9.71885875855294e-08,
+      "loss": 0.3033,
+      "step": 25560
+    },
+    {
+      "epoch": 0.9557679276584807,
+      "grad_norm": 0.5276907086372375,
+      "learning_rate": 9.637353720260867e-08,
+      "loss": 0.3117,
+      "step": 25565
+    },
+    {
+      "epoch": 0.9559548566488305,
+      "grad_norm": 0.6600742340087891,
+      "learning_rate": 9.556190229668027e-08,
+      "loss": 0.3827,
+      "step": 25570
+    },
+    {
+      "epoch": 0.9561417856391803,
+      "grad_norm": 0.43783825635910034,
+      "learning_rate": 9.475368314765365e-08,
+      "loss": 0.3325,
+      "step": 25575
+    },
+    {
+      "epoch": 0.9563287146295301,
+      "grad_norm": 0.4194599390029907,
+      "learning_rate": 9.394888003426028e-08,
+      "loss": 0.2532,
+      "step": 25580
+    },
+    {
+      "epoch": 0.95651564361988,
+      "grad_norm": 0.3996281921863556,
+      "learning_rate": 9.314749323405481e-08,
+      "loss": 0.2825,
+      "step": 25585
+    },
+    {
+      "epoch": 0.9567025726102297,
+      "grad_norm": 0.4395323395729065,
+      "learning_rate": 9.234952302341172e-08,
+      "loss": 0.2375,
+      "step": 25590
+    },
+    {
+      "epoch": 0.9568895016005795,
+      "grad_norm": 0.5239723920822144,
+      "learning_rate": 9.155496967752642e-08,
+      "loss": 0.3791,
+      "step": 25595
+    },
+    {
+      "epoch": 0.9570764305909293,
+      "grad_norm": 0.6656010746955872,
+      "learning_rate": 9.076383347042084e-08,
+      "loss": 0.3139,
+      "step": 25600
+    },
+    {
+      "epoch": 0.9572633595812791,
+      "grad_norm": 0.6281600594520569,
+      "learning_rate": 8.997611467493228e-08,
+      "loss": 0.3157,
+      "step": 25605
+    },
+    {
+      "epoch": 0.9574502885716288,
+      "grad_norm": 0.39084675908088684,
+      "learning_rate": 8.919181356272454e-08,
+      "loss": 0.2807,
+      "step": 25610
+    },
+    {
+      "epoch": 0.9576372175619786,
+      "grad_norm": 0.4941912293434143,
+      "learning_rate": 8.841093040427907e-08,
+      "loss": 0.2826,
+      "step": 25615
+    },
+    {
+      "epoch": 0.9578241465523284,
+      "grad_norm": 0.500826358795166,
+      "learning_rate": 8.76334654689015e-08,
+      "loss": 0.3496,
+      "step": 25620
+    },
+    {
+      "epoch": 0.9580110755426782,
+      "grad_norm": 0.6868979930877686,
+      "learning_rate": 8.685941902471628e-08,
+      "loss": 0.3204,
+      "step": 25625
+    },
+    {
+      "epoch": 0.9581980045330281,
+      "grad_norm": 0.30598723888397217,
+      "learning_rate": 8.608879133866988e-08,
+      "loss": 0.3358,
+      "step": 25630
+    },
+    {
+      "epoch": 0.9583849335233778,
+      "grad_norm": 0.3103514611721039,
+      "learning_rate": 8.53215826765308e-08,
+      "loss": 0.285,
+      "step": 25635
+    },
+    {
+      "epoch": 0.9585718625137276,
+      "grad_norm": 0.5536108016967773,
+      "learning_rate": 8.455779330288516e-08,
+      "loss": 0.3143,
+      "step": 25640
+    },
+    {
+      "epoch": 0.9587587915040774,
+      "grad_norm": 0.589036226272583,
+      "learning_rate": 8.379742348114339e-08,
+      "loss": 0.286,
+      "step": 25645
+    },
+    {
+      "epoch": 0.9589457204944272,
+      "grad_norm": 0.2255747765302658,
+      "learning_rate": 8.30404734735346e-08,
+      "loss": 0.26,
+      "step": 25650
+    },
+    {
+      "epoch": 0.959132649484777,
+      "grad_norm": 0.5313140749931335,
+      "learning_rate": 8.228694354111111e-08,
+      "loss": 0.2988,
+      "step": 25655
+    },
+    {
+      "epoch": 0.9593195784751267,
+      "grad_norm": 0.30977725982666016,
+      "learning_rate": 8.153683394374057e-08,
+      "loss": 0.2458,
+      "step": 25660
+    },
+    {
+      "epoch": 0.9595065074654765,
+      "grad_norm": 0.8567424416542053,
+      "learning_rate": 8.079014494011827e-08,
+      "loss": 0.2961,
+      "step": 25665
+    },
+    {
+      "epoch": 0.9596934364558264,
+      "grad_norm": 0.569592297077179,
+      "learning_rate": 8.004687678775158e-08,
+      "loss": 0.2552,
+      "step": 25670
+    },
+    {
+      "epoch": 0.9598803654461762,
+      "grad_norm": 0.2618090808391571,
+      "learning_rate": 7.930702974297544e-08,
+      "loss": 0.2402,
+      "step": 25675
+    },
+    {
+      "epoch": 0.9600672944365259,
+      "grad_norm": 0.5399622917175293,
+      "learning_rate": 7.85706040609413e-08,
+      "loss": 0.2917,
+      "step": 25680
+    },
+    {
+      "epoch": 0.9602542234268757,
+      "grad_norm": 0.4389611780643463,
+      "learning_rate": 7.783759999562046e-08,
+      "loss": 0.2601,
+      "step": 25685
+    },
+    {
+      "epoch": 0.9604411524172255,
+      "grad_norm": 0.3556120693683624,
+      "learning_rate": 7.710801779980514e-08,
+      "loss": 0.3564,
+      "step": 25690
+    },
+    {
+      "epoch": 0.9606280814075753,
+      "grad_norm": 0.3294890522956848,
+      "learning_rate": 7.638185772510854e-08,
+      "loss": 0.2801,
+      "step": 25695
+    },
+    {
+      "epoch": 0.960815010397925,
+      "grad_norm": 0.37796953320503235,
+      "learning_rate": 7.565912002196141e-08,
+      "loss": 0.2577,
+      "step": 25700
+    },
+    {
+      "epoch": 0.9610019393882748,
+      "grad_norm": 0.4705111086368561,
+      "learning_rate": 7.493980493961439e-08,
+      "loss": 0.2638,
+      "step": 25705
+    },
+    {
+      "epoch": 0.9611888683786247,
+      "grad_norm": 0.2994195520877838,
+      "learning_rate": 7.422391272614016e-08,
+      "loss": 0.2883,
+      "step": 25710
+    },
+    {
+      "epoch": 0.9613757973689745,
+      "grad_norm": 0.2776239812374115,
+      "learning_rate": 7.351144362842898e-08,
+      "loss": 0.2938,
+      "step": 25715
+    },
+    {
+      "epoch": 0.9615627263593243,
+      "grad_norm": 0.4236242175102234,
+      "learning_rate": 7.280239789219213e-08,
+      "loss": 0.2281,
+      "step": 25720
+    },
+    {
+      "epoch": 0.961749655349674,
+      "grad_norm": 0.7493228912353516,
+      "learning_rate": 7.209677576195617e-08,
+      "loss": 0.3591,
+      "step": 25725
+    },
+    {
+      "epoch": 0.9619365843400238,
+      "grad_norm": 0.598604679107666,
+      "learning_rate": 7.139457748107314e-08,
+      "loss": 0.2865,
+      "step": 25730
+    },
+    {
+      "epoch": 0.9621235133303736,
+      "grad_norm": 0.147347092628479,
+      "learning_rate": 7.069580329170933e-08,
+      "loss": 0.302,
+      "step": 25735
+    },
+    {
+      "epoch": 0.9623104423207234,
+      "grad_norm": 0.5269376635551453,
+      "learning_rate": 7.000045343485306e-08,
+      "loss": 0.3699,
+      "step": 25740
+    },
+    {
+      "epoch": 0.9624973713110732,
+      "grad_norm": 0.30106881260871887,
+      "learning_rate": 6.93085281503092e-08,
+      "loss": 0.2679,
+      "step": 25745
+    },
+    {
+      "epoch": 0.962684300301423,
+      "grad_norm": 0.6084346771240234,
+      "learning_rate": 6.862002767670351e-08,
+      "loss": 0.288,
+      "step": 25750
+    },
+    {
+      "epoch": 0.9628712292917728,
+      "grad_norm": 0.31380781531333923,
+      "learning_rate": 6.793495225148161e-08,
+      "loss": 0.2513,
+      "step": 25755
+    },
+    {
+      "epoch": 0.9630581582821226,
+      "grad_norm": 0.5058198571205139,
+      "learning_rate": 6.725330211090342e-08,
+      "loss": 0.2661,
+      "step": 25760
+    },
+    {
+      "epoch": 0.9632450872724724,
+      "grad_norm": 0.3875492811203003,
+      "learning_rate": 6.65750774900531e-08,
+      "loss": 0.2098,
+      "step": 25765
+    },
+    {
+      "epoch": 0.9634320162628222,
+      "grad_norm": 0.34512069821357727,
+      "learning_rate": 6.59002786228291e-08,
+      "loss": 0.2959,
+      "step": 25770
+    },
+    {
+      "epoch": 0.9636189452531719,
+      "grad_norm": 0.5447210669517517,
+      "learning_rate": 6.522890574195195e-08,
+      "loss": 0.2606,
+      "step": 25775
+    },
+    {
+      "epoch": 0.9638058742435217,
+      "grad_norm": 0.34275734424591064,
+      "learning_rate": 6.456095907895754e-08,
+      "loss": 0.3135,
+      "step": 25780
+    },
+    {
+      "epoch": 0.9639928032338715,
+      "grad_norm": 0.5490571856498718,
+      "learning_rate": 6.389643886420161e-08,
+      "loss": 0.2562,
+      "step": 25785
+    },
+    {
+      "epoch": 0.9641797322242214,
+      "grad_norm": 0.921448290348053,
+      "learning_rate": 6.323534532685971e-08,
+      "loss": 0.3327,
+      "step": 25790
+    },
+    {
+      "epoch": 0.9643666612145712,
+      "grad_norm": 0.8506829738616943,
+      "learning_rate": 6.25776786949217e-08,
+      "loss": 0.238,
+      "step": 25795
+    },
+    {
+      "epoch": 0.9645535902049209,
+      "grad_norm": 0.24562324583530426,
+      "learning_rate": 6.192343919519949e-08,
+      "loss": 0.2228,
+      "step": 25800
+    },
+    {
+      "epoch": 0.9647405191952707,
+      "grad_norm": 0.4626258909702301,
+      "learning_rate": 6.127262705332148e-08,
+      "loss": 0.2939,
+      "step": 25805
+    },
+    {
+      "epoch": 0.9649274481856205,
+      "grad_norm": 0.46563923358917236,
+      "learning_rate": 6.06252424937337e-08,
+      "loss": 0.2838,
+      "step": 25810
+    },
+    {
+      "epoch": 0.9651143771759703,
+      "grad_norm": 0.3600118160247803,
+      "learning_rate": 5.998128573969975e-08,
+      "loss": 0.2402,
+      "step": 25815
+    },
+    {
+      "epoch": 0.96530130616632,
+      "grad_norm": 0.25077444314956665,
+      "learning_rate": 5.9340757013304215e-08,
+      "loss": 0.261,
+      "step": 25820
+    },
+    {
+      "epoch": 0.9654882351566698,
+      "grad_norm": 0.4118346869945526,
+      "learning_rate": 5.8703656535444853e-08,
+      "loss": 0.3058,
+      "step": 25825
+    },
+    {
+      "epoch": 0.9656751641470197,
+      "grad_norm": 0.5594443678855896,
+      "learning_rate": 5.806998452584034e-08,
+      "loss": 0.2788,
+      "step": 25830
+    },
+    {
+      "epoch": 0.9658620931373695,
+      "grad_norm": 0.28521931171417236,
+      "learning_rate": 5.743974120302587e-08,
+      "loss": 0.2705,
+      "step": 25835
+    },
+    {
+      "epoch": 0.9660490221277193,
+      "grad_norm": 0.44842952489852905,
+      "learning_rate": 5.681292678435424e-08,
+      "loss": 0.2366,
+      "step": 25840
+    },
+    {
+      "epoch": 0.966235951118069,
+      "grad_norm": 0.6687315702438354,
+      "learning_rate": 5.618954148599587e-08,
+      "loss": 0.2382,
+      "step": 25845
+    },
+    {
+      "epoch": 0.9664228801084188,
+      "grad_norm": 0.3889986574649811,
+      "learning_rate": 5.556958552293878e-08,
+      "loss": 0.2423,
+      "step": 25850
+    },
+    {
+      "epoch": 0.9666098090987686,
+      "grad_norm": 0.41925016045570374,
+      "learning_rate": 5.4953059108987516e-08,
+      "loss": 0.2204,
+      "step": 25855
+    },
+    {
+      "epoch": 0.9667967380891184,
+      "grad_norm": 0.502301037311554,
+      "learning_rate": 5.4339962456763096e-08,
+      "loss": 0.2187,
+      "step": 25860
+    },
+    {
+      "epoch": 0.9669836670794681,
+      "grad_norm": 0.5296496748924255,
+      "learning_rate": 5.37302957777075e-08,
+      "loss": 0.2298,
+      "step": 25865
+    },
+    {
+      "epoch": 0.9671705960698179,
+      "grad_norm": 0.34486472606658936,
+      "learning_rate": 5.3124059282076975e-08,
+      "loss": 0.2691,
+      "step": 25870
+    },
+    {
+      "epoch": 0.9673575250601678,
+      "grad_norm": 0.41491997241973877,
+      "learning_rate": 5.2521253178944295e-08,
+      "loss": 0.2574,
+      "step": 25875
+    },
+    {
+      "epoch": 0.9675444540505176,
+      "grad_norm": 0.4363487660884857,
+      "learning_rate": 5.192187767619872e-08,
+      "loss": 0.2644,
+      "step": 25880
+    },
+    {
+      "epoch": 0.9677313830408674,
+      "grad_norm": 0.45662906765937805,
+      "learning_rate": 5.1325932980550444e-08,
+      "loss": 0.2408,
+      "step": 25885
+    },
+    {
+      "epoch": 0.9679183120312171,
+      "grad_norm": 0.5905013680458069,
+      "learning_rate": 5.073341929752174e-08,
+      "loss": 0.2843,
+      "step": 25890
+    },
+    {
+      "epoch": 0.9681052410215669,
+      "grad_norm": 0.43980005383491516,
+      "learning_rate": 5.0144336831453586e-08,
+      "loss": 0.2896,
+      "step": 25895
+    },
+    {
+      "epoch": 0.9682921700119167,
+      "grad_norm": 0.6901749968528748,
+      "learning_rate": 4.95586857855046e-08,
+      "loss": 0.3044,
+      "step": 25900
+    },
+    {
+      "epoch": 0.9684790990022665,
+      "grad_norm": 0.9550485014915466,
+      "learning_rate": 4.897646636164877e-08,
+      "loss": 0.2959,
+      "step": 25905
+    },
+    {
+      "epoch": 0.9686660279926163,
+      "grad_norm": 0.13232779502868652,
+      "learning_rate": 4.839767876067658e-08,
+      "loss": 0.271,
+      "step": 25910
+    },
+    {
+      "epoch": 0.9688529569829661,
+      "grad_norm": 0.3293937146663666,
+      "learning_rate": 4.782232318219615e-08,
+      "loss": 0.2995,
+      "step": 25915
+    },
+    {
+      "epoch": 0.9690398859733159,
+      "grad_norm": 0.4516461491584778,
+      "learning_rate": 4.7250399824629867e-08,
+      "loss": 0.3644,
+      "step": 25920
+    },
+    {
+      "epoch": 0.9692268149636657,
+      "grad_norm": 0.3880869746208191,
+      "learning_rate": 4.668190888521884e-08,
+      "loss": 0.3003,
+      "step": 25925
+    },
+    {
+      "epoch": 0.9694137439540155,
+      "grad_norm": 0.3466978967189789,
+      "learning_rate": 4.611685056001847e-08,
+      "loss": 0.2949,
+      "step": 25930
+    },
+    {
+      "epoch": 0.9696006729443652,
+      "grad_norm": 0.4129624664783478,
+      "learning_rate": 4.555522504390175e-08,
+      "loss": 0.2601,
+      "step": 25935
+    },
+    {
+      "epoch": 0.969787601934715,
+      "grad_norm": 0.30791693925857544,
+      "learning_rate": 4.499703253055709e-08,
+      "loss": 0.2434,
+      "step": 25940
+    },
+    {
+      "epoch": 0.9699745309250648,
+      "grad_norm": 0.7145932912826538,
+      "learning_rate": 4.4442273212488286e-08,
+      "loss": 0.2478,
+      "step": 25945
+    },
+    {
+      "epoch": 0.9701614599154146,
+      "grad_norm": 0.359036386013031,
+      "learning_rate": 4.3890947281016725e-08,
+      "loss": 0.2566,
+      "step": 25950
+    },
+    {
+      "epoch": 0.9703483889057645,
+      "grad_norm": 0.4460180401802063,
+      "learning_rate": 4.3343054926279216e-08,
+      "loss": 0.2508,
+      "step": 25955
+    },
+    {
+      "epoch": 0.9705353178961142,
+      "grad_norm": 0.38728684186935425,
+      "learning_rate": 4.279859633722794e-08,
+      "loss": 0.2721,
+      "step": 25960
+    },
+    {
+      "epoch": 0.970722246886464,
+      "grad_norm": 0.4190647602081299,
+      "learning_rate": 4.225757170163047e-08,
+      "loss": 0.3867,
+      "step": 25965
+    },
+    {
+      "epoch": 0.9709091758768138,
+      "grad_norm": 0.37819820642471313,
+      "learning_rate": 4.1719981206072015e-08,
+      "loss": 0.2931,
+      "step": 25970
+    },
+    {
+      "epoch": 0.9710961048671636,
+      "grad_norm": 0.620923638343811,
+      "learning_rate": 4.118582503594981e-08,
+      "loss": 0.2239,
+      "step": 25975
+    },
+    {
+      "epoch": 0.9712830338575134,
+      "grad_norm": 0.4498381018638611,
+      "learning_rate": 4.065510337548206e-08,
+      "loss": 0.338,
+      "step": 25980
+    },
+    {
+      "epoch": 0.9714699628478631,
+      "grad_norm": 0.5090503096580505,
+      "learning_rate": 4.0127816407696805e-08,
+      "loss": 0.2363,
+      "step": 25985
+    },
+    {
+      "epoch": 0.9716568918382129,
+      "grad_norm": 0.3429947793483734,
+      "learning_rate": 3.96039643144408e-08,
+      "loss": 0.263,
+      "step": 25990
+    },
+    {
+      "epoch": 0.9718438208285628,
+      "grad_norm": 0.392718642950058,
+      "learning_rate": 3.90835472763762e-08,
+      "loss": 0.2374,
+      "step": 25995
+    },
+    {
+      "epoch": 0.9720307498189126,
+      "grad_norm": 0.3899400532245636,
+      "learning_rate": 3.8566565472980545e-08,
+      "loss": 0.29,
+      "step": 26000
+    },
+    {
+      "epoch": 0.9722176788092624,
+      "grad_norm": 0.561836302280426,
+      "learning_rate": 3.805301908254455e-08,
+      "loss": 0.2512,
+      "step": 26005
+    },
+    {
+      "epoch": 0.9724046077996121,
+      "grad_norm": 0.6161816120147705,
+      "learning_rate": 3.754290828217655e-08,
+      "loss": 0.1973,
+      "step": 26010
+    },
+    {
+      "epoch": 0.9725915367899619,
+      "grad_norm": 1.0645028352737427,
+      "learning_rate": 3.7036233247799144e-08,
+      "loss": 0.4018,
+      "step": 26015
+    },
+    {
+      "epoch": 0.9727784657803117,
+      "grad_norm": 0.5088971257209778,
+      "learning_rate": 3.6532994154150347e-08,
+      "loss": 0.2651,
+      "step": 26020
+    },
+    {
+      "epoch": 0.9729653947706615,
+      "grad_norm": 0.3199444115161896,
+      "learning_rate": 3.603319117478244e-08,
+      "loss": 0.2447,
+      "step": 26025
+    },
+    {
+      "epoch": 0.9731523237610112,
+      "grad_norm": 0.3997444808483124,
+      "learning_rate": 3.5536824482061974e-08,
+      "loss": 0.2706,
+      "step": 26030
+    },
+    {
+      "epoch": 0.9733392527513611,
+      "grad_norm": 0.22671836614608765,
+      "learning_rate": 3.504389424717314e-08,
+      "loss": 0.3388,
+      "step": 26035
+    },
+    {
+      "epoch": 0.9735261817417109,
+      "grad_norm": 0.36706823110580444,
+      "learning_rate": 3.455440064011328e-08,
+      "loss": 0.2259,
+      "step": 26040
+    },
+    {
+      "epoch": 0.9737131107320607,
+      "grad_norm": 0.4109925329685211,
+      "learning_rate": 3.406834382969515e-08,
+      "loss": 0.1845,
+      "step": 26045
+    },
+    {
+      "epoch": 0.9739000397224105,
+      "grad_norm": 0.47747719287872314,
+      "learning_rate": 3.358572398354465e-08,
+      "loss": 0.2509,
+      "step": 26050
+    },
+    {
+      "epoch": 0.9740869687127602,
+      "grad_norm": 0.1783730685710907,
+      "learning_rate": 3.310654126810309e-08,
+      "loss": 0.273,
+      "step": 26055
+    },
+    {
+      "epoch": 0.97427389770311,
+      "grad_norm": 0.3882327079772949,
+      "learning_rate": 3.263079584862938e-08,
+      "loss": 0.2216,
+      "step": 26060
+    },
+    {
+      "epoch": 0.9744608266934598,
+      "grad_norm": 0.3611951768398285,
+      "learning_rate": 3.2158487889192294e-08,
+      "loss": 0.2725,
+      "step": 26065
+    },
+    {
+      "epoch": 0.9746477556838096,
+      "grad_norm": 0.5424373149871826,
+      "learning_rate": 3.168961755267819e-08,
+      "loss": 0.242,
+      "step": 26070
+    },
+    {
+      "epoch": 0.9748346846741595,
+      "grad_norm": 0.6144366264343262,
+      "learning_rate": 3.12241850007855e-08,
+      "loss": 0.2658,
+      "step": 26075
+    },
+    {
+      "epoch": 0.9750216136645092,
+      "grad_norm": 0.17419812083244324,
+      "learning_rate": 3.076219039403139e-08,
+      "loss": 0.2147,
+      "step": 26080
+    },
+    {
+      "epoch": 0.975208542654859,
+      "grad_norm": 0.23372456431388855,
+      "learning_rate": 3.0303633891742844e-08,
+      "loss": 0.2811,
+      "step": 26085
+    },
+    {
+      "epoch": 0.9753954716452088,
+      "grad_norm": 0.1994643360376358,
+      "learning_rate": 2.984851565206226e-08,
+      "loss": 0.3354,
+      "step": 26090
+    },
+    {
+      "epoch": 0.9755824006355586,
+      "grad_norm": 0.45933806896209717,
+      "learning_rate": 2.9396835831947412e-08,
+      "loss": 0.3562,
+      "step": 26095
+    },
+    {
+      "epoch": 0.9757693296259083,
+      "grad_norm": 0.4124797284603119,
+      "learning_rate": 2.8948594587170366e-08,
+      "loss": 0.2709,
+      "step": 26100
+    },
+    {
+      "epoch": 0.9759562586162581,
+      "grad_norm": 0.9764533638954163,
+      "learning_rate": 2.850379207231746e-08,
+      "loss": 0.4274,
+      "step": 26105
+    },
+    {
+      "epoch": 0.9761431876066079,
+      "grad_norm": 0.3561939001083374,
+      "learning_rate": 2.8062428440785994e-08,
+      "loss": 0.2174,
+      "step": 26110
+    },
+    {
+      "epoch": 0.9763301165969577,
+      "grad_norm": 0.35872069001197815,
+      "learning_rate": 2.7624503844790872e-08,
+      "loss": 0.2391,
+      "step": 26115
+    },
+    {
+      "epoch": 0.9765170455873076,
+      "grad_norm": 0.20135623216629028,
+      "learning_rate": 2.7190018435360178e-08,
+      "loss": 0.2471,
+      "step": 26120
+    },
+    {
+      "epoch": 0.9767039745776573,
+      "grad_norm": 0.866783082485199,
+      "learning_rate": 2.6758972362334046e-08,
+      "loss": 0.3307,
+      "step": 26125
+    },
+    {
+      "epoch": 0.9768909035680071,
+      "grad_norm": 0.34653496742248535,
+      "learning_rate": 2.633136577436912e-08,
+      "loss": 0.2525,
+      "step": 26130
+    },
+    {
+      "epoch": 0.9770778325583569,
+      "grad_norm": 0.3317651152610779,
+      "learning_rate": 2.590719881893522e-08,
+      "loss": 0.2919,
+      "step": 26135
+    },
+    {
+      "epoch": 0.9772647615487067,
+      "grad_norm": 0.22476986050605774,
+      "learning_rate": 2.5486471642314215e-08,
+      "loss": 0.3969,
+      "step": 26140
+    },
+    {
+      "epoch": 0.9774516905390565,
+      "grad_norm": 0.3792192339897156,
+      "learning_rate": 2.5069184389602264e-08,
+      "loss": 0.2839,
+      "step": 26145
+    },
+    {
+      "epoch": 0.9776386195294062,
+      "grad_norm": 0.5724767446517944,
+      "learning_rate": 2.4655337204712027e-08,
+      "loss": 0.2301,
+      "step": 26150
+    },
+    {
+      "epoch": 0.977825548519756,
+      "grad_norm": 0.30545639991760254,
+      "learning_rate": 2.4244930230364894e-08,
+      "loss": 0.2735,
+      "step": 26155
+    },
+    {
+      "epoch": 0.9780124775101059,
+      "grad_norm": 0.5367255210876465,
+      "learning_rate": 2.3837963608100977e-08,
+      "loss": 0.2457,
+      "step": 26160
+    },
+    {
+      "epoch": 0.9781994065004557,
+      "grad_norm": 0.431430459022522,
+      "learning_rate": 2.3434437478269124e-08,
+      "loss": 0.2527,
+      "step": 26165
+    },
+    {
+      "epoch": 0.9783863354908054,
+      "grad_norm": 0.5572644472122192,
+      "learning_rate": 2.3034351980035784e-08,
+      "loss": 0.2407,
+      "step": 26170
+    },
+    {
+      "epoch": 0.9785732644811552,
+      "grad_norm": 0.43069931864738464,
+      "learning_rate": 2.263770725137837e-08,
+      "loss": 0.247,
+      "step": 26175
+    },
+    {
+      "epoch": 0.978760193471505,
+      "grad_norm": 0.4567778706550598,
+      "learning_rate": 2.224450342908746e-08,
+      "loss": 0.2365,
+      "step": 26180
+    },
+    {
+      "epoch": 0.9789471224618548,
+      "grad_norm": 0.3975851535797119,
+      "learning_rate": 2.1854740648769023e-08,
+      "loss": 0.3065,
+      "step": 26185
+    },
+    {
+      "epoch": 0.9791340514522046,
+      "grad_norm": 0.3134264647960663,
+      "learning_rate": 2.1468419044839984e-08,
+      "loss": 0.296,
+      "step": 26190
+    },
+    {
+      "epoch": 0.9793209804425543,
+      "grad_norm": 0.27878788113594055,
+      "learning_rate": 2.1085538750531542e-08,
+      "loss": 0.3218,
+      "step": 26195
+    },
+    {
+      "epoch": 0.9795079094329042,
+      "grad_norm": 0.6269852519035339,
+      "learning_rate": 2.0706099897890296e-08,
+      "loss": 0.3444,
+      "step": 26200
+    },
+    {
+      "epoch": 0.979694838423254,
+      "grad_norm": 0.38990554213523865,
+      "learning_rate": 2.0330102617771575e-08,
+      "loss": 0.2459,
+      "step": 26205
+    },
+    {
+      "epoch": 0.9798817674136038,
+      "grad_norm": 0.3752122223377228,
+      "learning_rate": 1.995754703984609e-08,
+      "loss": 0.2583,
+      "step": 26210
+    },
+    {
+      "epoch": 0.9800686964039536,
+      "grad_norm": 0.6185774207115173,
+      "learning_rate": 1.9588433292598852e-08,
+      "loss": 0.2597,
+      "step": 26215
+    },
+    {
+      "epoch": 0.9802556253943033,
+      "grad_norm": 0.8538332581520081,
+      "learning_rate": 1.9222761503325803e-08,
+      "loss": 0.3271,
+      "step": 26220
+    },
+    {
+      "epoch": 0.9804425543846531,
+      "grad_norm": 0.39858320355415344,
+      "learning_rate": 1.886053179813718e-08,
+      "loss": 0.2141,
+      "step": 26225
+    },
+    {
+      "epoch": 0.9806294833750029,
+      "grad_norm": 0.4068199396133423,
+      "learning_rate": 1.850174430195528e-08,
+      "loss": 0.237,
+      "step": 26230
+    },
+    {
+      "epoch": 0.9808164123653527,
+      "grad_norm": 0.457328200340271,
+      "learning_rate": 1.814639913851557e-08,
+      "loss": 0.3351,
+      "step": 26235
+    },
+    {
+      "epoch": 0.9810033413557026,
+      "grad_norm": 0.2778027653694153,
+      "learning_rate": 1.779449643036668e-08,
+      "loss": 0.3103,
+      "step": 26240
+    },
+    {
+      "epoch": 0.9811902703460523,
+      "grad_norm": 0.6216102242469788,
+      "learning_rate": 1.7446036298869316e-08,
+      "loss": 0.2385,
+      "step": 26245
+    },
+    {
+      "epoch": 0.9813771993364021,
+      "grad_norm": 0.33470118045806885,
+      "learning_rate": 1.710101886419735e-08,
+      "loss": 0.3129,
+      "step": 26250
+    },
+    {
+      "epoch": 0.9815641283267519,
+      "grad_norm": 0.37651658058166504,
+      "learning_rate": 1.6759444245338928e-08,
+      "loss": 0.2858,
+      "step": 26255
+    },
+    {
+      "epoch": 0.9817510573171017,
+      "grad_norm": 0.3364080488681793,
+      "learning_rate": 1.642131256009094e-08,
+      "loss": 0.3334,
+      "step": 26260
+    },
+    {
+      "epoch": 0.9819379863074514,
+      "grad_norm": 0.34844186902046204,
+      "learning_rate": 1.608662392506677e-08,
+      "loss": 0.3211,
+      "step": 26265
+    },
+    {
+      "epoch": 0.9821249152978012,
+      "grad_norm": 0.31410226225852966,
+      "learning_rate": 1.575537845569075e-08,
+      "loss": 0.2541,
+      "step": 26270
+    },
+    {
+      "epoch": 0.982311844288151,
+      "grad_norm": 0.6750185489654541,
+      "learning_rate": 1.5427576266200394e-08,
+      "loss": 0.2703,
+      "step": 26275
+    },
+    {
+      "epoch": 0.9824987732785009,
+      "grad_norm": 0.7728235721588135,
+      "learning_rate": 1.510321746964416e-08,
+      "loss": 0.3094,
+      "step": 26280
+    },
+    {
+      "epoch": 0.9826857022688507,
+      "grad_norm": 0.7537628412246704,
+      "learning_rate": 1.4782302177884789e-08,
+      "loss": 0.2852,
+      "step": 26285
+    },
+    {
+      "epoch": 0.9828726312592004,
+      "grad_norm": 0.5353776812553406,
+      "learning_rate": 1.4464830501597082e-08,
+      "loss": 0.3261,
+      "step": 26290
+    },
+    {
+      "epoch": 0.9830595602495502,
+      "grad_norm": 0.6199917793273926,
+      "learning_rate": 1.4150802550267905e-08,
+      "loss": 0.2453,
+      "step": 26295
+    },
+    {
+      "epoch": 0.9832464892399,
+      "grad_norm": 0.15641337633132935,
+      "learning_rate": 1.3840218432195074e-08,
+      "loss": 0.2621,
+      "step": 26300
+    },
+    {
+      "epoch": 0.9834334182302498,
+      "grad_norm": 0.38754504919052124,
+      "learning_rate": 1.3533078254492905e-08,
+      "loss": 0.2353,
+      "step": 26305
+    },
+    {
+      "epoch": 0.9836203472205995,
+      "grad_norm": 0.6075884699821472,
+      "learning_rate": 1.3229382123082223e-08,
+      "loss": 0.3138,
+      "step": 26310
+    },
+    {
+      "epoch": 0.9838072762109493,
+      "grad_norm": 0.46125859022140503,
+      "learning_rate": 1.2929130142700363e-08,
+      "loss": 0.2389,
+      "step": 26315
+    },
+    {
+      "epoch": 0.9839942052012992,
+      "grad_norm": 0.44463780522346497,
+      "learning_rate": 1.2632322416896715e-08,
+      "loss": 0.2877,
+      "step": 26320
+    },
+    {
+      "epoch": 0.984181134191649,
+      "grad_norm": 0.4823720455169678,
+      "learning_rate": 1.2338959048029398e-08,
+      "loss": 0.2585,
+      "step": 26325
+    },
+    {
+      "epoch": 0.9843680631819988,
+      "grad_norm": 0.3346082866191864,
+      "learning_rate": 1.2049040137273038e-08,
+      "loss": 0.2628,
+      "step": 26330
+    },
+    {
+      "epoch": 0.9845549921723485,
+      "grad_norm": 0.7445553541183472,
+      "learning_rate": 1.1762565784612101e-08,
+      "loss": 0.2554,
+      "step": 26335
+    },
+    {
+      "epoch": 0.9847419211626983,
+      "grad_norm": 0.5069872736930847,
+      "learning_rate": 1.1479536088843112e-08,
+      "loss": 0.3881,
+      "step": 26340
+    },
+    {
+      "epoch": 0.9849288501530481,
+      "grad_norm": 0.4112212061882019,
+      "learning_rate": 1.1199951147574661e-08,
+      "loss": 0.277,
+      "step": 26345
+    },
+    {
+      "epoch": 0.9851157791433979,
+      "grad_norm": 0.4882351756095886,
+      "learning_rate": 1.0923811057227396e-08,
+      "loss": 0.2854,
+      "step": 26350
+    },
+    {
+      "epoch": 0.9853027081337477,
+      "grad_norm": 0.37586453557014465,
+      "learning_rate": 1.0651115913035137e-08,
+      "loss": 0.2478,
+      "step": 26355
+    },
+    {
+      "epoch": 0.9854896371240974,
+      "grad_norm": 1.524063229560852,
+      "learning_rate": 1.0381865809040437e-08,
+      "loss": 0.3372,
+      "step": 26360
+    },
+    {
+      "epoch": 0.9856765661144473,
+      "grad_norm": 0.4818149209022522,
+      "learning_rate": 1.0116060838103458e-08,
+      "loss": 0.2763,
+      "step": 26365
+    },
+    {
+      "epoch": 0.9858634951047971,
+      "grad_norm": 0.2843870222568512,
+      "learning_rate": 9.853701091888656e-09,
+      "loss": 0.2549,
+      "step": 26370
+    },
+    {
+      "epoch": 0.9860504240951469,
+      "grad_norm": 0.44535329937934875,
+      "learning_rate": 9.594786660880317e-09,
+      "loss": 0.3165,
+      "step": 26375
+    },
+    {
+      "epoch": 0.9862373530854966,
+      "grad_norm": 0.2467801719903946,
+      "learning_rate": 9.339317634367017e-09,
+      "loss": 0.225,
+      "step": 26380
+    },
+    {
+      "epoch": 0.9864242820758464,
+      "grad_norm": 0.4127419590950012,
+      "learning_rate": 9.087294100456056e-09,
+      "loss": 0.2825,
+      "step": 26385
+    },
+    {
+      "epoch": 0.9866112110661962,
+      "grad_norm": 0.4206787645816803,
+      "learning_rate": 8.838716146060134e-09,
+      "loss": 0.2727,
+      "step": 26390
+    },
+    {
+      "epoch": 0.986798140056546,
+      "grad_norm": 0.5794749855995178,
+      "learning_rate": 8.59358385690956e-09,
+      "loss": 0.3581,
+      "step": 26395
+    },
+    {
+      "epoch": 0.9869850690468958,
+      "grad_norm": 0.37697890400886536,
+      "learning_rate": 8.351897317541157e-09,
+      "loss": 0.271,
+      "step": 26400
+    },
+    {
+      "epoch": 0.9871719980372456,
+      "grad_norm": 0.5611318945884705,
+      "learning_rate": 8.11365661130603e-09,
+      "loss": 0.3294,
+      "step": 26405
+    },
+    {
+      "epoch": 0.9873589270275954,
+      "grad_norm": 0.2812879979610443,
+      "learning_rate": 7.878861820367345e-09,
+      "loss": 0.3604,
+      "step": 26410
+    },
+    {
+      "epoch": 0.9875458560179452,
+      "grad_norm": 0.3695289194583893,
+      "learning_rate": 7.647513025698105e-09,
+      "loss": 0.2133,
+      "step": 26415
+    },
+    {
+      "epoch": 0.987732785008295,
+      "grad_norm": 0.46063148975372314,
+      "learning_rate": 7.4196103070856005e-09,
+      "loss": 0.2794,
+      "step": 26420
+    },
+    {
+      "epoch": 0.9879197139986448,
+      "grad_norm": 0.3603619337081909,
+      "learning_rate": 7.195153743124739e-09,
+      "loss": 0.2773,
+      "step": 26425
+    },
+    {
+      "epoch": 0.9881066429889945,
+      "grad_norm": 0.515447735786438,
+      "learning_rate": 6.97414341122582e-09,
+      "loss": 0.3515,
+      "step": 26430
+    },
+    {
+      "epoch": 0.9882935719793443,
+      "grad_norm": 0.4539223611354828,
+      "learning_rate": 6.756579387607875e-09,
+      "loss": 0.2721,
+      "step": 26435
+    },
+    {
+      "epoch": 0.9884805009696941,
+      "grad_norm": 0.927209734916687,
+      "learning_rate": 6.542461747304218e-09,
+      "loss": 0.2898,
+      "step": 26440
+    },
+    {
+      "epoch": 0.988667429960044,
+      "grad_norm": 0.33204635977745056,
+      "learning_rate": 6.331790564155782e-09,
+      "loss": 0.3234,
+      "step": 26445
+    },
+    {
+      "epoch": 0.9888543589503938,
+      "grad_norm": 0.4017919898033142,
+      "learning_rate": 6.124565910818891e-09,
+      "loss": 0.2895,
+      "step": 26450
+    },
+    {
+      "epoch": 0.9890412879407435,
+      "grad_norm": 0.3101350665092468,
+      "learning_rate": 5.9207878587574926e-09,
+      "loss": 0.2818,
+      "step": 26455
+    },
+    {
+      "epoch": 0.9892282169310933,
+      "grad_norm": 0.38816362619400024,
+      "learning_rate": 5.7204564782498136e-09,
+      "loss": 0.2629,
+      "step": 26460
+    },
+    {
+      "epoch": 0.9894151459214431,
+      "grad_norm": 0.49870437383651733,
+      "learning_rate": 5.523571838385034e-09,
+      "loss": 0.3173,
+      "step": 26465
+    },
+    {
+      "epoch": 0.9896020749117929,
+      "grad_norm": 0.5789703726768494,
+      "learning_rate": 5.330134007062171e-09,
+      "loss": 0.2963,
+      "step": 26470
+    },
+    {
+      "epoch": 0.9897890039021426,
+      "grad_norm": 0.3744274079799652,
+      "learning_rate": 5.1401430509923075e-09,
+      "loss": 0.3257,
+      "step": 26475
+    },
+    {
+      "epoch": 0.9899759328924924,
+      "grad_norm": 0.4980723559856415,
+      "learning_rate": 4.953599035697476e-09,
+      "loss": 0.2974,
+      "step": 26480
+    },
+    {
+      "epoch": 0.9901628618828423,
+      "grad_norm": 0.71192467212677,
+      "learning_rate": 4.77050202551288e-09,
+      "loss": 0.265,
+      "step": 26485
+    },
+    {
+      "epoch": 0.9903497908731921,
+      "grad_norm": 0.3702625036239624,
+      "learning_rate": 4.590852083582453e-09,
+      "loss": 0.2456,
+      "step": 26490
+    },
+    {
+      "epoch": 0.9905367198635419,
+      "grad_norm": 0.459648996591568,
+      "learning_rate": 4.414649271863303e-09,
+      "loss": 0.2939,
+      "step": 26495
+    },
+    {
+      "epoch": 0.9907236488538916,
+      "grad_norm": 0.2367364466190338,
+      "learning_rate": 4.241893651120155e-09,
+      "loss": 0.2624,
+      "step": 26500
+    },
+    {
+      "epoch": 0.9909105778442414,
+      "grad_norm": 0.43819648027420044,
+      "learning_rate": 4.0725852809342384e-09,
+      "loss": 0.345,
+      "step": 26505
+    },
+    {
+      "epoch": 0.9910975068345912,
+      "grad_norm": 0.3360496461391449,
+      "learning_rate": 3.906724219694402e-09,
+      "loss": 0.2522,
+      "step": 26510
+    },
+    {
+      "epoch": 0.991284435824941,
+      "grad_norm": 0.2503529191017151,
+      "learning_rate": 3.744310524600447e-09,
+      "loss": 0.2847,
+      "step": 26515
+    },
+    {
+      "epoch": 0.9914713648152907,
+      "grad_norm": 0.49219122529029846,
+      "learning_rate": 3.585344251665346e-09,
+      "loss": 0.2354,
+      "step": 26520
+    },
+    {
+      "epoch": 0.9916582938056406,
+      "grad_norm": 0.42718371748924255,
+      "learning_rate": 3.4298254557108e-09,
+      "loss": 0.3146,
+      "step": 26525
+    },
+    {
+      "epoch": 0.9918452227959904,
+      "grad_norm": 0.6982577443122864,
+      "learning_rate": 3.2777541903716845e-09,
+      "loss": 0.2527,
+      "step": 26530
+    },
+    {
+      "epoch": 0.9920321517863402,
+      "grad_norm": 0.5282479524612427,
+      "learning_rate": 3.129130508092715e-09,
+      "loss": 0.2256,
+      "step": 26535
+    },
+    {
+      "epoch": 0.99221908077669,
+      "grad_norm": 0.24284540116786957,
+      "learning_rate": 2.983954460130667e-09,
+      "loss": 0.3069,
+      "step": 26540
+    },
+    {
+      "epoch": 0.9924060097670397,
+      "grad_norm": 0.4188280403614044,
+      "learning_rate": 2.8422260965510485e-09,
+      "loss": 0.303,
+      "step": 26545
+    },
+    {
+      "epoch": 0.9925929387573895,
+      "grad_norm": 0.8391254544258118,
+      "learning_rate": 2.7039454662336484e-09,
+      "loss": 0.2957,
+      "step": 26550
+    },
+    {
+      "epoch": 0.9927798677477393,
+      "grad_norm": 0.3748670518398285,
+      "learning_rate": 2.5691126168669866e-09,
+      "loss": 0.2781,
+      "step": 26555
+    },
+    {
+      "epoch": 0.9929667967380891,
+      "grad_norm": 0.44531556963920593,
+      "learning_rate": 2.437727594949424e-09,
+      "loss": 0.3105,
+      "step": 26560
+    },
+    {
+      "epoch": 0.993153725728439,
+      "grad_norm": 0.43134966492652893,
+      "learning_rate": 2.309790445794713e-09,
+      "loss": 0.2255,
+      "step": 26565
+    },
+    {
+      "epoch": 0.9933406547187887,
+      "grad_norm": 0.471179723739624,
+      "learning_rate": 2.1853012135220065e-09,
+      "loss": 0.302,
+      "step": 26570
+    },
+    {
+      "epoch": 0.9935275837091385,
+      "grad_norm": 0.3568282425403595,
+      "learning_rate": 2.0642599410658493e-09,
+      "loss": 0.2448,
+      "step": 26575
+    },
+    {
+      "epoch": 0.9937145126994883,
+      "grad_norm": 0.5308835506439209,
+      "learning_rate": 1.946666670169517e-09,
+      "loss": 0.2497,
+      "step": 26580
+    },
+    {
+      "epoch": 0.9939014416898381,
+      "grad_norm": 0.5474340319633484,
+      "learning_rate": 1.832521441387236e-09,
+      "loss": 0.3189,
+      "step": 26585
+    },
+    {
+      "epoch": 0.9940883706801878,
+      "grad_norm": 0.31903812289237976,
+      "learning_rate": 1.7218242940841845e-09,
+      "loss": 0.2821,
+      "step": 26590
+    },
+    {
+      "epoch": 0.9942752996705376,
+      "grad_norm": 0.33196452260017395,
+      "learning_rate": 1.6145752664376014e-09,
+      "loss": 0.3051,
+      "step": 26595
+    },
+    {
+      "epoch": 0.9944622286608874,
+      "grad_norm": 0.7084635496139526,
+      "learning_rate": 1.5107743954334565e-09,
+      "loss": 0.1976,
+      "step": 26600
+    },
+    {
+      "epoch": 0.9946491576512372,
+      "grad_norm": 0.3358515799045563,
+      "learning_rate": 1.4104217168708911e-09,
+      "loss": 0.2774,
+      "step": 26605
+    },
+    {
+      "epoch": 0.9948360866415871,
+      "grad_norm": 0.46138903498649597,
+      "learning_rate": 1.3135172653577777e-09,
+      "loss": 0.2587,
+      "step": 26610
+    },
+    {
+      "epoch": 0.9950230156319368,
+      "grad_norm": 0.47722405195236206,
+      "learning_rate": 1.2200610743129394e-09,
+      "loss": 0.2649,
+      "step": 26615
+    },
+    {
+      "epoch": 0.9952099446222866,
+      "grad_norm": 0.30284929275512695,
+      "learning_rate": 1.1300531759694811e-09,
+      "loss": 0.302,
+      "step": 26620
+    },
+    {
+      "epoch": 0.9953968736126364,
+      "grad_norm": 0.768482506275177,
+      "learning_rate": 1.043493601365908e-09,
+      "loss": 0.2393,
+      "step": 26625
+    },
+    {
+      "epoch": 0.9955838026029862,
+      "grad_norm": 0.6742646098136902,
+      "learning_rate": 9.60382380355007e-10,
+      "loss": 0.2402,
+      "step": 26630
+    },
+    {
+      "epoch": 0.995770731593336,
+      "grad_norm": 0.5979500412940979,
+      "learning_rate": 8.807195415994063e-10,
+      "loss": 0.3245,
+      "step": 26635
+    },
+    {
+      "epoch": 0.9959576605836857,
+      "grad_norm": 1.0489237308502197,
+      "learning_rate": 8.045051125726844e-10,
+      "loss": 0.313,
+      "step": 26640
+    },
+    {
+      "epoch": 0.9961445895740355,
+      "grad_norm": 0.5769110918045044,
+      "learning_rate": 7.317391195593714e-10,
+      "loss": 0.3009,
+      "step": 26645
+    },
+    {
+      "epoch": 0.9963315185643854,
+      "grad_norm": 0.7712821364402771,
+      "learning_rate": 6.624215876538387e-10,
+      "loss": 0.262,
+      "step": 26650
+    },
+    {
+      "epoch": 0.9965184475547352,
+      "grad_norm": 0.4885789752006531,
+      "learning_rate": 5.965525407614081e-10,
+      "loss": 0.2446,
+      "step": 26655
+    },
+    {
+      "epoch": 0.996705376545085,
+      "grad_norm": 0.2925719916820526,
+      "learning_rate": 5.341320015994633e-10,
+      "loss": 0.2598,
+      "step": 26660
+    },
+    {
+      "epoch": 0.9968923055354347,
+      "grad_norm": 0.22376123070716858,
+      "learning_rate": 4.751599916941185e-10,
+      "loss": 0.2542,
+      "step": 26665
+    },
+    {
+      "epoch": 0.9970792345257845,
+      "grad_norm": 0.45399028062820435,
+      "learning_rate": 4.1963653138354933e-10,
+      "loss": 0.257,
+      "step": 26670
+    },
+    {
+      "epoch": 0.9972661635161343,
+      "grad_norm": 0.3218282461166382,
+      "learning_rate": 3.675616398157722e-10,
+      "loss": 0.308,
+      "step": 26675
+    },
+    {
+      "epoch": 0.9974530925064841,
+      "grad_norm": 0.3195207118988037,
+      "learning_rate": 3.1893533495086505e-10,
+      "loss": 0.2604,
+      "step": 26680
+    },
+    {
+      "epoch": 0.9976400214968338,
+      "grad_norm": 0.31073522567749023,
+      "learning_rate": 2.737576335576364e-10,
+      "loss": 0.3143,
+      "step": 26685
+    },
+    {
+      "epoch": 0.9978269504871837,
+      "grad_norm": 0.3902645409107208,
+      "learning_rate": 2.320285512169562e-10,
+      "loss": 0.2869,
+      "step": 26690
+    },
+    {
+      "epoch": 0.9980138794775335,
+      "grad_norm": 0.3908766806125641,
+      "learning_rate": 1.9374810231953533e-10,
+      "loss": 0.2434,
+      "step": 26695
+    },
+    {
+      "epoch": 0.9982008084678833,
+      "grad_norm": 0.5233207941055298,
+      "learning_rate": 1.5891630006814595e-10,
+      "loss": 0.29,
+      "step": 26700
+    },
+    {
+      "epoch": 0.9983877374582331,
+      "grad_norm": 0.7636908292770386,
+      "learning_rate": 1.275331564754012e-10,
+      "loss": 0.3399,
+      "step": 26705
+    },
+    {
+      "epoch": 0.9985746664485828,
+      "grad_norm": 0.21469980478286743,
+      "learning_rate": 9.959868236375514e-11,
+      "loss": 0.2314,
+      "step": 26710
+    },
+    {
+      "epoch": 0.9987615954389326,
+      "grad_norm": 0.3713749349117279,
+      "learning_rate": 7.511288736661293e-11,
+      "loss": 0.3933,
+      "step": 26715
+    },
+    {
+      "epoch": 0.9989485244292824,
+      "grad_norm": 0.4366331994533539,
+      "learning_rate": 5.407577992944113e-11,
+      "loss": 0.2541,
+      "step": 26720
+    },
+    {
+      "epoch": 0.9991354534196322,
+      "grad_norm": 0.4484507739543915,
+      "learning_rate": 3.648736730643698e-11,
+      "loss": 0.2271,
+      "step": 26725
+    },
+    {
+      "epoch": 0.9993223824099821,
+      "grad_norm": 0.3740657567977905,
+      "learning_rate": 2.2347655563859096e-11,
+      "loss": 0.3167,
+      "step": 26730
+    },
+    {
+      "epoch": 0.9995093114003318,
+      "grad_norm": 0.5398233532905579,
+      "learning_rate": 1.165664957780699e-11,
+      "loss": 0.2958,
+      "step": 26735
+    },
+    {
+      "epoch": 0.9996962403906816,
+      "grad_norm": 0.46718209981918335,
+      "learning_rate": 4.414353035331331e-12,
+      "loss": 0.2451,
+      "step": 26740
+    },
+    {
+      "epoch": 0.9998831693810314,
+      "grad_norm": 0.2932209074497223,
+      "learning_rate": 6.207684344339271e-13,
+      "loss": 0.2126,
+      "step": 26745
+    },
+    {
+      "epoch": 0.9999953267752413,
+      "step": 26748,
+      "total_flos": 5.335451483137671e+18,
+      "train_loss": 0.2982786745278668,
+      "train_runtime": 62383.2691,
+      "train_samples_per_second": 3.43,
+      "train_steps_per_second": 0.429
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 26748,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.335451483137671e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}